an empirical analysis of real estate investment …17… · real estate investment trusts in asia:...

i

AN EMPIRICAL ANALYSIS OF

REAL ESTATE INVESTMENT TRUSTS

IN ASIA: STRUCTURE, PERFORMANCE

AND STRATEGIC INVESTMENT

IMPLICATIONS

By

ANH KHOI PHAM

A thesis submitted in fulfilment of the requirements for the Degree of Doctor of Philosophy at

the University of Western Sydney

March 2013

ii

UNIVERSITY OF WESTERN SYDNEY

This thesis has been produced by a student of this University to satisfy learning requirements of a

post graduate course.

While University staff may have offered supervision and advice to the author, the University is

unable to accept responsibility for any advice, recommendations, suggestions or conclusions

contained in this piece of work.

The University gratefully acknowledges the co-operation given to the student author since

studies such as these permit students to examine real world issues in a context which provides an

opportunity for meaningful research to be conducted and reported.

iii

DECLARATION

This thesis has been prepared to meet the requirements of a Doctor of Philosophy degree at the

University of Western Sydney.

I declare that this thesis represents my work, except where due acknowledgement is made, and

that it has not been previously included in a thesis, dissertation or report submitted to this

University or to any other institution for a degree, diploma or other qualification.

All possible care has been taken in the preparation of the information in this thesis; however, I

expressly disclaim any liability for the accuracy and sufficiency of the information and under no

circumstances shall either be liable in negligence or otherwise in and arising out of the

preparation or supply of the information in this thesis.

Signed: Anh Khoi Pham

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to show my sincere gratitude to Professor Graeme Newell,

Property Department, School of Business, University of Western Sydney, who has provided

inspirational guidance, support and encouragement throughout my doctoral candidature.

I would also like to thank Dr. Chyi Lin Lee, who is my co supervisor and has made available his

support in numerous valuable suggestions and constructive advice. Special appreciation goes to

Dr. Girijasankar Mallik for helping me with statistical and econometric techniques.

I am indebted to my parents, Van Toan Pham and Thi Lich Tran and my brothers Hoang Hai

Pham and Anh Tu Pham who always kept me away from family responsibilities and encouraged

me to concentrate on my study.

And most of all, I owe my deepest gratitude to my loving wife Ha My Le, who has been

faithfully supportive, encouraging and patient with me during the course of this work.

v

TABLE OF CONTENTS

CHAPTER 1 INTRODUCTION .................................................................................................... 1

1.1 Background ...................................................................................................................... 1

1.1.1 The Evolution of Real Estate Investment Trusts ...................................................... 3

1.1.2 The Development of REIT Markets in Asia ............................................................. 4

1.2 Research Gaps .................................................................................................................. 5

1.3 Statement of the Problem ................................................................................................. 7

1.4 Research Questions .......................................................................................................... 8

1.5 Objectives of the Study .................................................................................................... 8

1.6 Research Methodology ..................................................................................................... 9

1.7 Significance of the Study ............................................................................................... 10

1.8 Thesis Outline ................................................................................................................ 11

CHAPTER 2 AN OVERVIEW OF THE ASIAN REIT MARKETS .......................................... 15

2.1 Overview of REIT Markets Globally ............................................................................. 15

2.2 Significance of Asian Property Markets ........................................................................ 17

2.3 Development of REIT Markets in Asia.......................................................................... 25

2.3.1 Overview ................................................................................................................. 25

2.3.2 Japan ....................................................................................................................... 26

2.3.3 Singapore ................................................................................................................ 35

2.3.4 Hong Kong .............................................................................................................. 43

2.3.5 Malaysia .................................................................................................................. 49

2.3.6 Thailand .................................................................................................................. 55

2.3.7 Taiwan..................................................................................................................... 61

2.3.8 South Korea ............................................................................................................ 66

2.3.9 Prospective Asian REIT Markets............................................................................ 70

2.4 Significance of Asian REITs .......................................................................................... 72

2.5 Summary ........................................................................................................................ 76

vi

CHAPTER 3 LITERATURE REVIEW ....................................................................................... 80

3.1 Introduction .................................................................................................................... 80

3.2 Direct Property in an Investment Portfolio .................................................................... 80

3.2.1 The Performance of Direct Property ....................................................................... 80

3.2.2 The Inflation-hedging Characteristics of Direct Property ...................................... 84

3.2.3 The Diversification Potential of Direct Property .................................................... 86

3.2.4 The Role of Direct Property in a Portfolio.............................................................. 87

3.3 Indirect Property in an Investment Portfolio .................................................................. 89

3.3.1 The Relationship between Indirect and Direct Property ......................................... 89

3.3.2 The Performance of Real Estate Investment Trusts (REITs).................................. 90

3.3.3 The Inflation-hedging Characteristics of REITs ..................................................... 92

3.3.4 The Diversification Potential of REITs .................................................................. 93

3.3.5 The Role of REITs in a Portfolio ............................................................................ 95

3.4 International Property in an Investment Portfolio .......................................................... 96

3.4.1 International Direct Property Investment................................................................ 96

3.4.2 International Indirect Property Investment ............................................................. 98

3.5 Asian Property in an Investment Portfolio ................................................................... 101

3.5.1 Asian Direct Property ........................................................................................... 101

3.5.2 Asian Property Companies ................................................................................... 106

3.5.3 Asian REITs .......................................................................................................... 109

CHAPTER 4 DATA AND METHODOLOGY ......................................................................... 114

4.1 Introduction .................................................................................................................. 114

4.2 Sources of Data ............................................................................................................ 114

4.2.1 Domestic Context.................................................................................................. 114

4.2.2 Regional and Global Context ................................................................................ 115

4.3 Index Construction ....................................................................................................... 117

4.4 Performance Analysis .................................................................................................. 118

4.4.1 Return Measurements ........................................................................................... 118

4.4.2 Risk Measurements ............................................................................................... 119

4.4.3 Risk-adjusted Return Measurements .................................................................... 119

4.4.4 Correlation Coefficient ......................................................................................... 120

vii

4.4.5 Portfolio Theory and Diversification .................................................................... 121

4.5 Regression Analysis ..................................................................................................... 123

4.5.1 Serial Correlation Testing ..................................................................................... 123

4.5.2 Unit Root and Stationarity Testing ....................................................................... 124

4.5.3 The Augmented Dickey-Fuller (ADF) Test .......................................................... 124

4.5.4 The Phillips-Perron (PP) Test ............................................................................... 126

4.5.5 The Kwiatkowski, Phillips, Schmidt and Shin (KPSS) Test ................................ 126

4.5.6 Heteroskedasticity Testing .................................................................................... 127

4.5.7 Volatility Spillover Analysis................................................................................. 127

CHAPTER 5 THE PERFORMANCE AND SIGNIFICANCE OF ASIAN REITs IN

DOMESTIC MIXED-ASSET PORTFOLIOS ........................................................................... 130

5.1 Introduction .................................................................................................................. 130

5.2 Performance of Japan REITs........................................................................................ 131

5.2.1 Risk-adjusted Performance Analysis .................................................................... 131

5.2.2 Mixed-asset Portfolio Analysis ............................................................................. 132

5.2.3 Sub-period Analysis .............................................................................................. 135

5.3 Performance of Singapore REITs ................................................................................ 143




5.4 Performance of Hong Kong REITs .............................................................................. 154




5.5 Performance of Malaysia REITs .................................................................................. 163




5.6 Performance of Thailand REITs .................................................................................. 174



viii


5.7 Performance of Taiwan REITs ..................................................................................... 183




5.8 Performance of Korean REITs ..................................................................................... 194




5.9 Summary and Strategic Implications ........................................................................... 204

CHAPTER 6 THE SIGNIFICANCE OF ASIAN REITs IN REGIONAL AND

INTERNATIONAL INVESTMENT PORTFOLIOS ................................................................ 210

6.1 Introduction .................................................................................................................. 210

6.2 Asian REITs in a Regional and Inter-property Investment Strategy ........................... 210

6.2.1 Performance Analysis ........................................................................................... 210

6.2.2 Portfolio Analysis ................................................................................................. 215


6.3 Asian REITs in a Global Investment Strategy ............................................................. 226

6.3.1 Performance Analysis ........................................................................................... 226

6.3.2 Portfolio Analysis ................................................................................................. 231



CHAPTER 7 THE DYNAMICS OF RETURNS AND VOLATILITY IN ASIAN REIT

MARKETS ................................................................................................................................ 247

7.1 Introduction .................................................................................................................. 247

7.2 Preliminary Analysis .................................................................................................... 248

7.3 Regression Analysis ..................................................................................................... 251

7.3.1 Volatility Clustering Effects ................................................................................. 251

7.3.2 Cross-market Return and Volatility Spillovers ..................................................... 251

7.3.3 Effects of the Global Financial Crisis ................................................................... 257

7.3.4 Diagnostic Checks ................................................................................................ 257

ix


CHAPTER 8 CONCLUSION..................................................................................................... 259

8.1 Introduction .................................................................................................................. 259

8.2 Conclusion of Main Findings ....................................................................................... 259

8.2.1 Risk and Return Characteristics of Asian REITs .................................................. 259

8.2.2 Risk-adjusted Performance of Asian REITs ......................................................... 261

8.2.3 Diversification Benefits of Asian REITs .............................................................. 262

8.2.4 The Role of Asian REITs in a Mixed-asset Portfolio ........................................... 263

8.2.5 Impact of the GFC on Asian REITs ...................................................................... 263

8.2.6 Dynamic Linkages between the Asian REIT Markets .......................................... 264

8.3 Contribution of the Study ............................................................................................. 265

8.3.1 Practical Contributions.......................................................................................... 265

8.3.2 Theoretical Contributions ..................................................................................... 266

8.4 Limitations of Study ..................................................................................................... 267

8.5 Recommendations for Future Research ....................................................................... 268

8.6 Final Comments ........................................................................................................... 268

REFERENCES ........................................................................................................................... 270

APPENDICES ............................................................................................................................ 293

List of Publications ................................................................................................................. 293

List of Conference Presentations ............................................................................................ 293

x

LIST OF TABLES

Table 1.1: Leading REITs in Asia by Country: June 2012 ............................................................. 5

Table 2.1: Top 10 REIT Markets: September 2011 ...................................................................... 17

Table 2.2: Global REIT Markets: September 2011 ...................................................................... 17

Table 2.3: Economic and Demographic Indicators of Major Asian Economies .......................... 19

Table 2.4: Global Competitiveness Index: 2010 – 2012............................................................... 20

Table 2.5: Corruption Perception Index: 2010 – 2011 ................................................................. 21

Table 2.6: Property Transparency Index: 2012............................................................................. 22

Table 2.7: Global Listed Property Securities Markets: 2011 ........................................................ 25

Table 2.8: J-REITs by Property Sector: June 2012....................................................................... 28

Table 2.9: Profile of Listed J-REITs: June 2012 .......................................................................... 30

Table 2.10: S-REITs by Property Sector: June 2012 .................................................................... 36

Table 2.11: Profile of Listed S-REITs: June 2012 ........................................................................ 37

Table 2.12: HK-REITs by Property Sector: June 2012 ................................................................ 44

Table 2.13: Profile of Listed HK-REITs: June 2012 .................................................................... 45

Table 2.14: Profile of Listed M-REITs: June 2012 ...................................................................... 50

Table 2.15: M-REITs by Property Sector: June 2012................................................................... 50

Table 2.16: Thai-REITs by Property Sector: June 2012 ............................................................... 56

Table 2.17: Profile of Listed Thai-REITs: June 2012................................................................... 57

Table 2.18: T-REITs by Property Sector: June 2012 .................................................................... 61

Table 2.19: Profile of Listed T-REITs: June 2012 ....................................................................... 62

Table 2.20: K-REITs by Property Sector: June 2012 ................................................................... 66

Table 2.21: Profile of Listed K-REITs: June 2012 ....................................................................... 67

Table 2.22: Asian REIT Performance: Q1 2012 ........................................................................... 73

Table 2.23: Significance of Asian REITs: June 2012 ................................................................... 74

Table 2.24: Asian REIT Market by Property Sector ..................................................................... 75

Table 2.25: Asian REIT Legislation Overview ............................................................................ 78

Table 4.1: Data Descriptions – Local Context ............................................................................ 116

Table 4.2: Data Descriptions – Regional and Global Context .................................................... 117

Table 5.1: J-REIT Risk-adjusted Performance Analysis: Oct 2001-Apr 2012 ........................... 132

xi

Table 5.2: J-REIT Correlation Matrix: Oct 2001-Apr 2012 ....................................................... 132

Table 5.3: J-REIT Asset Allocations: Oct 2001-Apr 2012 ......................................................... 133

Table 5.4: J-REIT Asset Allocations with Constraints: Oct 2001-Apr 2012 ............................. 135

Table 5.5: J-REIT Sub-period Performance Analysis ................................................................ 136

Table 5.6: J-REIT Sub-period Correlation Analysis................................................................... 138

Table 5.7: S-REIT Risk-adjusted Performance Analysis: Aug 2003-Apr 2012 ......................... 143

Table 5.8: S-REIT Correlation Matrix: Aug 2003-Apr 2012 ..................................................... 144

Table 5.9: S-REIT Asset Allocations: Aug 2003-Apr 2012 ....................................................... 144

Table 5.10: S-REIT Asset Allocations with Constraints: Aug 2003-Apr 2012 .......................... 146

Table 5.11: S-REIT Impact of the Global Financial Crisis ........................................................ 148

Table 5.12: S-REIT Sub-period Correlation Analysis ................................................................ 150

Table 5.13: HK-REIT Risk-adjusted Performance Analysis: Aug 2004-Apr 2012 ................... 154

Table 5.14: HK-REIT Correlation Matrix: Aug 2004-Apr 2012 ................................................ 154

Table 5.15: HK-REIT Asset Allocations: Aug 2004-Apr 2012 ................................................. 155

Table 5.16: HK-REIT Asset Allocations with Constraints: Aug 2004-Apr 2012 ...................... 158

Table 5.17: HK-REIT Impact of the Global Financial Crisis ..................................................... 159

Table 5.18: HK-REIT Sub-period Correlation Analysis ............................................................ 160

Table 5.19: M-REIT Risk-adjusted Performance Analysis: Sep 2005-Apr 2012 ...................... 163

Table 5.20: M-REIT Correlation Matrix: Sep 2005-Apr 2012 ................................................... 164

Table 5.21: M-REIT Asset Allocations: Sep 2005-Apr 2012..................................................... 164

Table 5.22: M-REIT Asset Allocations with Constraints: Sep 2005-Apr 2012 ......................... 166

Table 5.23: M-REIT Impact of the Global Financial Crisis ....................................................... 168

Table 5.24: M-REIT Sub-period Correlation Analysis............................................................... 170

Table 5.25: Thai-REIT Risk-adjusted Performance Analysis: Dec 2003-Apr 2012 .................. 174

Table 5.26: Thai-REIT Correlation Matrix: Dec 2003-Apr 2012............................................... 175

Table 5.27: Thai-REIT Mixed-Asset Portfolio Allocations: Dec 2003-Apr 2012 ..................... 177

Table 5.28: Thai-REIT Impact of the Global Financial Crisis ................................................... 179

Table 5.29: Thai-REIT Sub-period Correlation Analysis ........................................................... 180

Table 5.30: T-REIT Risk-adjusted Performance Analysis: Mar 2006-Apr 2012 ....................... 183

Table 5.31: T-REIT Correlation Matrix: Mar 2006-Apr 2012 ................................................... 184

Table 5.32: T-REIT Asset Allocations: Mar 2006-Apr 2012 ..................................................... 185

xii

Table 5.33: T-REIT Asset Allocations with Constraints: Mar 2006-Apr 2012 .......................... 186

Table 5.34: T-REIT Impact of the Global Financial Crisis ........................................................ 188

Table 5.35: K-REIT Risk-adjusted Performance Analysis: Jul 2006-Apr 2012 ........................ 194

Table 5.36: K-REIT Correlation Matrix: Jul 2006-Apr 2012 ..................................................... 195

Table 5.37: K-REIT Asset Allocations: Jul 2006-Apr 2012....................................................... 197

Table 5.38: K-REIT Asset Allocations with Constraints: Jul 2006-Apr 2012 ........................... 198

Table 5.39: K-REIT Impact of the Global Financial Crisis ........................................................ 200

Table 5.40: K-REIT Sub-period Correlation Analysis ............................................................... 201

Table 5.41: Performance Summary of Asian REITs Versus Stocks .......................................... 204

Table 5.42: Performance Summary of Asian REITs Versus Property Companies .................... 205

Table 5.43: Asian REIT Diversification Summary..................................................................... 206

Table 5.44: Asset Allocation Summary: Average Allocation in Mixed-Asset Portfolios .......... 206

Table 5.45: Return and Risk post-GFC Versus pre-GFC ........................................................... 207

Table 5.46: Asian REIT Risk-adjusted Performance by Sub-period .......................................... 208

Table 5.47: GFC and Asset Allocation: Average Portfolio Allocation in REITs ....................... 209

Table 6.1: Asian REITs Risk-adjusted Performance Analysis - Common Sample: Jul 2006-Apr

2012............................................................................................................................................. 211

Table 6.2: Asian REITs Risk-adjusted Performance Analysis - Individual Samples: Aug 2001-

Apr 2012* ................................................................................................................................... 213

Table 6.3: Asian REIT Correlation Matrix: Jul 2006-Apr 2012 ................................................. 214

Table 6.4: Asian REIT Allocations - Common Samples: Jul 2006-Apr 2012 ............................ 216

Table 6.5: Asian REIT Allocations - Individual Samples .......................................................... 217

Table 6.6: Asian REIT Allocations with Constraints: July 2006 to April 2012 ......................... 219

Table 6.7: Asian REIT Sub-period Performance Analysis: ........................................................ 221

Table 6.8: Asian REIT Sub-period Correlation Analysis ........................................................... 224

Table 6.9: Performance Analysis – REITs vs. Shares: Jul 2006-Apr 2012 ................................ 227

Table 6.10: Correlation Matrix - Whole Period: : Jul 2006-Apr 2012 ....................................... 230

Table 6.11: Global Portfolio Asset Allocations: : Jul 2006-Apr 2012 ....................................... 231

Table 6.12: Global Portfolio Asset Allocations with Constraints: : Jul 2006-Apr 2012 ............ 232

Table 6.13: Regional Asset Allocations: : Jul 2006-Apr 2012 ................................................... 234

Table 6.14: Regional Asset Allocations: : Jul 2006-Apr 2012 ................................................... 235

xiii

Table 6.15: Sub-period Performance Analysis ........................................................................... 237

Table 6.16: Sub-period Correlation Analysis ............................................................................. 239

Table 6.17: Performance Summary - pan-Asia REITs Versus Stocks ....................................... 244

Table 7.1: Descriptive Statistics of Asia REIT Index Returns: June 15, 2006-April 27, 2012 .. 249

Table 7.2: Plots of REIT Index Returns: June 15, 2006-April 27, 2012 .................................... 250

Table 7.3: Unit Root Tests for Asia REIT Return Series: June 15, 2006-April 27, 2012 .......... 251

Table 7.4: Test for Serial Correlation and Heteroscedasticity: June 15, 2006-April 27, 2012 ... 251

Table 7.5: EGARCH Model: Mean and Volatility Spillovers between the Asian REIT Markets:

June 15, 2006-April 27, 2012 ..................................................................................................... 254

xiv

LIST OF FIGURES

Figure 1.1: Aggregate Value of Asset Classes, US Early 2000s (in trillions of US$).................... 2

Figure 1.2: Forms of Property Investment Vehicles ....................................................................... 3

Figure 1.3: Overall Structure of the Thesis ................................................................................... 14

Figure 2.1: Evolution of US REIT Market: 1962 – 2011 ............................................................. 16

Figure 2.2: Investable Property by Region: 2011 – 2031 ............................................................. 23

Figure 2.3: Asia Property Markets Relative to GDP: 2011 .......................................................... 24

Figure 2.4: Evolution of Asian REIT Markets.............................................................................. 26

Figure 2.5: Evolution of J-REIT Industry: October 2001 – April 2012 ....................................... 28

Figure 2.6: J-REIT Segmentation ................................................................................................. 29

Figure 2.7: Investment Corporation Structure for Listed J-REITs ............................................... 32

Figure 2.8: S-REIT Segmentation................................................................................................. 36

Figure 2.9: Typical Structure for a Listed S-REIT ....................................................................... 38

Figure 2.10: S-REIT Unitholder Ownership ................................................................................. 39

Figure 2.11: S-REIT Overseas Portfolio by Country.................................................................... 40

Figure 2.12: HK-REIT Segmentation ........................................................................................... 44

Figure 2.13: Typical Structure of a Listed HK-REIT ................................................................... 46

Figure 2.14: M-REIT Segmentation ............................................................................................. 51

Figure 2.15: Typical Structure for an M-REIT ............................................................................. 52

Figure 2.16: Thai-REIT Segmentation ......................................................................................... 56

Figure 2.17: TW-REIT Segmentation ........................................................................................... 62

Figure 2.18: K-REIT Segmentation .............................................................................................. 67

Figure 2.19: Evolution of Asian REIT Markets: October 2001 – June 2012 ............................... 73

Figure 2.20: Asian REIT Market Share ........................................................................................ 75

Figure 2.21: Asian REIT Market Break-down by Property Type ................................................ 76

Figure 4.1: Hypothetical Mean-variance Efficient Frontier ....................................................... 123

Figure 5.1: J-REIT Asset Allocation Diagram: Oct 2001-Apr 2012 .......................................... 133

Figure 5.2: J-REIT Efficient Frontiers: Oct 2001-Apr 2012 ...................................................... 134

Figure 5.3: J-REIT Asset Allocation Diagrams with Constraints: Oct 2001-Apr 2012 ............. 135

Figure 5.4: J-REIT 12-Month Rolling Volatility Analysis: Sep 2002-Apr 2012 ....................... 137

xv

Figure 5.5: J-REIT 12-Month Rolling Correlation Analysis: Sep 2002-Apr 2012 .................... 139

Figure 5.6: J-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC: Oct

2001-Feb 2008 ............................................................................................................................ 141

Figure 5.7: J-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012 ..................................................................................................................... 142

Figure 5.8: S-REIT Asset Allocation Diagram: Aug 2003-Apr 2012 ........................................ 145

Figure 5.9: S-REIT Efficient Frontiers: Aug 2003-Apr 2012..................................................... 146

Figure 5.10: S-REIT Asset Allocation Diagrams with Constraints: Aug 2003-Apr 2012 ......... 147

Figure 5.11: S-REIT 12-Month Rolling Volatility Analysis: Aug 2004-Apr 2012 .................... 149

Figure 5.12: S-REIT 12-Month Rolling Correlation Analysis: Aug 2004-Apr 2012 ................. 151

Figure 5.13: S-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Aug 2003-Feb 2008 .................................................................................................................... 152

Figure 5.14: S-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012 ..................................................................................................................... 153

Figure 5.15: HK-REIT Asset Allocation Diagram: Aug 2004-Apr 2012 ................................... 156

Figure 5.16: HK-REIT Efficient Frontiers: Aug 2004-Apr 2012 ............................................... 157

Figure 5.17: HK-REIT Asset Allocation Diagrams with Constraints: Aug 2004-Apr 2012 ...... 158

Figure 5.18: HK-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Aug 2004-Feb 2008 .................................................................................................................... 161

Figure 5.19: HK-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-

GFC: Apr 2009-Apr 2012 ........................................................................................................... 162

Figure 5.20: M-REIT Asset Allocation Diagram: Sep 2005-Apr 2012 ...................................... 165

Figure 5.21: M-REIT Efficient Frontiers: Sep 2005-Apr 2012 .................................................. 166

Figure 5.22: M-REIT Asset Allocation Diagrams with Constraints: Sep 2005-Apr 2012 ......... 167

Figure 5.23: M-REIT 12-Month Rolling Volatility Analysis: Sep 2006-Apr 2012 ................... 169

Figure 5.24: M-REIT 12-Month Rolling Correlation Analysis: Sep 2006-Apr 2012 ................ 171

Figure 5.25: M-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Sep 2005-Feb 2008 ..................................................................................................................... 172

Figure 5.26: M-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012 ..................................................................................................................... 173

Figure 5.27: Thai-REIT 12-Month Rolling Volatility Analysis: Dec 2004-Apr 2012 ............... 175

xvi

Figure 5.28: Thai-REIT 12-Month Rolling Correlation Analysis: Dec 2004-Apr 2012 ............ 176

Figure 5.29: Thai-REIT Asset Allocation Diagram: Dec 2003-Apr 2012 .................................. 177

Figure 5.30: Thai-REIT Efficient Frontiers: Dec 2003-Apr 2012 .............................................. 178

Figure 5.31: Thai-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-

GFC: Dec 2003-Feb 2008 ........................................................................................................... 181

Figure 5.32: Thai-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-

GFC: Apr 2009-Apr 2012 ........................................................................................................... 182

Figure 5.33: T-REIT Efficient Frontiers: Mar 2006-Apr 2012................................................... 184

Figure 5.34: T-REIT Asset Allocation Diagram: Mar 2006-Apr 2012 ...................................... 185

Figure 5.35: T-REIT Asset Allocation Diagrams with Constraints: Mar 2006-Apr 2012 ......... 186

Figure 5.36: T-REIT 12-Month Rolling Volatility Analysis: Mar 2007-Apr 2012 .................... 189

Figure 5.37: T-REIT Sub-period Correlation Analysis .............................................................. 190

Figure 5.38: T-REIT 12-Month Rolling Correlation Analysis: Mar 2007-Apr 2012 ................. 191

Figure 5.39: Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC: Mar 2006-

Feb 2008...................................................................................................................................... 192

Figure 5.40: Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC: Apr

2009-Apr 2012 ............................................................................................................................ 193

Figure 5.41: K-REIT 12-Month Rolling Volatility Analysis: Jul 2007-Apr 2012 ..................... 195

Figure 5.42: K-REIT 12-Month Rolling Correlation Analysis: Jul 2007-Apr 2012 .................. 196

Figure 5.43: K-REIT Asset Allocation Diagram: Jul 2007-Apr 2012 ........................................ 197

Figure 5.44: K-REIT Asset Allocation Diagrams with Constraints: Jul 2007-Apr 2012 ........... 198

Figure 5.45: K-REIT Efficient Frontiers: Jul 2007-Apr 2012 .................................................... 199

Figure 5.46: K-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Jul 2006-Feb 2008....................................................................................................................... 202

Figure 5.47: K-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012 ..................................................................................................................... 203

Figure 6.1: Asian REIT Risk and Return Profile - Common Sample: Jul 2006-Apr 2012 ........ 212

Figure 6.2: Asian REIT Risk and Return Profile - Individual Samples: Aug 2001-Apr 2012* . 214

Figure 6.3: Regional Correlations REITs vs. Shares: Jul 2006-Apr 2012 .................................. 215

Figure 6.4: Asian REIT Efficient Frontier Compositions - Common Samples: July 2006 to April

2012............................................................................................................................................. 216

xvii

Figure 6.5: Asian REIT Efficient Frontier Composition - Individual Samples .......................... 217

Figure 6.6: Efficient Frontiers - Common vs. Individual Samples ............................................. 218

Figure 6.7: Asian REIT Efficient Frontier Composition with Constraints: July 2006 to April 2012

..................................................................................................................................................... 219

Figure 6.8: Risk and Return Profile of Asian REITs over Sub-periods...................................... 222

Figure 6.9: Asset Allocation Sub-period .................................................................................... 225

Figure 6.10: Efficient Frontiers Pre-GFC vs. Post-GFC Period ................................................. 226

Figure 6.11: Risk and Return Profile of Asian REITs: Jul 2006-Apr 2012 ................................ 228

Figure 6.12: Global Asset Allocation: : Jul 2006-Apr 2012 ....................................................... 232

Figure 6.13: Global Asset Allocation with Constraints: : Jul 2006-Apr 2012 ........................... 233

Figure 6.14: Regional Asset Allocations: Jul 2006-Apr 2012 .................................................... 234

Figure 6.15: Regional Asset Allocations: Jul 2006-Apr 2012 .................................................... 235

Figure 6.16: Sub-period Asset Allocation Diagrams .................................................................. 242

Figure 6.17: Summary Performance Post-GFC vs. Pre-GFC ..................................................... 245

Figure 6.18: Summary Sub-period Performance of REITs vs. Stocks ....................................... 246

xviii

LIST OF ABBREVIATIONS

¥: Japanese Yen.

₱: Philippine Peso.

₩: Korean Won.

: Thai Baht.

ABS: Asset-backed Securities.

ADF: Augmented Dickey-Fuller. ,

AIC: Akaike Information Criterion.

APT: Arbitrage Pricing Theory.

AR: Autoregressive.

ARCH: Autoregressive Conditional

Heteroskedasticity.

ASX: Australian Securities Exchange.

bln: billion.

BOJ: Bank of Japan.

CAPM: Capital Asset Pricing Model.

CBRC: China Banking Regulatory

Commission.

CCIS: Code on Collective Investment

Schemes.

CPI: Capital Perspectives Index.

C-REIT: China REIT.

CSRC: China Securities Regulatory

Commission.

CV: Coefficient of Variation.

DA: Developed Asia.

DF: Dickey-Fuller.

DJTM: Dow Jones Total Market.

EA: Emerging Asia.

EGARCH: Exponential Generalized

Autoregressive Conditional

Heteroskedastic.

Ex: excluding.

FASA: Financial Asset Securitization Act.

FLM: Factor Loading Model.

FSA: Financial Services Agency.

FSC: Financial Supervisory Commission.

GCI: Global Competitiveness Index.

GDP: Gross Domestic Product.

GFC: Global Financial Crisis.

GPR: Global Property Research.

GTI: Global Transparency Index.

GUI: Government Uniform Invoice.

HK: Hong Kong.

HKEx: Hong Kong Stock Exchange.

HK-REIT: Hong Kong REIT.

IFA: Internal Floor Area.

IFC: International Financial Centres.

IPD: Investment Property Databank.

ITL: Investment Trust Law.

JLL: Jones Lang LaSalle.

JLW: Jones Lang Wootton.

JP: Japan.

KO: South Korea.

KPSS: Kwiatkowski, Phillips, Schmidt and

Shin.

K-REIT: South Korean REIT.

xix

KRX: Korea Stock Exchange. ; Korean

Stock Exchange.

LB: Ljung-Box.

LM: Lagrange multiplier.

MAS: Monetary Authority of Singapore.

ME: manager-entrusted.

ML: Malaysia.

MLIT: Ministry of Land, Infrastructure,

Transport and Tourism.

mln: million.

MOLTM: Ministry of Land, Transport and

Maritime Affairs.

MPT: Modern Portfolio Theory.

M-REIT: Malaysian REIT.

MVM: Macrovariable Model.

NAREIT: National Association of Real

Estate Investment Trusts.

NAV: Net Asset Value.

NCREIF: National Council of Real Estate

Investment Fiduciaries.

NPI: NCREIF Property Index.

NT$: New Taiwan Dollar.

OLS: Ordinary Least Squares.

PA: Pan-Asia.

PAexJP: Pan-Asia excluding Japan.

PBC: People’s Bank of China.

PFPO: Property Funds for Public Offering.

PP: Phillips-Perron.

P-REIT: Philippine REIT.

PSEC: Securities and Exchange

Commission of the Philippines.

REAT: Real Estate Asset Trust.

REICA: Real Estate Investment Company

Act.

REIT: Real Estate Investment Trust.

REMF: Real Estate Mutual Fund.

REPEF: Real Estate Private Equity Fund.

RESA: Real Estate Securitisation Act.

RPGT: Real Property Gains Tax.

S&P: Standard & Poor's.

SC: Securities Commission.

SEBI: Securities and Exchange Board of

India.

SECP: Securities and Exchange

Commission of Pakistan.

SET: Stock Exchange of Thailand.

SFA: Securities and Future Act.

SFC: Securities and Future Commission.

SG: Singapore.

SM: self-managed.

SPV: Special Purpose Vehicles.

S-REIT: Singaporean REIT.

STMA: Special Taxation Measures Act.

T-Bill: Treasury Bill.

TI: Transparency International.

TL: Thailand.

TR: total return.

T-REIT: Taiwanese REIT.

TSE: Tokyo Stock Exchange.

TSEC: Securities and Exchange

Commission of Thailand.

TW: Taiwan.

xx

TWSE: Taiwanese Stock Exchange.

UK: United Kingdom.

US$: US Dollar.

UWS: University of Western Sydney.

VBA: Visual Basic for Applications.

WEF: World Economic Forum.

xxi

EXECUTIVE SUMMARY

Since the early 2000s, Real Estate Investment Trusts (REITs) have become an important

property investment vehicle in Asia. Despite strong interest from property investors, research on

Asian REIT markets has been thin and modest. This thesis provides an empirical analysis on the

performance of Asian REITs and their significance in a mixed-asset portfolio. Strategic

investment implications for both local and international property investors are also identified.

The analysis covers all seven Asian REIT markets of Japan, Singapore, Hong Kong, Malaysia,

Taiwan, Thailand and South Korea over various periods from September 2001 to April 2012.

Over the course of this thesis, three major studies are carried out. The first study investigates the

performance and significance of Asian REITs in a domestic mixed-asset portfolio. The second

study extends the analysis to a regional and international investment context, while the third one

explores the dynamics of returns and volatility between Asian REIT markets.

The empirical results from the individual markets indicate that Asian REITs delivered better

absolute returns and lower risk than both shares and property companies in most Asian REIT

markets. As a result, Asian REITs outperformed shares and property companies in a risk-

adjusted basis in most markets. In particular, REITs outperformed shares in Japan, Singapore,

Hong Kong and Taiwan and property companies in all Asian markets except South Korea. The

study also found a low or negative correlation between Asian REITs and the other asset classes

of shares, property companies and bonds, suggesting strong diversification benefits. The asset

allocation analysis found that the inclusion of REITs in a mixed-asset portfolio would enhance

portfolio returns. As a result, Asian REITs play a significant role in the efficient portfolios for

Asian markets, accounting for about one-third of the optimal domestic mixed-asset portfolios.

Within a global context, Asian REIT markets also delivered stronger risk-adjusted performance

in comparison to the developed markets of the US and Australia over the study period. This

highlights the sound investment credentials of Asian REITs underpinned by low borrowing

levels, steady earning streams and favourable property market fundamentals. With regards to a

pan-Asia asset strategy for global investors, the analysis reaffirmed the significant role of pan-

xxii

Asia REITs in the asset mix. On average, the optimum portfolio would be composed of 61%

Asian REITs, 31% US shares and 8.2% Asian shares.

The sub-period analysis has revealed the significant impact of the GFC during March 2008-

March 2009 on Asian REIT performance. The crisis caused Asian REITs’ correlations with other

asset classes to increase significantly, leading to a significant loss of portfolio diversification

benefits. The post-GFC period since April 2009 has seen Asian REITs gaining back some level

of these diversification benefits, as the correlations with shares and property companies have

been falling. The strong recovery of Asian REIT performance has also been witnessed in both

the return and risk dimensions. Most Asian REIT markets have registered improved returns and

lower risk post-GFC versus the pre-GFC period.

The econometric analysis of Asian REIT markets’ linkages showed that Asian REIT returns

tended to transmit from the developed markets to the emerging markets in Asia. This suggests

that investors can take advantage of available information from the more dominant markets to

predict movements of REIT returns in the smaller markets. On the other hand, the volatility

spillovers among Asian REIT markets were found to be more multidirectional. Volatility

spillovers were from the developed markets of Hong Kong and Singapore to the other markets,

but there were some volatility feedback effects from the emerging REIT markets to the

developed REIT markets as well. This indicates that the emerging markets would also contain

useful information to forecast the volatility in the developed markets.

The original contributions of this thesis to the existing literature on the discipline of property

investment are twofold; practical and theoretical. The main practical contribution was the

extension of the current knowledge base to the new and under-researched Asian REIT markets.

This would be of great interest to global property investors who are seeking access to the vast

potential of property markets in Asia. The theoretical contribution of this research was made in

the synthesising of previous works and theories from different academic disciplines, reaffirming

their applicability to property research and contributing to the ongoing development of property

investment as a discipline. New Asian REIT series were also developed.

xxiii

The quality of this research has been further externally validated by three publications in leading

property research journals in the US and Australia; namely:

1. Pham, AK 2011, 'The performance of Thai-REITs in a mixed-asset portfolio', Pacific

Rim Property Research Journal, Vol. 17, No. 2, pp. 197-214.

2. Pham, AK 2011, 'The significance and performance of South Korean REITs in a mixed-

asset portfolio', Journal of Real Estate Literature, Vol. 19, No. 2, pp. 373-90.

3. Pham, AK 2012, 'The dynamics of return and volatility in the emerging and developed

Asian REIT markets', Journal of Real Estate Literature, Vol. 20, No. 1, pp. 79-96.

1

1 CHAPTER 1

INTRODUCTION

1.1 Background

Property investment has grown dramatically over the last few decades and has become one of the

core asset classes in the investment world. As of the early 2000s, property is the second largest

investment sector in the US, next only to fixed-income securities, but larger than shares and the

money market (Geltner et al. 2001). The total value of the property market in the US was

approximately US$17 trillion, making up over a quarter of the US$64 trillion investable asset as

of early 2000s (Figure 1.1). Globally, the investment-grade property market is estimated to be

approximately $27 trillion as of 2011 (Pramerica 2012).

Property has an essential part in a balanced investment portfolio as it offers investors unique

investment characteristics which differ fundamentally from the other main asset classes of shares

and bonds. One of the main reasons to invest in property is for portfolio diversification benefits.

As property tends to move in different cycles to the other major asset class, the inclusion of

property in the mixed-asset portfolio would potentially reduce risk and enhance portfolio returns.

In addition, property investment is considered to offer several other key benefits including

income stability, capital growth, tax reduction and an effective hedge against inflation.

There are two basic categories of property; residential and commercial property. Residential

property is mainly held by individual and small investors, while commercial property is mostly

dominated by large-scale institutions. However, with the development of new property

investment vehicles, recent years have seen an increase in the participation of retail investors in

the commercial property market.

Figure 1.2 shows the various forms of commercial property investment vehicles. Even though

different markets might have different market structures, generally investors can invest either

directly or indirectly in property. Direct property investment refers to the purchasing of direct

ownership of the physical property. This type of investment is generally capital-intensive and

2

requires active management skills. Another way by which investors can invest in property is

through indirect investment vehicles. Rather than owning the entire property, investors can hold

fractional ownership of the underlying properties through property securities or an unlisted

property fund. Indirect property investment does not require property management and can be

more affordable to retail investors.

Indirect property investment can be either privately held (unlisted) or publicly traded (listed) on a

stock market. Unlisted property investment includes products such as unlisted wholesale funds,

direct private funds, property syndicates and unlisted retail funds. These products however are

less liquid and less transparent, as transaction prices are not reported frequently. As a result, they

are more commonly purchased by high net-worth individuals and institutions eg: pension funds,

sovereign wealth funds. In contrast, the listed property market is more liquid and informationally

efficient, as securities are traded on a more frequent basis. The common types of listed property

products were property securities funds, property companies and Real Estate Investment Trusts

(REITs), with the last one emerging as the predominant property vehicle for both private and

institutional investors.

Figure 1.1: Aggregate Value of Asset Classes, US Early 2000s (in trillions of US$)

Source: Geltner et al. (2001)

$25.65 40.28%

$17.07 26.81%

$16.95 26.63%

$4.00 6.28%

Bonds & Mortgages

Real Estate Equity*

Stocks

Cash

*Excludes value in mortgages and corporate real estate.

3

Figure 1.2: Forms of Property Investment Vehicles

Source: Newell (2008)

1.1.1 The Evolution of Real Estate Investment Trusts

In the 1960s, the US Congress commenced the creation of Real Estate Investment Trusts

(REITs) which would provide affordable access to investment in commercial property to all

investors. Prior to the establishment of the US REIT regime, the commercial property market

was dominated by high net-worth individuals or institutional investors. US REITs allow retail

investors to purchase partial ownership of large-scale income-producing properties. In addition,

US REITs would also offer investors tax advantages as they are pass-through vehicles, meaning

that income tax is exempted at the corporate level.

Over the last fifty years, investment in US REITs has increased, with US REITs becoming the

most important indirect investment vehicle in property. As of December 2011, US REITs made

up 88.5% of the total listed property market in the US (Macquarie 2011) . The last decade has

seen a rapid expansion of REIT structures worldwide. By 2011, there were 22 countries and

territories around the world that have established REIT regimes. Despite the major impact of the

Global Financial Crisis (GFC), the global REIT market has grown dramatically in the last

decade. The global REIT market has grown to a total market capitalisation of US$700 billion,

with much of the recent growth being in Asia. The region has been the nucleus of the global

Commercial Property

Investment

Direct Property

Core Property Value-added

Property Opportunistic

Property

Indirect Property

Listed Property

Securities

Property Securities

Funds

Property Companies

REITs

Unlisted Property Securities

Unlisted Wholesales

Funds

Direct Private Funds

Property Syndicates

Unlisted Retail Funds

4

economic growth in the last decade, recovering strongly from the GFC and remaining solid amid

the debt crisis in the US and Europe. This strong economic growth combined with rapid

urbanisation have been a major stimulus to the property markets in Asia (Newell 2012).

1.1.2 The Development of REIT Markets in Asia

The steady growth in the Asian property market has been the major driver for the development of

REITs in Asia over the last ten years. In November 2000, Japan was the first country in Asia to

establish a REIT market. Since then, there have been six other REIT markets established in the

region. These are Singapore, South Korea, Thailand, Taiwan, Malaysia and Hong Kong. Other

Asian nations such as China, India, Pakistan and the Philippines are also in the processing of

implementing their own REIT regimes.

In just over a decade, the number of Asian REITs has grown to 133 REITs across the seven

Asian REIT markets with a combined market capitalisation of over US$100 billion as of June

2012. Asian REITs has become a major component of the global property portfolio, accounting

for 12.5% of the global REIT market in 2011 (Macquarie 2011). Japan, Singapore and Hong

Kong are among the top 10 REIT markets globally by total market capitalization. Asian REITs

have also delivered strong performance over the last five years, significantly outperforming the

major REIT markets of the US, UK and Australia (S&P 2012). The development of Asian REITs

is further supported by favourable changes in the regulatory structures in recent years. This has

been received positively by both local and international investors.

Table 1.1 shows the top REITs by market capitalisation in each of the seven Asian REIT market

as of June 2012. The largest REITs in Asia were Link REIT (ranked #1 and had US$8.6 billion

in market value), Nippon Building Fund (#2 and US$5.3 billion), Japan REIT (#3 and US$4.7

billion), Capita Mall (#4 and US$4.6 billion) and Accendas (#5 and US$3.5 billion). Three Asian

REITs were in the top 50 REITs globally (Newell, 2012). Asian REITs invest across a wide

range of property sectors including office, retail, industrial, lodging and residential sectors.

Diversified REITs were also available in most markets. There were also REITs engaging in

specialised assets such as healthcare and plantation. In addition, Asia was a pioneer in the

establishment of specialised REIT products such as Islamic REITs and RMB-denominated

5

REITs. This diverse profile for Asian REITs offer investors many investment opportunities and

significant property exposure.

1.2 Research Gaps

Despite the significant interest of property investors in Asian REITs, empirical research

regarding the dynamics of Asian REIT markets is still very limited compared to the more

developed REIT markets of the US and Australia. A comprehensive review of the literature on

Table 1.1: Leading REITs in Asia by Country: June 2012

Property No. of Mkt Cap. Asia

REIT Name Sector Properties (US$ mln) Rank

Japan NIPPON BUILDING FD. Office 67 5,346 #2

JAPAN REIT Office 57 4,676 #3

JAPAN RETAIL FD. INV. Retail 70 2,859 #7

Singapore CAPITAMALL TRUST Retail 16 4,634 #4

ASCENDAS REAL ESTATE INVESTMENT Industrial 95 3,476 #5

CAPITACOMMERCIAL TRUST Retail 9 2,593 #8

Hong Kong LINK REIT Retail 180 8,631 #1

HUI XIAN REIT Diversified 2 2,880 #6

CHAMPION REIT Office 2 2,000 #11

Malaysia PAVILION REIT Retail 2 1,135 #26

SUNWAY REIT Diversified 8 1,087 #27

CAPITAMALLS MALAYSIA TRUST Retail 4 839 #40

Thailand CPN RETAIL GROWTH LEASEHOLD Retail 3 729 #44

TESCO LOTUS RETAIL GROWTH Retail 17 666 #49

SAMUI AIRPORT PROPERTY FUND Industrial 1 333 #75

Taiwan CATHAY NO.1 REIT. Diversified 3 828 #41

SHIN KONG NO.1 REIT. Diversified 6 487 #56 CATHAY NO.2 REIT.TRUST Office 3 347 #72

South Korea KOCREF 15 CR-REIT Office 1 52 #107

KOCREF REIT 8 Office 2 41 #110

TRUS Y 7 REIT Diversified NA 34 #113

Source: Author’s compilation from DataStream

6

Asian property markets revealed a number of research gaps. First of all, most of the previous

studies on Asian property securities has been on property companies (Addae-Dapaah & Kion

1996; Garvey et al. 2001; Lee et al. 2011b; Liow 1997, 1998, 2000, 2001, 2007; Liow 2011;

Liow & Adair 2009; Liow & Sim 2006; Newell et al. 2007; Newell et al. 2009b; Nguyen 2010,

2011, 2012) rather than on REITs. As REITs and property companies are two distinct asset

classes, it is important to understand the different risk and return characteristics between the two

investment products. There has also been a lack of studies that compare the risk-adjusted

performance between property companies and REITs in Asia.

Another gap in the literature is a lack of comprehensive studies addressing Asian REITs in a pan-

Asian and international context. Among a few number of studies on Asian REITs, the analysis

were conducted mostly for individual countries such as Singapore (Chiang et al. 2008), Hong

Kong (Newell et al. 2010), Malaysia (Lee & Ting 2009; Newell & Osmadi 2009, 2010), Japan

(Newell & Peng 2012), Taiwan (Newell & Peng 2012), Thailand (Pham, 2011a) and South

Korea (Pham, 2011b). Nevertheless, these studies covered various time periods and had different

methods, which made it impossible to compare the performance between Asian REIT markets.

The third area which have not been analysed in the past is on the dynamic linkages between the

Asian REIT markets. There were a number of studies on the market integration in Asian property

markets (Garvey et al. 2001; Liow & Adair 2009). However, these studies were on listed

property companies and not on REITs. Li and Yung (2007) studied the transmissions of property

returns from the Atlantic REIT markets of the US and UK to the Asia Pacific REIT markets of

Australia, Hong Kong, Japan and Singapore. There are no studies providing empirical findings

on the dynamic interactions between the Asian REIT markets within a regional context.

Moreover, there were not many studies on the effects of the recent GFC on Asian REITs. This

was owing to the fact that most of the earlier studies on Asian property were conducted prior to

the GFC period. The GFC has caused substantial damage to the global financial market and

significantly dampened investor confidence in the property market. Therefore, it is important to

understand how the GFC affected the Asian listed property market in general and the Asian

REIT market in particular.

7

Last but not least, this study will also address another gap in the literature which is the lack of

research on volatility spillovers in REIT returns. Most of the previous studies on market

integration focused on linkages at the first moment rather than second moment (volatility). The

use of quantitative and econometric techniques in studying REIT returns in Asia has also been

limited in the past. Most of the previous studies that employed regression techniques in

modelling REIT returns were on more developed markets such as the US and Australia.

1.3 Statement of the Problem

Given the current gaps in the body of literature regarding Asian REITs, the general problems that

will be addressed by this thesis are:

There is a lack of available market benchmarks for Asian REIT markets. Specifically,

REIT indices for Thailand, South Korea and Malaysia were either unavailable or

inadequate for analysis. In addition, there is a lack of a pan-Asia composite REIT index

that tracks the performance of all listed Asian REITs.

The risk and return attributes of Asian REIT markets, in particular the emerging markets,

have not been studied extensively in the past.

The significance and benefits of Asian REITs within a regional and global investment

framework has not been investigated.

The optimal levels of Asian REITs in the local and global mixed-asset portfolios are yet

to be explored.

The effects of the GFC on Asian REIT markets have not been thoroughly examined and

reported.

There is a lack of understanding about the dynamic linkages between the developed and

emerging Asian REIT markets.

There is a lack of research on volatility spillover effects between the REIT returns.

The use of advanced econometric techniques to examine Asian REIT returns has been

limited.

8

1.4 Research Questions

From the statement of the problem, a number of research questions are defined. These general

questions are as follows:

What are the risk and return characteristics of Asian REITs?

What is the risk-adjusted performance of Asian REITs comparing to shares, bonds and

property companies?

Do Asian REITs provide diversification benefits to local and international investors?

What are the optimum levels of Asian REITs in mixed-asset portfolios in both a domestic

and international context?

What are the impacts of the GFC on Asian REITs regarding risk, return and

diversification potential?

What is the nature of the dynamic linkages between Asian REIT markets at both the first

and second moment?

1.5 Objectives of the Study

The primary goal of this study is to provide a quantitative study on the significance and risk-

adjusted performance of Asian REITs in order to identify the potential role of Asian REITs in a

mixed-asset portfolio for both local and international property investors. Corresponding to the

general research gaps and research questions, the research objectives are defined as follows:

To develop performance benchmarks for REIT markets in Asia. In particular, the market

indices for listed REITs in Malaysia, South Korea and Taiwan will be developed. In

addition, a pan-Asian composite REIT index as well as two sub-indices for the developed

and emerging Asian REITs will be constructed and analysed.

To analyse the risk and return characteristics of Asian REITs.

To compare the risk-adjusted performance of Asian REITs to other major asset classes of

bonds, shares and property companies.

To assess the diversification potential of Asian REITs in a mixed-asset portfolio in both a

local and international investment perspective.

To ascertain the optimal weight of Asian REITs in mixed-asset and inter-property

portfolios.

9

To investigate the effects of the GFC on Asian REITs at both the first and second

moment.

To explore the return and volatility spillover effects among Asian REIT markets.

1.6 Research Methodology

Given the research framework of the above objectives, the overall research process is divided

into three major studies. The first study investigates the performance and significance of Asian

REITs in a domestic mixed-asset portfolio. In this study, the risk-adjusted performance for local

REITs, bonds, shares and property companies for the seven Asian REIT markets of Japan,

Singapore, Hong Kong Malaysia, Thailand, Taiwan and South Korea are assessed. Both the

Sharpe ratio and reward-to-risk ratio are calculated to further reflect the risk-adjusted

performance of all asset classes. In addition, the correlation between REITs and the other

investment assets were computed to ascertain the possible diversification benefits that can be

gained from including REITs in a domestically mixed-asset portfolio. Efficient frontiers and

asset allocation diagrams were developed to investigate the importance of REITs in an efficient

portfolio. To analyse the impact of the global financial crisis (GFC) on Asian REIT performance,

the timeframe was divided into three sub-periods: pre-GFC, GFC and post-GFC.

The second study examines the significance of Asian REITs in regional and international

investment portfolios. This study extends the analysis to a regional and global investment

perspective. The first part of this study investigates an inter-Asia REIT investment strategy. An

efficient portfolio comprising seven Asian REIT markets is calculated. The role of each REIT

market in the efficient portfolio is analysed and reported. The second part of the study assesses

the potential of a pan-Asian REIT strategy for international property investors. To serve this

purpose, a pan-Asia REIT index is constructed in order to determine whether a pan-Asia REIT

strategy will bring diversification benefits to international property investors. In addition, three

sub-indexes; a pan-Asia excluding Japan REIT index, an emerging Asia REIT index and a

developed Asia REIT index were created to analyse and compare various investment strategies.

The third study explores the dynamics of returns and volatility in Asian REIT markets. The study

employs exponential generalized autoregressive conditional heteroskedasticity (EGARCH)

10

models to examine the dynamic transmission of REIT returns and volatility between the seven

Asian REIT markets. In addition, tests for autocorrelation, heteroskedasticity, volatility

clustering effects are also carried out. Lastly, the effects of the GFC on Asian REIT returns and

volatility will also be assessed using dummy variable regression analysis.

1.7 Significance of the Study

This study provides an empirical analysis on the performance of Asian REITs and their potential

role in mixed-asset portfolios. The analysis covers all seven REIT markets in Asia; i.e. Japan,

Singapore, Hong Kong, Malaysia, Thailand, Taiwan and South Korea. The last decade has

witnessed the significant development of REITs in Asia, underpinned by favourable economic

conditions and improved property market transparency. Asian REITs have also become a major

asset class and important indirect investment vehicle for both local and international investors to

get exposure to the growing commercial property markets in Asia. Given the significance and

growing importance of Asian REITs within a global investment context, studies on Asian

property securities and especially Asian REITs are extremely limited.

This study is significant for a number of reasons. First of all, this study provides a quantitative

analysis on one of the most important property investment vehicles, which is REITs. Over half a

century, REITs have been providing a convenient mean for investors to easily access the

commercial property market without the intensive capital requirement and the hassle of direct

property management. In addition, publicly-traded REITs offer investors more liquidity than

direct exposure to property, as they are listed on the stock exchange and trade more frequently.

The rapid emergence of a global REIT market with a sizable market value of US$700 billion

across 22 countries would offer investors many opportunities for growth and diversification

(Macquarie 2011).

Secondly, this thesis is among a limited but growing number of studies focusing on the Asian

property markets. The strong economic performance coupled with rapid urbanisation and

improved market transparency has created a strong platform for the development of the Asian

property markets. Since direct property investment remains difficult to access in Asia, REITs

provide an efficient way to participate in the growth of the Asian property markets. The Asian

11

REIT markets account for 12.5% of the global REIT market and are set to grow as the major

economies of China and India are expected to launch their REIT markets in the future.

The third significance of this study is that it provides a critical insight to the investment

dynamics of Asian REITs. Given the growing interest of institutional investors in Asian REITs

as an asset class in recent years, it is important to have a comprehensive understanding of the

investment characterises of Asian REITs and their potential role in a mixed-asset portfolio. The

analysis is fundamental to any investment decision making process. In this study, the strategic

implications for both local and international investors will be discussed and presented.

Last but not least, this study applies EGARCH models in modelling temporal linkages between

the property markets in Asia. The use of ARCH-type modelling has been popular in other

research disciplines such as finance and economics, but is still limited in property research. In

addition, most of the previous studies emphasise on the first-moment (return) rather than the

second-moment (volatility), even though volatility also contains important market information.

This study therefore would make important contributions to the current body of literature in

property.

1.8 Thesis Outline

Figure 1.3 provides the overall structure of this research. The thesis is organised into eight

chapters including the introduction (Chapter 1). The contextual framework, covering the

background of REITs, research gaps, statements of problems, research question, objectives,

methodology and the significance of the study has been set out in the first chapter. The

remaining chapters of the thesis is summarised as follows:

Chapter 2 provides a comprehensive review of the Asian REIT markets. It presents an overview

of the global REIT markets, the significance of Asian property markets and the development of

REITs in Asia. In particular, the structure and profile of the current and prospective REIT

regimes in Asia will be discussed and reported. The significance of Asian REITs within a global

context will also be highlighted.

12

Next, Chapter 3 critically reviews the body of literature related to the role and performance of

property in an investment portfolio. It will primarily focus on the importance of direct and

indirect property in investment portfolios and the benefits of international property investment,

with special attention given to the Asian property markets. In addition, a critical review of

previous studies on Asian property companies and Asian REITs will also be presented.

Chapter 4 describes the data and research methodologies employed in this study. It presents the

data sources and description of variables. The research methodologies will be proposed and

elaborated. The performance measures and the regression models applied in this study will be

explained in detail.

After this, Chapter 5 reports the empirical findings on the investment performance of Asian

REITs in a local context. The risk and return characteristics of individual REIT markets will be

evaluated. The significance of Asian REITs in a domestic mixed-asset portfolio for individual

markets will be assessed. The results from the sub-period analysis will also be reported. In

addition, the findings on the impact of the GFC on Asian REIT markets are also examined and

presented.

Chapter 6 discusses the significance and performance of Asian REITs in regional and

international investment portfolios. The possibility and effectiveness of a pan-Asian REIT

strategy will be determined. The risk-adjusted performance of the Asian REIT markets is

assessed. An inter-Asia REIT portfolio is developed and analysed. In addition, a pan-Asia REIT

index and four sub-indexes are developed to assess the risk-adjusted performance and

diversification benefits of a pan-Asia REIT strategy in a regional and global investment context.

Sub-period analysis is also performed to assess the impact of the GFC on the pan-Asia REITs.

Then, Chapter 7 presents results from the application of EGARCH models to measure the

transmission of returns and volatility between Asian REIT markets. It provides empirical

evidence on the nature of information transmission and market linkages between Asian REIT

markets at both the first and second moment. The effects of the GFC on Asian REIT volatility

are also studied.

13

Finally, Chapter 8 finalises the thesis with the conclusions. It summarises the main findings on

the investment performance and temporal linkages of the Asian REIT markets. It will further

review the theoretical contributions and strategic implications of the key findings. This chapter

ends with a discussion on the limitations of the thesis and recommendations for further research.

14

Figure 1.3: Overall Structure of the Thesis

Conclusions

Analysis and Findings

Data and Methodology

Oveview of the REIT Markets

Review of the Literature

Bacground and Context

Research Objectives Chapter 1: Introduction

Chapter 2: An Overview of the Asian REIT Markets

Chapter 4: Data and Methodology

Chapter 5: The Performance and Significance of Asian

REITs in Domestic Mixed-asset Portfolios

Chapter 6: The Significance of Asian REITs in Regional and International Investment

Portfolios

Chapter 8: Conclusion

Chapter 7: The Dynamics of Returns and Volatility in

Asian REIT Markets

Chapter 3: Literature Review

15

2 CHAPTER 2

AN OVERVIEW OF THE ASIAN REIT MARKETS

2.1 Overview of REIT Markets Globally

Real estate investment trusts (REITs) were introduced in the 1960s by the US Congress in order

to make investment in large-scale, income-producing properties more accessible to all investors.

Prior to the introduction of REITs, access to the property market was provided through direct

investments or private partnerships, which were dominated by institutional investors and wealthy

individuals. With the creation of REITs, investors could purchase interests in a portfolio of

properties on the capital market in the same way as they could purchase shares in a company. As

a result, REITs allowed small or retail investors to participate in the ownership of property which

would not otherwise be easily accessible.

US REITs are established as corporations, trusts or associations for the purpose of owning and

managing commercial property. US REITs have the incentive of being pass-through vehicles,

meaning that they can be exempted from corporate income tax. However, in order to qualify for

the special tax treatment, a US REIT must distribute at least 90% of its taxable income in the

form of dividends. In addition, it shall invest at least 75% of total assets in property and derive

more than 75% of gross income from property related sources, including rents from real property

and interest on mortgages financing real property (US SEC 2012).

Figure 2.1 shows the evolution of US REIT industry from 1962 to 2011. Over half a century, the

US REIT industry has changed significantly. By the end of 2006, the US REIT market

capitalisation has risen to its peak at US$438 billion. However, the global financial crisis (GFC)

sent market value tumbling by more than half to US$192 billion by the end of 2008. The period

from 2008 to 2011 saw a remarkable post-GFC recovery of the REIT market, increasing by

130% in value from its lowest point. As of September 2011, there were 180 publicly traded US

REITs with a combined equity value of approximately US$385.2 billion, representing 88.5% of

the total US listed property market (Macquarie 2011).

16

Figure 2.1: Evolution of US REIT Market: 1962 – 2011

Source: Author’s compilation from NAREIT (2012)

The success of REITs in the US has set the stage for the adoption of REIT and REIT-like

structures worldwide. Outside the US, The Netherlands (1969) and Australia (1971) were the

first countries to establish their own REIT regimes. In 2011, there were 510 listed REITs across

22 markets globally (Macquarie 2011). Table 2.1 depicts the top 10 REIT markets in the world

by market capitalisation as of September 2011. The leading REIT markets were the US (#1 and

55.0% market share), Australia (#2 and 9.9%), France (#3 and 8.1%), Japan (#4 and 5.8%) and

UK (#5 and 5.1%). Apart from Japan, Asia also had two other markets among the top 10 largest

REIT markets in the world. These were Singapore (#7 and 3.8%) and Hong Kong (#8 and 1.9%).

The global REIT market represented 44.5% of the overall listed property securities market as of

September 2011 (Macquarie 2011) .

0

50

100

150

200

250

1962 1969 1976 1983 1990 1997 2004 2011

0

50

100

150

200

250

300

350

400

450

500

US$

Bill

ion

Nu

mb

er

No of US-REITs Market Capitalisation

17

Table 2.1: Top 10 REIT Markets: September 2011

REIT Market Market Cap (US$ bln) % of Local Listed RE % of Global REIT Mkt Global Rank

US 385.2 88.5% 55.0% #1

Australia 69.2 90.3% 9.9% #2

France 57.0 94.0% 8.1% #3

Japan 40.4 32.4% 5.8% #4

UK 35.9 62.6% 5.1% #5

Canada 35.4 70.6% 5.1% #6

Singapore 26.7 27.1% 3.8% #7

Hong Kong 13.6 8.6% 1.9% #8

Netherlands 9.5 80.0% 1.3% #9

Belgium 7.1 83.2% 1.0% #10

Source: Author’s compilation from Macquarie (2011)

Table 2.2 exhibit the global REIT markets as of September 2011. With a total market

capitalisation of approximately US$700 billion, global REITs accounted for 44.5% of the total

listed property securities markets. Even though the US was still the dominant REIT market

globally, the last decade has seen dramatic expansion of REITs outside the US, in particular

Europe and Asia. This expansion has been manifested by the increasing investment appetite for

global property and the introduction of REIT legislation in many European and Asian countries.

Table 2.2: Global REIT Markets: September 2011

REIT Markets Market Cap (US$ bln) % of Local Listed RE % of Global REIT Mkt

Americas 421.4 79.9% 60.2%

Europe 114.3 51.8% 16.3%

Asia 87.2 13.2% 12.5%

Oceania 72.3 90.5% 10.3%

Middle East & Africa 4.9 5.6% 0.7%

Global REITs 700.1 44.5% 100.0%

Source: Author’s compilation from Macquarie (2011)

2.2 Significance of Asian Property Markets

Asia has been emerging as an economic powerhouse of the global economy in the last decade.

Table 2.3 presents an overview of the major economic indicators of Asia. The main Asian

economies as a whole represented almost 40% of the world’s total gross domestic product (GDP)

in nominal terms as of 2011. The region is home to some of the largest economies in the world in

terms of nominal GDP. China (#2), Japan (#3), India (#11), South Korea (#15) and Indonesia

18

(#16) are the largest economies in Asia and also among the top 20 economies globally. Together

with Japan (GDP per Capita = US$45,920), four other Asian economies; Singapore

(US$49,271), Hong Kong (US$34,049), South Korea (US$22,778) and Taiwan (US$20,101),

a.k.a the Asian Tigers, have graduated to the list of advanced and high-income economies (IMF

2012).

Despite the negative impact of the GFC and the ongoing European debt crisis, Asia remained in

good shape with solid growth and low unemployment in recent years. GDP growth rates in the

region were significant higher than the rest of the world. Over the 2007-2012 period, the average

annual growth rate in the region was 4.9% as compared to the global average of 3.3%. According

to IMF’s forecasts in 2012, Asia will continue to offer strong opportunities for growth in the

coming years. Average rate of growth over the next five years is forecast to be 5% p.a., well

above the global average of 4.2%. Strong economic fundamentals have enabled the region to

enjoy low unemployment rates, averaging 4.4% over the 2007-2012 period (IMF 2012).

Rapid economic growth in Asia has resulted in structural demographic changes in the region.

Asia is now home to more than half of the world’s population. China (#1), India (#2) and

Indonesia (#4) remain among the five most-populated nations in the world. In 2010, only about

40% of Asia’s population lived in cities, compared with 70% in Europe, North America and

Latin America. As the region is transitioning from an agrarian-based economy to an urban

economy based on industry and services, the coming decades are expected to see rapid

urbanisation across Asia. By 2050, it is predicted that Asia cities will be home to more than 60%

of the world urban dwellers (UN 2010). Accelerated growth in the industry and services sector

and speedy urbanisation are the major driving forces of the property sector in Asia.

19

Table 2.3: Economic and Demographic Indicators of Major Asian Economies

GDP GDP Growth Unem. Rate Inflation Population

Country Nominal as % of Rank per Capita 2007-11 2012-16 2007-11 2007-11 Total as % of Rank

(US$ bln) Global Global (US$) (%) (%) (%) (%) (mln) Global Global

China 7,298 10.5% #2 5,414 10.5% 8.6% 4.1% 3.7% 1,320 19.2% #1

Japan 5,869 8.4% #3 45,920 -0.1% 1.5% 4.5% -0.2% 127 1.8% #10

India 1,676 2.4% #11 1,389 8.1% 7.4% NA 9.2% 1,202 17.5% #2

South Korea 1,116 1.6% #15 22,778 3.5% 3.9% 3.4% 3.4% 50 0.7% #26

Indonesia 846 1.2% #16 3,509 5.9% 6.7% 7.8% 6.3% 246 3.6% #4

Taiwan 467 0.7% #26 20,101 3.9% 4.6% 4.7% 1.4% 23 0.3% #49

Thailand 346 0.5% #31 5,394 2.6% 5.4% 1.2% 2.8% 68 1.0% #20

Malaysia 279 0.4% #36 9,700 4.4% 4.8% 3.4% 2.6% 29 0.4% #44

Singapore 260 0.4% #38 49,271 5.8% 3.7% 2.3% 3.5% 5 0.1% #112

Hong Kong 243 0.3% #40 34,049 3.6% 4.0% 4.1% 2.9% 7 0.1% #95

Philippines 213 0.3% #47 2,223 4.7% 4.8% 7.3% 4.8% 102 1.5% #12

Pakistan 211 0.3% #48 1,201 3.7% 3.5% 5.5% 12.0% 175 2.6% #6

Vietnam 123 0.2% #58 1,374 6.6% 6.7% 4.5% 13.2% 89 1.3% #13

Asia 26,647 38.3%

4.9% 5.0% 4.4% 5.0% 3,442 50.1%

Global 69,660 100.0% 3.3% 4.2% 4.2% 6,865 100.0%

Source: Author’s compilation from IMF (2012)

20

The last few years have also seen Asian countries excelling in their competitiveness based on the

Global Competiveness Index (GCI) published by the World Economic Forum (WEF). The GCI

measures how productively a country uses available resources for economic development. Table

2.4 reports the GCI from 2010 to 2012. Asia has seven countries and territories in the top 30

most competitive markets (WEF 2012). Singapore was the most competitive market in Asia and

second globally, after Switzerland. Japan (#9) and the other economic powers in Asia; Hong

Kong (#11), Taiwan (13#), Malaysia (#21), South Korea (#24) and China (26#) have all made it

to the top 30 list. Among the developing economies, Malaysia was the leader, ranking at 21st

position in 2012. Significant improvements in market competitiveness were also seen for the

Philippines (up 10 places to #75) and Pakistan (gaining 5 ranks to #118).

Table 2.4: Global Competitiveness Index: 2010 – 2012

GCI 2011-2012 GCI 2010-2011 Change

Market Rank Score Rank Score Rank Score

Singapore #2 5.63 #3 5.56 +1 +0.07

Japan #9 5.40 #6 5.37 -3 +0.03

Hong Kong #11 5.36 #11 5.30 0 +0.06

Taiwan #13 5.26 #13 5.21 0 +0.05

Malaysia #21 5.08 #26 4.88 +5 +0.20

South Korea #24 5.02 #22 4.93 -2 +0.09

China #26 4.90 #27 4.48 +1 +0.42

Thailand #39 4.52 #38 4.51 -1 +0.01

India #56 4.30 #51 4.33 -5 -0.03

Vietnam #65 4.24 #59 4.27 -6 -0.03

Philippines #75 4.08 #85 3.96 +10 +0.12

Pakistan #118 3.58 #123 3.48 +5 +0.10

Source: Author’s compilation from WEF (2011, 2012)

Table 2.5 shows the 2010-2011 corruption perception index results for Asia (TI 2012).

Corruption levels in the region varied significantly between countries. This was a result of the

region’s diverse market systems and institutional structures. Singapore was considered to be the

least corrupt market in Asia and ranked 5th

globally. Hong Kong (#12) and Japan (#14) joined

Singapore in the top 20 least corrupt nations in the world. The two economic powerhouses of

Asia China and India ranked 75th

and 95th

respectively. Progressive reforms in the public sectors

have also seen countries such as Vietnam (#112), the Philippines (#129) and Pakistan (#134)

improving their scores and gaining 4, 5 and 9 places respectively during 2010-2011.

21

Table 2.5: Corruption Perception Index: 2010 – 2011

2011 2010 Change

Market Rank Score Rank Score Rank Score

Singapore #5 9.2 #1 9.3 -4 -0.10

Hong Kong #12 8.4 #13 8.4 +1 0.00

Japan #14 8.0 #17 7.8 +3 +0.20

Taiwan #32 6.1 #33 5.8 +1 +0.30

South Korea #43 5.4 #39 5.4 -4 0.00

Malaysia #60 4.3 #56 4.4 -4 -0.10

China #75 3.6 #78 3.5 +3 +0.10

Thailand #80 3.4 #78 3.5 -2 -0.10

India #95 3.1 #87 3.3 -8 -0.20

Vietnam #112 2.9 #116 2.7 +4 +0.20

Philippines #129 2.6 #134 2.4 +5 +0.20

Pakistan #134 2.5 #143 2.3 +9 +0.20

Source: TI (2012)

Market transparency is one of the most important factors for property investors. Jones Lang

LaSalle (JLL) tracks the international property market transparency in their Global Transparency

Index (GTI) reports. According to the 2012 figures, the US (#1), UK (#2) and Australia (#3)

were the most transparent property markets in the world. Asia also has some of the most

transparent property markets (Table 2.6). Hong Kong (#11) and Singapore (#13) were the most

transparent markets in Asia, ranking relatively high globally, at par with most Western European

countries. Malaysia (#23) and Japan (#25) were classified as transparent markets. Since 2010,

China and the emerging South East Asia (the Philippines, Indonesia and Vietnam) have seen the

largest improvements in global rankings. Notably, China’s tier one cities have moved from a

“low transparency” to “semi-transparent” level in 2012, rising from the 45th

to the 32nd

place.

Additionally, strong progress has also been seen for Indonesia (rising 19 places), the Philippines

(climbing 13 ranks) and Vietnam (up 8 ranks) (JLL 2012).

22

Table 2.6: Property Transparency Index: 2012

Transparency

2012 2010 Change

Level Market Rank Score Rank Score Rank Score

Transparent

Hong Kong #11 1.76 #18 1.76 +7 0.00

Singapore #13 1.85 #16 1.73 +3 +0.12

Malaysia #23 2.32 #25 2.30 +2 +0.02

Japan #25 2.39 #26 2.30 +1 +0.09

Semi

Taiwan #29 2.60 #33 2.71 +4 -0.11

China Tier 1 Cities #32 2.83 #45 3.14 +13 -0.31

Philippines #35 2.86 #48 3.15 +13 -0.29

Indonesia #38 2.92 #57 3.46 +19 -0.54

Thailand #39 2.94 #39 3.02 0 -0.08

South Korea #41 2.96 #42 3.11 +1 -0.15


India Tier 2 Cities #49 3.08 #49 3.17 0 -0.09

India Tier 1 Cities #48 3.07 #41 3.11 -7 -0.04

India Tier 3 Cities #50 3.15 #55 3.39 +5 -0.24


Low Vietnam #68 3.76 #76 4.15 +8 -0.39

Source: JLL (2012)

According to Pramerica (2012), the distribution of investable property is forecast to shift toward

the Asia Pacific region, as it is expected to grow much faster than other parts of the world

(Figure 2.2). As of 2011, the region contained a total of US$7 trillion of institutional-grade

commercial property as compared to US$9 trillion in Europe and US$8 trillion in the US.

However, the property market in Asia is projected to grow rapidly to US$19 trillion by 2021 and

US$45 trillion by 2031, overtaking Europe and North America to become the largest property

market in the world. By 2021, it is expected to make up almost 40% of the global property

market, increasing from the current level of 27% in 2011. By 2031, half the world’s investable

property is expected to be located in the Asia Pacific region (Pramerica 2012).

23

Figure 2.2: Investable Property by Region: 2011 – 2031

Source: Pramerica (2012)

The rapid growth of the property sector in Asia is attributed to the increased relative wealth and

swift changes in demographic structure of the region. Currently, many Asian countries still have

small property markets relative to their GDP (Figure 2.3). As of 2011, the average size of Asian

property markets was only 34% of their GDP as compared to 45% in Europe and North America.

98 7

2 1

1312

19

41

2018

45

7

2

0

5

10

15

20

25

30

35

40

45

50

Europe North America Asia Pacific Latin America GCC

US$

Tri

llio

n

2011 2021 2031

27%39%

49%

35%

27%

22%

28% 24%20%

7% 8% 7%3% 2% 2%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2011 2021 2031

Asia Pacific Europe North America Latin America GCC

24

This is going to change as the population is becoming more urbanised and their wealth is

catching up with the developed nations. As seen in Figure 2.3, countries such as China, India,

Indonesia and Thailand still possess vast potential for property growth.

Figure 2.3: Asia Property Markets Relative to GDP: 2011

Source: Author, data from Pramerica (2012)

Asia’s strong appetite for property was further shown by the fact that the region is home to some

of the largest listed property securities markets in the world (Table 2.7). Four of the top 5 and 9

of the top 20 listed property markets are in Asia. In 2011, the Hong Kong securitised property

market reached US$356 billion in value, accounting for almost one fifth of the global market.

This made Hong Kong the largest listed property market in Asia and second largest globally only

to the US (#1). Japan (US$135 billion), Singapore (US$129 billion) and China (US$100 billion)

took up the 3rd

, 4th

and 5th

places globally respectively. All together, Asian listed property

markets contained US$837 billion market capitalisation, accounting for about 44% of the global

listed property volume.

Japan 45% S. Korea

45%

China 26% India

18% Hong Kong 90% Taiwan

42% Thailand 26%

Philippines 20%

Vietnam 17%

Singapore 90%

Malaysia 32%

Indonesia 23%

GDP Property

25

Table 2.7: Global Listed Property Securities Markets: 2011

Market Market Cap (US$ bln) % of Global Rank Global

Hong Kong 356.2 18.5% #2

Japan 135.0 7.0% #3

Singapore 128.7 6.7% #4

China 100.2 5.2% #5

India 21.5 1.1% #12

Philippines 21.1 1.1% #13

Taiwan 20.7 1.1% #14

Malaysia 20.0 1.0% #15

Indonesia 15.5 0.8% #20

Thailand 15.1 0.8% #21

Vietnam 2.4 0.1% #41

South Korea 0.5 <0.0% #47

Sri Lanka 0.5 <0.0% #47

Asia Total 837.0 43.5%

North America 564.1 29.3%

Europe 278.6 14.5% Others 246.1 12.8%

Global Total 1,925.8 100.0%

Source: Macquarie (2011)

Overall, this section has highlighted the economic importance of Asia within a global context, as

well as the significance of the commercial and securitised property markets in the region. The

following section will provide an overview of the development of REIT regimes in Asia and

their regulatory structures.

2.3 Development of REIT Markets in Asia

2.3.1 Overview

The REIT structures were introduced to Asia during the early 2000s, with Japan being the first

Asian country to introduce REIT legislations. Since then, REITs have been implemented across

six other markets in Asia; Singapore, South Korea, Thailand, Taiwan, Malaysia and Hong Kong

(Figure 2.4). In principle, the REIT structures in Asia were modelled on those in the US and

Australia. However, due to the different regulatory and market environments, Asian REIT

markets have their own distinctive and unique features among themselves. The following section

26

will provide comprehensive reviews of the seven REIT regimes in Asia, focusing on the

backgrounds, regulatory structures, special requirements and tax treatments. It will further

discuss the latest development of the prospective REIT markets in India, Pakistan, the

Philippines and China.

Figure 2.4: Evolution of Asian REIT Markets

Source: Author

2.3.2 Japan

2.3.2.1 Background

In November 2000, Japan was the first country in Asia to establish a REIT market. The Japanese

real estate investment trusts (J-REITs) were introduced by the Act on Investment Trusts and

Investment Corporations also known as the Investment Trust Law (ITL). According to the ITL,

there are two different types of J-REITs; (1) investment trusts and (2) investment corporations.

So far, all listed J-REITs have been established as investment corporations (CBRE 2011).

In September 2001, Japan Real Estate Investment (sponsored by Mitsui Fudosan) and Nippon

Building Fund (sponsored by Mitsubishi Estate) were the first two J-REITs listed on the Tokyo

Stock Exchange (TSE). The Japanese REIT market capitalisation has expanded rapidly to reach

US$50 billion in May 2007. During the GFC, market value fell rapidly by more than half to its

2001:

Japan

2002:

Singapore

South Korea

2003:

Thailand

2005:

Taiwan

Malaysia

Hong Kong

Future:

India

Pakistan

Philippines

China

27

trough at US$21 billion in October 2008. In the period after the GFC, the J-REIT market has

recovered strongly, supported by stimulus programmes (e.g. Real Estate Market Stabilisation

Fund, Asset Purchase Program) by the Japanese government and the Bank of Japan (BOJ)

(ARES 2012).

On March 11 2011, a magnitude 9.0 earthquake struck the east coast of Japan triggering a

devastating tsunami and causing a number of subsequent nuclear accidents. In the immediate

aftermath of the earthquake, fear of a nuclear crisis has shocked the Japanese share market. The

Nikkei 225 share index plunged by 10.22% within a week after the earthquake. The Japanese

REIT market was also affected but to a lesser extent than shares, dropping by only 4.38% for the

week. This is partly because the direct impact of the crisis on J-REITs was only limited, as they

had only 3% of their assets in the affected areas (JLL 2011). The estimated repair cost of

damages on J-REITs’ properties totalled ¥2,980 million (US$38 million), averaging 0.11% of

their total asset size (ARES 2011). By the end of April 2011, the J-REIT index has actually

increased by 1.72% over the previous month.

With a strong boost by the BOJ’s asset purchasing scheme, J-REITs have been actively

expanding their property portfolios. Acquisitions by J-REIT jumped by 67% yoy to US$9 billion

in 2011, accounting for around two-thirds of the total regional volume during the period (CBRE

2012). Despite a temporary setback from the natural disaster, the post-GFC period from October

2008 to April 2012 witnessed a remarkable recovery of the J-REIT market, as total market

capitalisation almost doubled from US$21 billion to US$41 billion (Figure 2.5).

With 34 listed J-REITs and a total market capitalisation of US$41 billion (Table 2.8) as of

September 2011, Japan has become one of the major global REIT markets. It is home to the

fourth largest REIT market in the world after the US (#1), Australia (#2) and France (#3). In

Asia, Japanese REITs have been playing a leading role, making up 41.1% of the total Asian

REIT market capitalisation as at June 2012. The country has 4 REITs in the top 10 largest REITs

in Asia, 9 in the top 20 and 24 in the top 50. Within a domestic context, J-REITs accounted for

32.4% of the local listed property securities market. The industry is strongly supported by the

28

government and the private sector, in particular by professional associations such as the

Association for Real Estate Securitisation (ARES).

Figure 2.5: Evolution of J-REIT Industry: October 2001 – April 2012


Table 2.8: J-REITs by Property Sector: June 2012

Sector Market Cap. (US$ mln) No. of REITs

Diversified 15,946 16

Office 14,599 7

Residential 4,926 7

Retail 4,434 2

Industrial 2,104 2

Lodging 277 1

Total 42,287 35


Figure 2.6 depicts the J-REIT property segmentation. J-REITs invested across the diversified

(36%), office (35%), residential (12%), retail (11%), industrial (5%) and lodging (1%) property

sectors (Figure 2.6). The majority of J-REITs hold assets in diversified property sectors (16),

followed by the office (7) and residential (7) sector. Table 2.9 presents the profile of J-REITs as

of June 2012. In total, J-REITs own 1,981 properties, averaging 58 properties per J-REIT.

GFC

0

10

20

30

40

50

60

No of J-REITs Market Cap (US$ billion)

29

Figure 2.6: J-REIT Segmentation

By Market Capitalisation By Number of REITs

Diversified36%

Office35%

Residential12%

Retail11%

Industrial5%

Lodging1%

Diversified46%

Office20%

Residential20%

Retail5%

Industrial6%

Lodging3%

30

Table 2.9: Profile of Listed J-REITs: June 2012

Listed Property No. of Mkt Cap. Asia

REIT Name Date Sector Properties (US$ mln) Rank

NIPPON BUILDING FD. Sep-01 Office 67 5,346 #2

JAPAN REIT Sep-01 Office 57 4,676 #3

JAPAN RETAIL FD. INV. Mar-02 Retail 70 2,859 #7

UNITED URBAN INV. Dec-03 Diversified 90 2,150 #10

JAPAN PRIME REALTY INV. Jun-02 Diversified 58 1,981 #12

ADVANCE RESIDENCE INV. Mar-10 Residential 190 1,966 #13

MORI TRT. SOGO REIT Feb-04 Diversified 13 1,941 #14

NOMURA RE OFFICE FD. Dec-03 Office 51 1,643 #17

FRONTIER REIT Aug-04 Retail 26 1,576 #18

JAPAN LOGISTICS FD. May-05 Industrial 29 1,255 #22

NIPPON ACC. FD. Aug-06 Residential 88 1,224 #23

ORIX JREIT Jun-02 Diversified 66 1,188 #24

ACTIVIA PROPERTIES INC. Jun-12 Diversified 18 1,172 #25

DAIWA HOUSE RES. Mar-06 Diversified 124 1,061 #28

DAIWA OFFICE INV. Oct-05 Office 36 945 #30

JAPAN EXCELLENT Jun-06 Office 23 929 #33

MORI HILLS REIT INV. Nov-06 Diversified 10 884 #35

KENEDIX REALTY INV. Jul-05 Diversified 83 879 #36

IND. & INF. FD. Oct-07 Industrial 22 849 #37

FUKUOKA REIT Jun-05 Diversified 19 842 #39

TOP REIT Mar-06 Diversified 18 828 #42

TOKYU REIT Sep-03 Diversified 26 781 #43

NOMURA RE RES. Feb-07 Residential 152 725 #45

PREMIER INV. Sep-02 Diversified 54 671 #47

GLOBAL ONE REIT Sep-03 Office 8 594 #51

JAPAN RTL. HOUSING INV. Jun-06 Residential 177 591 #52

MID REIT Aug-06 Office 12 466 #59

SEKISUI HOUSE SI INV. Jul-05 Diversified 64 459 #61

ICHIGO REIT Oct-05 Diversified 67 373 #68

HANKYU REIT Oct-05 Diversified 15 370 #69

HEIWA RE REIT Mar-05 Diversified 88 367 #70

JAPAN HOTEL AND RESORT Feb-06 Lodging 28 277 #79

KENEDIX RES. INV. Apr-12 Residential 20 168 #85

STARTS PROCEED INV. Nov-05 Residential 76 143 #89

INVINCIBLE INV. May-04 Residential 54 108 #96

TOTAL

1,999 42,287

Sources: Author’s compilation from DataStream, ARES (2012) and various companies’ websites

31

2.3.2.2 Regulatory Structure

2.3.2.2.1 Legal Form

According to the Act on Investment Trusts and Investment Corporations (ITL), a J-REIT must be

formed as a trust type or a corporation type. They are further sub-divided to open-ended or

closed-ended funds depending on the possibility of cancellation or redemption of shares. Up to

2011, all of the current J-REITs were corporate type since the legal process for the trust type was

more burdensome and expensive (EPRA 2011). The corporate governance rules for the corporate

type were also considered to be more favourable for investors.

2.3.2.2.2 Minimum Initial Capital

J-REITs are subject to the minimum investment capital of ¥100 million (US$1.26 million).

2.3.2.2.3 Sponsorship

The REIT market in Japan has a unique sponsorship system. Under this system, a J-REIT is

required to have a sponsor or a parent company. The sponsor will establish a REIT manager and

acquire a “Building Lots and Building Transactions Agent Licence” and a “Discretionary

Transaction Agent Licence” from the Ministry of Land, Infrastructure, Transport and Tourism

(MLIT). After these licences are obtained, the manager may apply for registration as an

investment manager with the Financial Services Agency (FSA). Typically, J-REITs are

sponsored by major property companies in Japan. For example, the first two J-REITs Japan Real

Estate Investment and Nippon Building Fund, are sponsored by two major property players in

Japan; the Mitsui Fudosan and the Mitsubishi Estate respectively. The REIT sponsor has an

important role because it wholly owns the REIT management company and often have a large

holding in the REIT (CFA Institute 2011).

2.3.2.2.4 Management Structure

Figure 2.7 outlines the typical corporation structure for a listed J-REIT. J-REITs have an external

management structure, whereby the trust or investment corporation has no employees and must

use a registered asset management company. The investment manager is required to have a

minimum paid-in-capital or net assets of ¥50 million (US$0.63 million) and sufficient

experienced personnel.

32

Figure 2.7: Investment Corporation Structure for Listed J-REITs

Source: CFA Institute (2011)

2.3.2.3 Requirements

2.3.2.3.1 Ownership

There are no requirements for the shareholders of J-REITs.

2.3.2.3.2 Assets

A J-REIT is required to invest more than 50% of its total assets in qualified assets as dictated

under the ITL. Qualified assets include; (1) securities, (2) property, (3) leasehold rights in

property, (4) surface rights, (5) monetary debts, (6) promissory notes, (7) trust beneficiary rights,

J - REIT Investment

Corporations

Board of Directors

Unit Holders

Properties

Asset Manager Custodian

Distributable

Income

Asset Custody

Fee

Distributions

Management Fee

Outsourcing of Management

Services

Outsourcing of Custodian Function

Flow of funds Services Ownership

33

(8) interest in a tokumei kumiai (anonymous partnership) and (9) trust beneficiary rights to

monies. In addition, a J-REIT cannot own more than 50% of the shares of another corporation.

2.3.2.3.3 Geographic Restrictions

There are no restrictions under the ITL. However, no J-REIT has made any overseas acquisitions

(CBRE 2011).

2.3.2.3.4 Property Development

Ownership of development of properties is restricted. At least 50% of total assets must be

income producing and are not to be sold within one year.

2.3.2.3.5 Gearing

There are no restrictions on the borrowing or gearing level of J-REITs. In order to receive a tax

deduction on dividends, J-REITs cannot borrow from lenders other than institutional investors.

However, most J-REITs have a limitation on the gearing ratio of around 55% to 60% in their

articles of incorporation. As of May 2012, the average debt to total asset ratio of J-REITs was

approximately 45% (APREA 2012).

2.3.2.3.6 Distribution

J-REITs are required to distribute at least 90% of their net profits after tax to their investors

within the same fiscal period. Capital gain is a part of the taxable income. However, there are no

minimum distribution requirements for capital gain.

2.3.2.4 Tax Treatment

2.3.2.4.1 Income

In principle, J-REITs are subject to regular corporate tax. However, under the Special Taxation

Measures Act (STMA), J-REITs are allowed to deduct distributed dividends from its taxable

income, upon meeting the STMA’s requirements. The retained capital after the deduction of

distributed dividends will be subject to the normal corporate taxes. As from April 1, 2012 Japan

national corporate tax is 25.5%, which has been recently reduced from the previous 30% rate.

34

However, there is also a 10% surtax which applies until March 31, 2015 resulting in an effective

tax rate of 28.05% for the first three years and 25.5% thereafter (Deloitte 2012).

2.3.2.4.2 Capital Gains

Capital gains are taxable as ordinary income; capital losses are generally deductable.

2.3.2.4.3 Withholding Tax

A 20% withholding tax is normally levied on dividend distributions to non-residents unless the

rate is reduced under a tax treaty. From January 1, 2013, a 2.1% surtax will increase the rate to

20.42%.

2.3.2.4.4 Other Taxes

Other taxes include real property acquisition tax (fudosan shutoku zei), registration and licence

tax (touroku menkyo zei), fixed asset tax (koteishisan zei), city planning tax (toshi keikaku zei)

and stamp duty (inshi zei).

35

2.3.3 Singapore

2.3.3.1 Background

The Singaporean REIT (S-REIT) sector started in 1999, after the first set of regulatory guidelines

for REITs was issued by the Monetary Authority of Singapore (MAS). S-REITs are subject to

the Securities and Future Act (SFA) and the Code on Collective Investment Schemes (CCIS). S-

REITs may or may not be listed on the Singapore Exchange (SGX).

In July 2002, CapitaMall Trust made a successful debut public offering and became the first S-

REIT to be listed on the SGX. The following five years saw a wave of new S-REITs, as both

developers and investors took advantage of the low interest rate environment. During the period

from 2002 to 2007, there were 19 additional S-REITs launched. The industry however suffered

in 2008 as the capital market was impacted by the GFC. The market endured a dry spell over the

next two years with only one successful listing, the IndiaBulls Properties Trust on June 2008.

Since 2011, the market however has become more active, thanks to ample liquidity in the

financial system following the implementation of measures to stimulate economic growth. The

period over 2010-2011 saw the establishments of 5 new S-REITs, among which were the two

heavyweights backed by Temasek; the Mapletree Industrial Trust (raising approx. US$1 billion)

and the Mapletree Commercial Trust (raising US$ 1.6 billion).

As of June 2012, there were 26 publicly traded S-REITs with an overall market capitalisation of

approximate US$31 billion (Table 2.10). Singapore is the second largest REIT market in Asia

after Japan, accounting for 30.9% of the total size of the regional REIT market. As of June 2012,

9 of the 26 S-REITs were retail REITs, representing almost half (US$14.97 billion) of the S-

REIT market capitalisation (Figure 2.8). The industrial sector accounts for 8 S-REITs (US$9.08

billion), corresponding to 29% of the total market capitalisation.

Table 2.11 depicts the property profile and market capitalisation of public S-REITs as of June

2012. S-REITs had a significant profile in terms of portfolio size and market capitalisation. The

number of properties ranged from 2 (Indiabulls Properties) to 133 (Saizen REIT), averaging 28

properties per S-REIT. The largest S-REIT, in terms of market capitalisation, was CapitaMall

Trust (US$4.6 billion), while the smallest one was Saizen REIT (US$144 million). S-REITs also

36

have a significant position regionally, accounting for 4 of the top 10 Asian REITs, 8 of the top 20

and 16 of the top 50 in terms of market value. It has been considered the best REIT market in

Asia consistently for overall potential, property market growth, REIT opportunity and regulatory

support (Trust 2011).

Table 2.10: S-REITs by Property Sector: June 2012


Retail 14,972 9

Industrial 9,083 8

Office 2,393 2

Lodging 1,348 1

Specialty 1,273 2

Residential 1,066 2

Diversified 738 2

Total 30,873 26


Figure 2.8: S-REIT Segmentation


Retail49%

Industrial29%

Office8%

Lodging4%

Specialty4%

Residential4%

Diversified2%

Retail34%

Industrial31%

Office7%

Lodging4% Specialty

8%

Residential8% Diversified

8%

37

Table 2.11: Profile of Listed S-REITs: June 2012



CAPITAMALL TRT. Jul-02 Retail 16 4,634 #4

ASCENDAS REIT Nov-02 Industrial 95 3,476 #5

CAPITACOMMERCIAL TRT. May-04 Retail 9 2,593 #8

SUNTEC REIT Dec-04 Retail 7 2,218 #9

K-REIT ASIA Apr-06 Office 8 1,930 #15

MAPLETREE LOGISTIC TRT. Jul-05 Industrial 96 1,811 #16

MAPLETREE IND. TRT. Oct-10 Industrial 70 1,444 #19

CDL HOSPITALITY TRT.S Jul-06 Lodging 12 1,348 #20

MAPLETREE COMERCIAL TRT. Apr-11 Retail 2 1,327 #21

FRASERS CENTREPOINT TRT. Jul-06 Retail 4 1,043 #29

FORTUNE REIT Aug-03 Retail 14 944 #31

STARHILL GLOBAL REIT Sep-05 Retail 6 936 #32

ASCOTT RESIDENCE TRT. Mar-06 Residential 48 922 #34

PARKWAY LIFE REIT Aug-07 Specialty 36 849 #38

CAPITARETAIL CHINA TRT. Dec-06 Retail 9 667 #48

LIPPO MALLS INDO. RETAIL Nov-07 Retail 15 610 #50

CACHE LOGISTICS TRT. Apr-10 Industrial 10 558 #53

CAMBRIDGE IND. TRT. Jul-06 Industrial 47 490 #55

SABANA SHARIAH IND. TRT. Nov-10 Industrial 20 467 #58

FRASERS COMMERCIAL TRT. Mar-06 Office 9 463 #60

ASCENDAS INDIA TRT. Aug-07 Industrial 5 450 #62

FIRST REIT TRT. Dec-06 Specialty 10 424 #63

PERENNIAL CHINA RETAIL TRT. Jun-11 Diversified 8 394 #64

AIMS AMP CAPITAL IND. Apr-07 Industrial 26 388 #66

INDIABULLS PROPERTIES Jun-08 Diversified 2 344 #73

SAIZEN REIT Nov-07 Residential 133 144 #88

TOTAL

717 30,873

Source: Author’s compilation from DataStream and various companies’ websites



An S-REIT must be formed as a trust.


There is no minimum capital required for S-REITs.

38

2.3.3.2.3 Management

Figure 2.9 illustrates a typical structure for a listed S-REIT. An S-REIT may be managed

externally or internally, but in practice all are externally managed. Under the SFA, the

management company is required to be a public company incorporated in Singapore. In addition,

the manager should have a physical office in Singapore and have minimal capital funds of S$1

million (US$0.78 million).

Figure 2.9: Typical Structure for a Listed S-REIT



2.3.3.3.1 Ownership

Singaporean dollar-denominated S-REITs listed on the SGX must have at least 25% of the total

shares held by more than 500 public unit holders. This ownership restriction maybe exempted if

the capital market capitalisation is greater than the minimum listing requirement. There are no

restrictions on foreign unit holders.

S - REIT

Unit Holders

Trustee Manager

Properties

Distributable Income

Trustee Fee Management Fee

Management Services

Holds Property on Trusts for Unit Holders

Distributions Identifies Investment Opportunities

Fee

Adviser

Flow of funds Services Ownership

39

According to research by CBRE (2012), S-REITs have become popular attractions for

international investors, especially investors from the US. Figure 2.10 shows the unitholder

ownership of S-REITs as of Q4 2011. North American investors were the largest unitholders of

S-REITs, accounting for almost half (46%) of the total units. Only a third (33%) of the total

shares was owned by domestic investors. Regional and European investors held 11% and 10% of

the market value respectively (CBRE 2012).

Figure 2.10: S-REIT Unitholder Ownership

Source: CBRE (2012)

2.3.3.3.2 Assets

An S-REIT can only invest in; (1) property, (2) property-related assets, (3) listed or unlisted debt

securities and listed shares of non-property corporations, (4) Singapore statutory board and (5)

cash or cash-equivalent items. Furthermore, at least 75% of its deposited property should be

invested in income-producing property.

North America46%

Singapore33%

Asia-Pacific11%

Europe10%

40

2.3.3.3.3 Income

An S-REIT should not derive more than 10% of its revenue from sources other than rental

payments from the tenants of the property held by the REIT or interest, dividends and other

similar payments from special purpose vehicles (SPVs) and other permissible investments.


In the spirit of promoting Singapore as a regional REIT hub, the Singaporean authority imposes

no geographical restrictions for S-REITs. As a result, S-REITs have been very active in cross-

border acquisitions. The S-REIT overseas portfolio by country is displayed in Figure 2.11. As of

February 2012, around 38% of S-REIT assets are located outside Singapore, one of the highest

proportions among global REITs. While most of S-REIT assets are located in Asia,

approximately 17% of their portfolios are outside the region, in markets including the UK,

Western Europe and Australia (CBRE 2012).

Figure 2.11: S-REIT Overseas Portfolio by Country

Source: CBRE (2012)


No property development activities should be undertaken, unless the S-REIT intends to hold the

developed property upon completion. The total contract value of property development activities

China23%

India19%

Japan18%

SE Asia17%

Outside Asia17%

Hong Kong5%

South Korea1%

41

and investments in uncompleted property developments should not exceed 10% of the S-REIT’s

deposited property.

2.3.3.3.6 Gearing

The aggregate leverage of an S-REIT should not exceed 35% of the fund’s deposited property.

The aggregate leverage of an S-REIT may be raised to a maximum of 60%, if a credit rating of

the REIT from Fitch Inc, Moody’s or Standard and Poor’s is obtained and disclosed to the

public.


In order to enjoy tax transparency treatment for property investments in Singapore, an S-REIT

must distribute at least 90% of its taxable income to unit holders. However, there is no such

requirement for overseas properties, as tax exemption is not applicable.


2.3.3.4.1 Income

Rental and property related income from Singaporean properties is exempted from income tax as

long as S-REITs comply with the 90% or more payout ratio. Foreign dividends, interest and trust

distributions received from foreign properties may also be exempted from Singaporean income

tax if certain conditions are met.


There is no capital gain tax in Singapore. However, capital gains from trading nature will be tax

at the prevailing corporate tax rate, currently at 17%.


There is no withholding tax on distribution to individuals. However, distributions to a foreign

non-individual unit holder are subject to withholding tax at 10% until March 31, 2015.

42


The acquisition of property is liable to stamp duty of approximately 3%. However, newly listed

S-REITs (within 6 months) can apply for stamp duty remission until March 31, 2015.

43

2.3.4 Hong Kong

2.3.4.1 Background

In July 2003, The Hong Kong Securities and Future Commission (SFC) issued the Code on Real

Estate Investment Trusts (Code on REITs), setting up the foundation for the establishment of

Hong Kong Real Estate Investment Trusts (HK-REITs). The code was ratified in 2005 and 2007

for the formal regulatory operation of HK-REITs. In 2010, further amendments were made to

encourage the growth and attractiveness of HK-REITs to investors. The changes included lifting

the maximum borrowing ratios from 35% to 45% and allowing overseas investments and new

property sectors.

In November 2005, the Link REIT became the first Hong Kong REIT (HK-REIT) to be

established through the world’s largest initial offering at that time. At June 2012, there were 7

HK-REITs listed on the Hong Kong Stock Exchange (HKEx) with a total market capitalisation

of US$15 billion (Table 2.12). This saw Hong Kong being the third largest REIT market in Asia,

after Japan and Singapore. HK-REITs engaged across the retail (2 REITs), diversified (2), office

(2) and lodging (1) industry. The retail REITs dominated the market, accounting for almost 60%

of the total market capitalisation.

The profiles of HK-REITs at June 2012 are provided in Table 2.13. HK-REITs are often

established and supported by major property players in the country. Link REIT, the largest REIT

in Hong Kong and Asia, was established by the Hong Kong Housing Authority. At June 2012, it

had a market capitalisation of US$8.6 billion and the portfolio consisted of 180 properties with

an internal floor area (IFA) of approximately 11 million square feet (1 million sqm) of retail

space. Several other HK-REITs are also founded by leading property companies, such as

Prosperity REIT (Cheung Kong), Champion REIT (Great Eagle) and Sunlight REIT (Henderson

Land).

With many positive regulatory changes in recent years, Hong Kong is considered the second

most attractive REIT market in Asia, after Singapore, in terms of overall potential, property

market growth and regulatory support (Trust 2011). Hong Kong is also an important gateway for

investors to gain access to the mainland China’s property market. Many HK-REITs hold

44

substantial assets in mainland China. Hui Xian REIT (backed by Cheung Kong) was the first

RMB-denominated REIT in the world and established in April 2011. As of June 2012, it was the

second largest REIT in Hong Kong and ranked 6th

in Asia. All of its investment properties are

located in mainland China; i.e. the Beijing Oriental Plaza and the Sheraton Shenyang Lido Hotel.

Other China-focus HK-REITs include the Regal REIT and the GZI/Yuexiu REIT.

Table 2.12: HK-REITs by Property Sector: June 2012


Retail 9,106 2

Diversified 3,386 2

Office 2,295 2

Lodging 714 1

Total 15,501 7

Figure 2.12: HK-REIT Segmentation


Retail59%

Diversified22%

Office15%

Lodging4%

Retail28%

Diversified29%

Office29%

Lodging14%

45



A typical structure of a HK-REIT is shown in Figure 2.13. A HK-REIT is a collective investment

scheme established as a unit trust. It may hold property directly or indirectly through SPVs.

However, such SPVs must be legally and beneficially owned by the HK-REIT and it must have

the majority ownership of the SPVs. Generally, no more than two layers of SPVs are allowed

unless specifically approved by the SFC.

As all HK-REITs must be listed, they are governed by the Code on REITs as well as the Listing

Rules issued by the HKEx. Unit holders of a HK-REIT cannot require redemption of their units,

but they can only exit via a sale of their units on the HKEx.


There is no minimum initial capital for HK-REITs.


The appointed management company can be internal or external. The management company

may choose to itself perform all the functions required of it under the Code on REITs or delegate

or contract out to one or more outside entities one or more of these functions. Furthermore, a

trustee that is functionally independent of the management company of the REIT must be

appointed, but may be part of the same corporate group if certain requirements are met.

Table 2.13: Profile of Listed HK-REITs: June 2012



LINK REIT Nov-05 Retail 180 8,631 #1

HUI XIAN REIT Apr-11 Diversified 2 2,880 #6

CHAMPION REIT May-06 Office 2 2,000 #11

REGAL REIT Mar-07 Lodging 6 714 #46

SUNLIGHT REIT Dec-06 Diversified 20 506 #54

YUEXIU REIT Dec-05 Retail 5 475 #57

PROSPERITY REIT Dec-05 Office 7 295 #78

TOTAL

222 15,501


46

Figure 2.13: Typical Structure of a Listed HK-REIT



2.3.4.3.1 Ownership

There are no limitations to the holding of units in HK-REITs. Additionally, there are no

restrictions on foreign unit holders.

2.3.4.3.2 Assets

Under the REIT Code, a HK-REIT must primarily invest in income-generating properties. Non-

income generating assets should not exceed 10% of the total Net Asset Value (NAV) of the HK-

REIT at the time of acquisition. The Code does not specify eligible properties but hotels,

recreational parks, serviced apartments, office buildings shopping malls and car parks have been

authorised by the SFC. Industrial buildings, however, have not yet been authorised by the SFC.

Where the REIT invests in hotel, recreation parks or serviced apartments, such investments shall

be held by SPVs. A REIT must hold its real estate for a period of at least two years, unless

consent is obtained from its unit holders by way of a special resolution at a general meeting.

Special Purpose Vehicles

(SPVs)

The Properties

Unit Holders Trustee

Management

Company

The H-REIT

Management

Services

Fees

Owner of the SPVs,

holding the Properties

on Trust for Unit

Holders

Holds Units Fees

47

2.3.4.3.3 Geographic Restriction

HK-REITs may invest in properties outside of Hong Kong. Many HK-REITs have made cross-

border investments in China (i.e. Regal REIT, GZI/Yuexiu REIT and Hui Xian REIT).


A HK-REIT is prohibited from investing in vacant land or engaging in or participating in

property development activities. Refurbishing, retro-fitting and renovations are accepted as they

are not regarded as property development activities.

2.3.4.3.5 Gearing

The aggregate borrowings must not exceed 45% of the total gross asset value of the HK-REIT at

the time when the borrowing is incurred.


A HK-REIT is required to distribute not less than 90% of its audited annual net income after tax

in the form of dividends to its unit holders each year. Where the HK-REIT holds property via

SPVs, each SPV must distribute to the HK-REIT all its income as permitted by the laws and

regulations of the relevant jurisdictions.


2.3.4.4.1 Income

HK-REITs are not tax transparent. HK-REITs are exempt from Hong Kong profits tax under the

Inland Revenue Ordinance of Hong Kong. However, where the HK-REIT holds real estate in

Hong Kong directly and derives rental income thereon, such rental income will be subject to

Hong Kong property tax at the prevailing rate of 15%. Where the REIT holds real estate in Hong

Kong indirectly via SPVs, such SPVs will be subject to profits tax at the prevailing rate of 16.5%

in respect of the profits derived from the real estate. The lack of tax transparency by HK-REITs

has been an impediment to the development and institutional investor acceptance of HK-REITs

(Newell et al. 2010).

48


Hong Kong does not impose tax on capital gains.


There is no withholding tax on interest, dividends or distributions from a HK-REIT. Hong Kong

has a territorial tax system and does not tax foreign-sourced income.


Property tax is charged annually on the owner of land and buildings situated in Hong Kong at the

standard rate on the NAV of such land and building. However, the SPVs can apply to the Inland

Revenue for exemption from property tax. Acquisition of new property in Hong Kong or

disposal of any property will attract stamp duty of up to 3.75% of the value of the property.

49

2.3.5 Malaysia

2.3.5.1 Background

The history of the Malaysian REIT (M-REIT) industry dated back to the establishment of listed

property trusts (LPTs) in 1989. However, it was not until 2005 that the Securities Commission of

Malaysia (SC) issued the Guidelines on Real Estate Investment Trusts (ordinary M-REITs) and

Guidelines for Islamic Real Estate Investment Trusts (Islamic M-REITs), providing the principal

legislations governing the official establishment of REITs in Malaysia. Both types of M-REITs

share similar regulatory structures and requirements, except that the Islamic M-REITs have to

comply with the Sharia requirements. Malaysia was the first country to introduce guidelines for

Islamic REITs at that time (Newell & Osmadi 2009).

As of June 2012, there were 15 M-REITs listed on the Bursa Malaysia (MYX), with a total

market capitalisation of US$5 billion (Table 2.14). This saw Malaysia being the fourth largest

REIT market in Asia. However, the REIT industry in Malaysia was still relatively small as

compared to those in Japan, Singapore and Hong Kong. Three of the largest M-REITs; Pavilion

REIT (US$1.14 billion), Sunway REIT (US$1.09 billion) and CapitaMalls Malaysia Trust

(US$839 million), were among the top 50 REITs in Asia in terms of market value. The three

Islamic M-REITs were Axis (US$390 million), Al-Hadharah Boustead (US$351 million) and

Al’Aqar Healthcare REIT (US$274). The number of properties in M-REIT portfolios ranged

from 2 to 29, averaging 10 properties per M-REIT.

50

Table 2.14: Profile of Listed M-REITs: June 2012



PAVILION REIT Dec-12 Retail 2 1,135 #26

SUNWAY REIT Jul-10 Diversified 8 1,087 #27

CAPITAMALLS MALAYSIA TRT. Jul-10 Retail 4 839 #40

AXIS REIT Aug-05 Office 29 390 #65

STARHILL REIT Dec-05 Retail 10 383 #67

AL-HADHARAH BOUSTEAD REIT Feb-07 Specialty 12 351 #71

AL’AQAR HEALTHCARE REIT Aug-06 Specialty 19 274 #80

UOA REIT Dec-05 Office 6 181 #83

AMANAHRAYA REIT Feb-07 Diversified 15 161 #86

AMFIRST REIT Dec-06 Office 8 156 #87

HEKTAR REIT Dec-06 Retail 5 141 #90

QUILL CAPITAL TRT. Jan-07 Office 10 139 #91

TOWER REIT Apr-06 Office 3 123 #94

ATRIUM REIT TRT. Apr-07 Industrial 5 44 #109

AMANAH HARTA TANAH PNB Sep-01 Diversified 11 33 #114

TOTAL

147 5,436


Based on the types of properties, M-REITs were categorised as retail (4 REITs), diversified (3),

office (5), specialty (2) and industrial (1) sector (Table 2.15). Retail M-REITs comprised the

largest segmentation, accounting for 46% of M-REIT market capitalisation (Figure 2.14). This

group was followed by the diversified and office REITs, representing 25% and 18% of the total

market value respectively.

Table 2.15: M-REITs by Property Sector: June 2012


Retail 2,497 4

Diversified 1,282 3

Office 988 5

Specialty 625 2

Industrial 44 1

Total 5,436 15


51

Figure 2.14: M-REIT Segmentation




REITs in Malaysia have to be registered as unit trusts. Malaysian trustees must be approved by

the SC (Figure 2.15). M-REITs can be listed or unlisted. A listed M-REIT is subject to the listing

requirements of the MYX. Unlisted M-REITs are opened-ended, meaning there is a mandatory

requirement for the management company, upon the proper request of an investor, to repurchase

units issued to unit holders.


The minimum initial capital of an M-REIT is MYR100 million (US$32 million). For subsequent

offerings, the minimum size of the funds is MYR25 million (US$7.8 million).


An M-REIT must be externally managed and administered by a management company approved

by the SC. Foreigners can hold up to 70% of the equity of the management company. At least

30% of the equity of the management company must be held by bumiputra (ethnic Malay)

shareholders. The properties held by the trust must be managed by a qualified property manager.

Retail46%

Diversified24%

Office18%

Specialty11%

Industrial1%

Retail27%

Diversified20%Office

33%

Specialty13%

Industrial7%

52

An Islamic M-REIT is subject to similar requirements as above, plus a Shariah committee or a

Sharia to ensure that any asset acquired is Shariah-compliant.

Figure 2.15: Typical Structure for an M-REIT

REIT

Authorised

Investments

Property

Management

Companies

REIT Managers Trustee

Unit Holders

Flow of funds

Services

Ownership

Dividends

and other

distribution

Trustee Fee

Management Fee

Acts on behalf

of Unit Holders

Management

Service

Property

Management

Service

Property Management

Fee


2.3.5.3.1 Ownership

There are no restrictions for both local and foreign unit holders of M-REITs. However,

foreigners cannot hold more than 70% in the M-REIT’s management company.

2.3.5.3.2 Assets

A M-REIT can only invest on the following authorised assets; (1) property, (2) single-purpose

asset owning companies, which principal assets comprise of property, (3) property-related assets,

(4) non-property-related assets and (5) cash, deposits and money market instruments. At least

53

60% of M-REITs’ total asset value must be invested in property and or single-purpose

companies, holding property at all times. Investment in non-property-related assets and cash

cannot be in excess of 25% of an M-REIT’s total asset value. These requirements apply to both

listed and unlisted M-REITs.

Additional limitations apply to Islamic M-REITs. The manager must ensure that the rental

income from non-Shariah-compliant activities does not exceed 20% of the total revenue of the

Islamic M-REIT. Non-compliant activities are financial services which are based on riba

(interest), gambling, gaming, non-permissible entertainment activities, conventional insurance,

the manufacture or sale of tobacco-based products, stock brokerage, share trading in Shariah

non-compatible securities, as well as hotels and resorts.

2.3.5.3.3 Gearing

Generally, the total borrowings of an M-REIT may not exceed 50% of the total asset value. An

M-REIT may only borrow from institutions that are licensed under the Banking and Financial

Institution Act 1989 and the Islamic Banking Act 1983.


M-REITs may acquire property located outside Malaysia. The management company must

ensure that relevant rules and regulations are complied with and authorisations from the SC have

been obtained prior to acquisition.


M-REITs are prohibited from property development. They can only acquire property that is

under construction or uncompleted properties of up to 10% of its total asset value, provided that

certain criteria listed in the SC Guidelines are met.


Legally, M-REITs are not required to have any minimum distribution of income, but they may

only receive tax exemption provided at least 90% of their total income for the year is distributed

to their unit holders. There are no requirements for capital gains to be distributed every year.

54


2.3.5.4.1 Income

M-REITs will be exempted from income tax, provided that at least 90% of the total income for

the year is distributed. In addition, start-up expenses incurred during the establishment, such as

consultancy, legal and valuation fees can be deducted.


There is no capital gains tax in Malaysia.


Withholding tax of 10% for non-corporate unit holders is a final tax. This rate has been extended

from January 1, 2012 to December 31, 2016. If this rate is not extended further, REIT dividends

received after December 3, 2012 will be subject to a final withholding tax of 15% from January

1, 2016 onwards.


M-REITs may attract real property gains tax (RPGT) at an effective rate of 5% on disposal of

properties or shares in a property company held within five years of the date of acquisition.

However, if the sale of real properties or shares occurs after five years, then it will be exempted

from RPGT.

55

2.3.6 Thailand

2.3.6.1 Background

The concept of real estate investment trusts in Thailand (Thai-REIT) was introduced in 2002, as

a recovery vehicle for distressed properties after the 1997 Asian financial crisis. The Thai

version of a REIT is known in the country as Listed Property Funds for Public Offering (PFPO).

This is a type of mutual fund and is listed on the Stock Exchange of Thailand. The law regulating

the PFPO is the Securities and Exchange Act B.E. 2535, enacted in 1992. Although PFPO was

initially established as a resemblance of the REIT structures in the US and Australia, the PFPO

code is considered to be more rigid than other REIT regulations overseas.

In October 2010, the Securities and Exchange Commission of Thailand (TSEC) announced that

it has approved the new REIT regulatory framework, under the Trust for Transactions in Capital

Market Act B.E. 2550 (2007). The new legislations aimed to follow international REIT practices

and provide greater flexibility for REITs as compared to the current PFPO code. There will be

less restriction on investment activities. More importantly, the maximum total borrowing of

Thai-REITs will be raised to 50% of NAV, as opposed to 10% in the case of PFPO. This will

give Thai-REITs more flexibility to acquire additional properties and to optimise their capital

structure. These changes had been well received by property trust managers as well as

institutional investors. The current PFPO will be able to convert into the REIT. However, in

March 2011, the government announced that the enactment of the new Thai-REIT framework

will be delayed until the Revenue Department clarifies the corporate income tax status of

institutional investors.

The first Thai-REIT, UOB Apartment Property I Leasehold, was listed on the SET in 2003. As

of June 2012, there were 37 listed Thai-REITs investing across retail (5), industrial (8), office

(6), lodging (6), residential (7) and diversified (5) property sectors (Table 2.16). Even though

small in number (5), the retail REITs accounted for the almost half (44%) of the total market

capitalisation of US$4.1 billion, followed by industrial (23%) and office (12) REITs (Figure

2.16).

56

The profile of listed Thai-REITs as of June 2012 can be seen in Table 2.17. Thai-REITs are

characterised by small size, with an average market capitalisation of US$111 million and lack of

property exposure. The average number of properties in a Thai-REIT portfolio is only 8, with the

majority of them (11) holding single property portfolios. The largest Thai-REIT in terms of the

number of properties is the Ticon Property Fund, with 174 assets in the industrial property

sector.

Table 2.16: Thai-REITs by Property Sector: June 2012


Retail 1,817 5

Industrial 937 8

Office 509 6

Lodging 395 6

Residential 275 7

Diversified 177 5

Total 4,111 37


Figure 2.16: Thai-REIT Segmentation


Retail44%

Industrial23%

Office12%

Lodging10%

Residential7%

Diversified4%

Retail13%

Industrial22%

Office16%

Lodging16%

Residential19%

Diversified14%

57

Table 2.17: Profile of Listed Thai-REITs: June 2012



CPN RETAIL GROWTH LH Aug-05 Retail 3 729 #44

TESCO LOTUS RETAIL GROWTH Mar-12 Retail 17 666 #49

SAMUI AIRPORT PROP. FD. Nov-06 Industrial 1 333 #75

TICON PROP. FD. May-05 Industrial 174 328 #76

QUALITY HOUSES LH Dec-06 Office 3 234 #81

FUTURE PARK LH Dec-06 Retail 1 212 #82

MAJOR CINEPLEX Jul-07 Retail 2 124 #92

DUSIT THANI FH & LH Jan-11 Lodging 3 123 #93

PRIME OFFICE LH Apr-11 Office 2 118 #95

WHA PREMIUM FACTORY Dec-10 Industrial 3 108 #97

LAND AND HOUSE PROP. FD. Apr-12 Residential 3 104 #98

TALAAD THAI LH Nov-10 Diversified NA 91 #99

CENTARA HOTEL & RESORT Oct-08 Lodging 1 87 #100

THAI RETAIL INV. FD. Jul-11 Retail 1 87 #101

TPARK LOGISTICS PROP. FD. Dec-09 Industrial 15 86 #102

MILLIONAIRE PROP. FD. Mar-05 Office 1 73 #103

MFC AMAZING A-LA ANDAMAN Nov-11 Lodging 2 67 #104

THAI COMMERCIAL INV. Jun-10 Office 4 63 #105

QUALITY HOSPITALITY LH Apr-08 Lodging 1 55 #106

GOLD PROP. FD. May-07 Residential 1 50 #108

LUXURY RE Jun-08 Lodging 1 39 #111

MFC-NICHADA THANI Aug-05 Residential 2 35 #112

MULTI-NATIONAL RESIDENCE Jun-08 Residential 3 33 #115

NICHADA THANI PROP. FD. 2 Apr-09 Residential 2 30 #116

BANGKOK COMMERCIAL PROP. Nov-03 Diversified 2 29 #117

TU DOME RESIDENCE COMPLEX Dec-06 Diversified 3 26 #118

MERCURE SAMUI PROP. Aug-10 Lodging 1 24 #119

MFC-STRATEGIC STORAGE FD. Aug-09 Industrial 4 22 #120

THAI IND. FD. 1 Jun-05 Industrial 8 21 #121

SALA @ SATHORN PROP. FD. Aug-09 Office 1 21 #122

TRINITY PROP. FD. Mar-11 Diversified 1 21 #123

SUB SRI THAI May-11 Industrial NA 21 #124

101 MONTRI STORAGE PROP. Aug-09 Industrial 1 18 #125

PROP. PERFECT FD. Mar-08 Residential 2 14 #129

URBANA PROP. FD. Oct-07 Diversified 1 10 #130

UOB APARTMENT PROP. Oct-03 Residential 1 8 #132

JC LH PROP. FD. Jan-07 Office 1 NA NA

TOTAL

272 4,111

Source: Author’s compilation from DataStream and various companies’ websites

58



The current Thai-REITs are established as mutual funds. However, under the new regime, Thai-

REITs will be closed-ended trusts with no maturity. The Thai-REITs must be listed on the SET

within 60 days of completion or the date they are offered to investors.


The management company, appointed by the trustee, must be licensed by the Ministry of Finance

and regulated by the Office of Securities and Exchange Commission of Thailand. The

management company and the trustee cannot be related.


A capital of minimum 500 million (US$3.2 million) is required.


2.3.6.3.1 Ownership

The minimum number of unit holders is 250 unit holders for an IPO and 35 unit holders after

listing in the Stock Exchange of Thailand (SET). Currently, persons in the same group or

affiliated companies cannot hold more than 1/3 of the total trust certificates sold. The

requirement on holding restriction will be lifted to 50% of trusts certificates sold. A foreign limit

of 49% of the total share certificates sold is imposed on freehold REITs.

2.3.6.3.2 Assets

Thai-REITs must invest no less than 75% of the NAV in property. Furthermore, they may only

invest in completed property or property that is at least 80% complete. The property must be

hold at least one year from the day of investment, unless waived by the TSEC due to unavoidable

circumstances. In the case of real property development, the REIT shall hold the real property at

least one year after completion of the construction. A Thai-REIT may invest in subsidiaries.

2.3.6.3.3 Income

At least 75% of the total income of the fund must be generated from rental income.

59

2.3.6.3.4 Gearing

Under the current structure, borrowings are limited to a maximum of 10% of NAV. However, the

proposed regulations will allow the trusts to borrow up to 35% of NAV for investment or

improvement of the invested property.


At least 90% of the net profit must be distributed to unit holders within 90 days after the end of

each annual accounting period. In addition, at least 90% capital gains are required to be

distributed.


Thai-REITs can only invest in property located in Thailand.


Development of property is not allowed for current Thai-REITs. However, under the new

structures, Thai-REITs can only invest in real property that is 80%-100% complete and develop

real property representing not more than 10% of NAV. Investment in vacant land can only be

accepted if it is for development property and it can be demonstrated that it supports the income

of Thai-REITs’ property.


2.3.6.4.1 Income

A Thai-REIT is not subject to income tax, but it pays a 12.5% Land and Building Tax on the

annual rental income of immovable properties.


Capital gains are not taxed at the level of the trusts.


No withholding tax is legislatively imposed.

60


The Thai-REIT is to pay a once-off registration fee of 1% on the amount of rental fee of

immovable properties (only if the lease period is more than three years); and 2% (reduced to

0.01% for two years commencing from March 29, 2008 for certain types of immovable

properties) transfer fee of the official appraised price for the income on disposal of immovable

properties. The 2% transfer fee is reduced to 0.01% for the transfer of immovable properties to

the property fund.

61

2.3.7 Taiwan

2.3.7.1 Background

In Taiwan, the Real Estate Securitisation Act (RESA) came into effect in 2003 and was last

amended in 2009. Fubon No. 1 was the first Taiwanese REIT (T-REIT), which was established

on March 2003 and listed on the Taiwanese Stock Exchange (TWSE) on July 2006. To date, a

total of eight T-REITs have been launched and floated on the TWSE. However, two T-REITs

have been delisted since 2011 (i.e. Trident REIT and Kee Tai REIT).

Table 2.18 shows the market break-up of the current T-REITs as of June 2012. There were six

currently listed T-REITs with a total market capitalisation of US$2.4 million, accounting for

2.5% of the Asian REIT market. This saw Taiwan having the second smallest REIT market in

Asia (#6/7 market), above only South Korea. Among the 6 T-REITs, 4 of them invested in the

diversified property sector, representing 73% total market capitalisation, while the other 2

invested in offices, accounting for the remaining 27% of the T-REIT market capitalisation

(Figure 2.17).

As seen from Table 2.19, all of the currently listed T-REITs were launched prior to May 2007;

no new T-REITs have been launched subsequently as of June 2012. T-REITs have relatively

small capitalisation, averaging US$415 million and small property portfolios ranging from 3 to 6

properties.

Table 2.18: T-REITs by Property Sector: June 2012


Diversified 1,830 4

Office 661 2

Total 2,491 6


62

Figure 2.17: T-REIT Segmentation


Table 2.19: Profile of Listed T-REITs: June 2012



CATHAY NO.1 REIT Jul-06 Diversified 3 828 #41

SHIN KONG NO.1 REIT Jul-06 Diversified 6 487 #56

CATHAY NO.2 REIT Oct-06 Office 3 347 #72

FUBON NO.1 REIT Jul-06 Diversified 4 338 #74

FUBON NO.2 REIT Jul-06 Office 3 313 #77

GALLOP NO.1 REIT May-07 Diversified 3 177 #84

TOTAL 22 2,491

Sources: Author’s compilation from DataStream and various companies’ websites



The RESA allows two structures for T-REITs; real estate investment trusts (REITs), following

US model, which require the establishment of a mutual fund; and real estate asset trust (REAT),

following the Japan’s regime of special purpose trusts. Both types of T-REITs are established as

trusts and administered by a trustee. A trustee is an institution that may manage and dispose the

trust property and raise public offerings or private placements the in the shares of T-REITs. In

practice, all trustees have been local banks or branch offices of foreign banks in Taiwan.

Diversified73%

Office27%

Diversified67%

Office33%

63


Under the RESA, the manager is external and must be appointed by the trustee. A manager is not

required to be specially licensed or approved.


The minimum initial capital ranges from NT$300 million (US$10 million) to NT$2 billion

(US$67 million) depending on the scope of business engaged by the trust. The minimum paid-in

capital required for a T-REIT engaging only in REIT or both the REIT and REAT business under

the RESA is NT$1 billion (US$67 million) and the minimum paid-in capital for a trust company

engaging only in REAT business is NT$300 million (US$67 million).


2.3.7.3.1 Ownership

The public T-REITs face a number of ownership restrictions. The trust certificates shall be held

by at least 50 persons for at least 335 days during a fiscal year, except for independent

professional investors. In addition, any five certificate holders shall not own more than 50% of

the total value of the certificates issues. As for the private T-REITs, the total investors shall not

exceed 35 persons in number and the same person cannot hold more than 25% of the total

number of shares issued.

2.3.7.3.2 Assets

According to the RESA, the investments of a T-REIT shall be limited to certain objects; (1)

existing property with stable income or property to be developed, (2) rights associated the

property, (3) property securities, (4) permitted utilisation and (5) other investment objects

approved by the competent authority. At least 75% of the NAV of a T-REIT must be composed

of cash, government bonds, property, property-related rights, beneficiary securities or asset-

backed securities (ABS) issued under the RESA or Financial Asset Securitization Act (FASA).

2.3.7.3.3 Gearing

T-REITs may borrow money and the underlying assets can be used as collateral for that purpose.

However, the purpose of the borrowed money is limited to the needs of real estate operations and

64

the distribution of profits, interests or other proceeds. The Taiwan’s Financial Supervisory

Commission (FSC) has the authority to set an upper limit of the borrowing ratio of a T-REIT,

depending on its credit rating. The normal upper limit is 35% of the net worth of a T-REIT, but it

can be up to 50% for T-REITs with credit rating of “twAA” or above.


At least 90% of the proceeds derived from the T-REIT investment shall be distributed within 6

months after the closing of the fiscal year.


There are no geographic restrictions under the RESA.


Development activities are allowed for urban renewal, infrastructure or public amenities

construction. However, investments should not exceed 15% of the T-REIT’s NAV.


2.3.7.4.1 Income

The Trustee is considered as a pass-through entity in terms of tax. Therefore, the income

generated from the operation of the REIT funds is not subject to corporate income tax at the

trustee level.


Capital gains generated by the operation of the T-REIT funds are not subject to corporate income

tax. However, the Land Value Increment Tax, applicable to the increase in sale value over

purchase value of land, will be paid by the T-REIT upon the disposal of the real estate.


The withholding tax rate applied depends on the category of the income. Generally, the interest

income of a REIT will be subject to a 10% withholding rate. Rental revenues received by the

Trustee will not be subject to withholding tax if the GUIs (Government Uniform Invoice) are

65

issued by the Trustee and the tenants are individuals. Withholding tax withheld may be

recovered by the Trustee from the tax authority.


The Trustee is the taxpayer of the Land Value Tax imposed on the registered owner of the

property.

66

2.3.8 South Korea

2.3.8.1 Background

The ground work for REITs in South Korea (K-REITs) was laid in 2001 by the amendment of

the Real Estate Investment Company Act (REICA). The introduction of REITs was a part of the

government’s reforms to create an effective scheme for troubled property companies to spin off

their real assets. Despite high expectations, the K-REIT market attracted little attention from

both property companies and investors due to a burdensome legal process.

In 2007, the Korean Ministry of Construction and Transportation revised the REICA in an

attempt to further simplify regulations governing the establishment and operation of K-REITs.

Major elements of the revision included the reduction of restrictions on borrowing, initial capital

requirements and lengthy approval processes. Currently, there are 3 REIT regimes in South

Korea; (1) corporate restructuring (CR) REITs, self-managed (SM) REITs and manager-

entrusted (ME) REITs.

The first listed K-REIT was the Kyobo-Meritz 1 CR-REIT, launched in April 2002. However, it

has since been delisted from the Korea Stock Exchange (KRX). As of June 2012, there were 7

active K-REITs in the diversified (4), office (2) and residential (1) property sector (Table 2.20).

Despite having an early start, the K-REIT industry was still lagging behind its regional

counterparts. As of June 2012, it was the smallest REIT market in Asia, with a total market

capitalisation of just US$465 million, accounting for less than 0.5% of the total Asian REIT

market capitalisation. The diversified sector made up the largest segmentation, representing 74%

of K-REIT total market capitalisation (Figure 2.18).

Table 2.20: K-REITs by Property Sector: June 2012


Office 93 2

Diversified 81 4

Residential 10 1

Total 184 7


67

Figure 2.18: K-REIT Segmentation


Table 2.21: Profile of Listed K-REITs: June 2012



KOCREF 15 CR-REIT Jan-10 Office 1 52 #107

KOCREF REIT 8 Jun-06 Office 2 41 #110

TRUS Y 7 REIT Sep-11 Diversified NA 34 #113

K-TOP REIT Jan-12 Diversified NA 18 #126

GOLDEN NARAE RE DEV. May-10 Diversified NA 15 #127

KWANG HEE DEV. REIT Jul-11 Diversified NA 14 #128

E-COREA REIT Mar-11 Residential NA 10 #131

TOTAL

3 184

Sources: Author’s compilation from DataStream and various companies’ websites



A K-REIT must be established as a stock corporation (chusik hoesa) under the Korean

Commercial Code and REICA. CR K-REITs are joint stock companies that have finite lives of

five years and must be dissolved when the period ends, while the ordinary K-REITs, SM and

ME, are paper companies with infinite lives. K-REITs must be established in Korea.

Office51%

Diversified44%

Residential5%

Office29%

Diversified57%

Residential14%

68


The minimum initial capital of ₩500 million (US$0.4 million) is required for obtaining a

business license from the Ministry of Land, Transport and Maritime Affairs (MOLTM). Within 6

months after the license is granted, the equity capital should be increased to ₩7 billion (US$6

million) for the SM K-REITs and ₩5 billion (US$4.3 million) for the CR and SM K-REITs.


The management can be either internal or external.


2.3.8.3.1 Ownership

One shareholder or any related person cannot possess more than 30% of the total stocks issued

by a K-REIT. In addition, at least 30% (20% for the period up to December 31, 2012) of the

shares must be offered to the public within 6 months from official permission (no limit for CR

K-REITs). There are no special restrictions on foreign ownership.

2.3.8.3.2 Assets

K-REITs shall have at least 80% of the total assets in property, property-related securities and

cash and 70% or more of total assets in property (including buildings under construction). In

addition, in case of a CR K-REIT, at least 70% of the total assets must be corporate recovery

related property. Corporate recovery property is property in which a company disposes to repay

its debts under relevant laws. The minimum holding period of property assets for ordinary K-

REITs are three years. For CR K-REITs, there are no restrictions.

2.3.8.3.3 Gearing

Generally, a K-REIT can borrow funds or issue bonds within twice the equity value.

Nevertheless, if there is a special resolution by the general shareholders’ meeting a K-REIT can

borrow funds or issue bonds up to 10 times the NAV.

69


90% of distributable income of a K-REIT must be distributed to shareholders. There is no

difference between a domestic and foreign-sourced profit distribution.


K-REITs are not forbidden to hold property in foreign jurisdictions.


In general, a K-REIT can invest in a property development project to a maximum of 30% of its

total assets. However, K-REITs that are specifically registered as a development specialised

REIT can be exempted from this restriction. A K-REIT may not hold more than 10% of voting

shares in other companies with an exception including a merger and an acquisition of a business.


2.3.8.4.1 Income

A K-REIT can claim tax deduction for its dividend paid, as long as 90% of the distributable

income is distributed as dividends. Undistributed income is subject to corporate income tax at a

rate of 10% for the first taxable income up to ₩200 million (US$180,000) and 22% on the

amount exceeding ₩200 million (US$180,000). Additionally, a 10% local residential surtax will

be levied.


According to the Korean Corporate Income Tax Law, capital gains are treated as ordinary

income and subject to the normal corporate income tax rate. However, there is no tax on capital

gains if the 90% distribution obligation is met. Additionally, capital gain surtax at a rate of 33%

will be applied on the sale of certain tainted assets such as housing or non-business purpose land.


Corporations with permanent establishments in South Korean are not subject to withholding tax

on J-REITs’ dividends. A withholding tax rate of 15.4% will be levied on distributions to

70

resident individuals. For non-residents, the withholding tax rate is 22% which may be reduced

under applicable treaty.


There are no other taxes levied on the corporate income.

2.3.9 Prospective Asian REIT Markets

2.3.9.1 India

In December 2007, the Securities and Exchange Board of India (SEBI) issued the long awaited

draft regulations on Real Estate Investment Trusts for public comments. Based on the public

consultation, the SEBI determined that Indian REITs should be established in the form of mutual

funds, of which the legal framework has existed. In April 2008, the SEBI further amended the

existing SEBI Mutual Fund Regulation to incorporate Real Estate Mutual Funds (REMFs), the

Indian model of REITs. Currently, there is as yet any REMF to be established in India. However,

there is a number of Real Estate Private Equity Funds (REPEFs) providing access to India’s

property market.

2.3.9.2 Pakistan

A REIT framework has been introduced in Pakistan in 2008, following the enactment of the Real

Estate Investment Trust Regulation by the Securities and Exchange Commission of Pakistan

(SECP). REITs in Pakistan are established as closed-ended trusts. The requirements for Pakistan

REITs are very similar to those in other Asian REIT markets. For example, Pakistan REITs are

also required to distribute not less than 90% of their annual income to unit holders. As long as

the 90% minimum distribution obligations are met, Pakistan REITs will be subject to income tax

exemption. In addition, the REIT regulations also require that each REIT must have a minimum

size of PKR2 billion (US$21 million) and the management company must hold at least 20% in

each REIT.

In March 2009, the SECP granted permission to two REIT management companies. Among four

applications received,f only two were approved, whereas the other two were rejected due to their

failure to meet the requirements. The two approved managers were Arif Habib REIT

71

Management Company Ltd. and AKD REIT Management Company Ltd. Both companies have

filed a number of REIT schemes with the SECP for its approval.

2.3.9.3 The Philippines

The Real Estate Investment Trust Act of 2009, also known as the Republic Act 9856, was

enacted in December 2009, providing the legal base for Philippine REITs (P-REITs). The REIT

Act became effective in February, 2010. In April 2010, the Securities and Exchange Commission

of the Philippines (PSEC) released a draft of the Implementing Rules and Regulations in relation

to the REIT Act 2009.

The legal structure of P-REITs follows that of J-REITs, where they will be formed as

corporations with shares of stocks instead of unit trusts. P-REITs are required to have a

minimum paid-up capital of ₱3 million (US$72,000) and at least 1,000 public shareholders. The

Act further advises that at least 75% of a P-REIT’s assets must be in income-generating property.

With respect to leveraging, borrowings shall not exceed 75% of the market value of each P-

REIT. Furthermore, 90% of distributable income must be allocated to shareholders.

In addition to the general requirements, which are in line with international standards, the

Philippine’s regulators has imposed a number of highly stringent rules in terms of ownership and

taxation. In particular, the public ownership of P-REITs is required to be at least 40% during the

first two years from listing and 67% by the third year. The Bureau of Internal Revenue (BIR)

further insists that P-REITs could only receive pass-through status if they maintain public

ownership of at least 67% and distribute at least 90% of their distributable income. Additionally

to the minimum ownership, the BIR also requires a 12% value-added tax on the initial transfer of

property assets in P-REITs.

These strict rules have been publicly criticised by property market participants. A number of

major property developers in the country, including SM Prime, Ayala Land and Megaworld,

have postponed their plans for P-REIT initial offerings. Consequently, the Philippines was

placed second last, above only Thailand, in Asia in terms of overall potential, REIT opportunity

and regulatory support, despite having high property market growth potential (Trust 2011).

72

2.3.9.4 China

As at June 2012, China does not have any legal framework for REITs. The first attempt to

establish a REIT model in China (C-REIT) was in 2004 when the China Banking Regulatory

Commission (CBRC) issued a consultation draft governing companies that engage in REITs

(Clifford Chance 2010). In 2007, two local companies Union Trust & Investment and CITIC

Securities, were given permission by the People’s Bank of China (PBC) to establish pilot

schemes. However, these pilot programmes were later abandoned due to the 2007 GFC and fear

of the Chinese property bubble.

In December 2008, the Chinese government reconsidered the introduction of C-REITs to the

financial market. The PBC and other central government authorities, including the CBRC and

the China Securities Regulatory Commission (CSRC), have been drafting rules and

administrative guidelines for pilot C-REITs. Local governments in Beijing, Tianjin, Shanghai

and Shenzhen have expressed their interest to participate in the pilot programmes formulated by

the PBC. In 2009, it was reported that a number of listed companies in Shanghai, including

Shanghai Zhangjiang Hi-Tech Park Development, Shanghai Jinqiao Export Processing Zone

Development, Shanghai Waigaoqiao Free Trade Zone Development and Shanghai Lujiazui

Finance & Trade Zone Development, have submitted their applications to the State Council for

approval. However, no pilot C-REIT scheme has been endorsed up to date.

2.4 Significance of Asian REITs

Since the establishment of the first REIT in 2001, the Asian REIT market has grown

significantly to reach its peak at U$81 billion in May 2007 (Figure 2.19). As the GFC unfolded,

the total market value plunged by half to its lowest point at US$38 billion in October 2008.

However, the subsequent period after the GFC has witnessed a robust recovery of the market,

supported by strong fiscal stimulus and quantitative easing measures from governments across

Asia. Taking advantage of the low interest rates and abundant capital, Asian REITs have

intensified new public offerings and acquisition activities during 2009 – 2011. Total market

capitalisation has almost trebled from its trough to an all-time high at US$106 billion in April

2012, well above the pre-GFC’s level.

73

Figure 2.19: Evolution of Asian REIT Markets: October 2001 – June 2012

Source: Author’s complitation from DataStream

The strong recovery of Asian REITs after the GFC is further shown in Table 2.22. It can be seen

that Asian REIT markets were stronger performers than the mature REIT markets of the US, UK

and Australia. As of the first quarter of 2012, the Asian REIT markets, except Japan, have

recovered from the previous 5-year’s levels, while the mature markets were still in negative

territory.

GFC

0

20

40

60

80

100

120

US$

bill

ion

Japan Singapore Hong Kong Malaysia Taiwan Thailand South Korea

Table 2.22: Asian REIT Performance: Q1 2012

YTD 1 Year 3 Year 5 Year

Japan 12.2% -0.2% 18.7% -5.5%

Singapore 15.8% 1.3% 37.6% 0.9%

Hong Kong 4.1% 14.5% 31.1% 12.0%

Malaysia 3.4% 20.2% 13.0% 4.1%

Taiwan 12.3% 45.0% 30.9% 13.8%

US 10.7% 12.9% 43.4% -0.3%

UK 14.3% -5.4% 24.1% -18.1%

Australia 8.4% 2.3% 31.4% -9.4%

Source: S&P (2012)

74

As of June 2012, there were 133 REITs across Asia with a combined market value of US$100

billion. Table 2.23 and Figure 2.20 depict the significance of Asian REITs within a regional and

global context. Currently, the Asian REIT market is dominated by the three developed markets

of Japan, Singapore and Hong Kong, making up more than 90% of the total market

capitalisation. Japan is still the largest REIT market in Asia, representing 41% of the regional

market value. However, there has been strong competition coming from Singapore (31%) and

Hong Kong (15.5%), as they position themselves as regional hubs for offshore REIT listings.

The last few years have also seen the emerging markets of Malaysia, Thailand, Taiwan and

South Korea implementing significant regulatory changes in order to catch up with the leading

players.

Within a global context, Asia REITs also have a significant profile, adding up to 12% of the

global REIT market as of June 2011. Japan (#5), Singapore (#7) and Hong Kong (#8) were

among the top 10 REIT markets globally in terms of market size.

Table 2.23: Significance of Asian REITs: June 2012

Country No. of REITs

Mkt Cap. (US$ bln)

% of Asia REITs

Asia REIT Rank

% of Global REIT*

Global REIT Rank*

Japan 35 42.3 41.9% #1 5.8% #5

Singapore 26 30.9 30.6% #2 3.8% #7

Hong Kong 7 15.5 15.4% #3 1.9% #8

Malaysia 15 5.4 5.4% #4 0.5% #12

Thailand 37 4.1 4.1% #5 0.1% #15

Taiwan 6 2.5 2.5% #6 0.3% #20

South Korea 7 0.2 0.2% #7 <0.1% #22

Asia Total 133 100.9 100.0% 12.5%

*As of September 2011. Sources: Author’s compilation from DataStream, Macquarie (2011)

75

Figure 2.20: Asian REIT Market Share

Table 2.24 and Figure 2.21 present the break-down of Asian REIT assets by property types. In

terms of market value, retail property was the largest REIT sector in Asia, followed by the

diversified and office sector. As of June 2012, there were 22 retail REITs with a combined

market value of US$32.8 billion, representing 33% of the total Asian REIT market. Among the

133 listed REITs, 36 REITs held diversified property portfolios with a total equity value of

US$22.5 billion (23%), while 26 REITs invested in office buildings worth US$21.6 billion

(22%). Asian REITs also invest across industrial (US$12.2 billion; 12%), residential (US$6.3

billion; 6.3%), lodging (US$2.7 billion; 2.7%) and specialty (US$1.9 billion; 1.9%) properties.

Table 2.24: Asian REIT Market by Property Sector

Sector No. of REITs Mkt Cap. (US$ mln) % of Total

Retail 22 32.8 32.8%

Diversified 36 22.5 22.6%

Office 26 21.5 21.5%

Industrial 19 12.2 12.2%

Residential 17 6.3 6.3%

Lodging 9 2.7 2.7%

Specialty 4 1.9 1.9%

Total 133 100.0 100.0%

Source: Sources: Author’s compilation from DataStream

Japan41.9%

Singapore30.6%

Hong Kong15.4%

Malaysia5.4%

Thailand4.1%

Taiwan2.5%

South Korea0.2%

76

Figure 2.21: Asian REIT Market Break-down by Property Type

2.5 Summary

This chapter has provided an overall description of the evolution of REITs globally, the

development of REITs in Asia, as well as the significance of Asian REITs within a global

context. Since its inception in the 1960s in the US, the REIT marketplace has grown

dramatically. REITs now play an important role in the commercial property market as well as the

broader economy. The success of REITs in the US has led to their adoption around the world. As

of 2011, there were more than 500 listed REITs across 22 markets, with a combined market

capitalisation of US$815 billion (Macquarie 2011).

In Asia, REITs have been introduced during the early 2000s. At the moment, REIT regulations

have been developed across Japan, Singapore, Hong Kong, Malaysia, Taiwan, Thailand and

South Korea. Table 2.25 provides a legislation summary of the current REIT regimes in Asia.

Other major Asian economies including India, Pakistan, the Philippines and China are also in

progress towards setting up their own REIT markets. The rapid development of the REIT sector

in Asia is underpinned by strong economic growth, increased competitiveness, transparency and

Retail32.5%

Diversified23.2%

Office21.4%

Industrial12.1%

Residential6.2%

Lodging2.7%

Specialty1.9%

77

reduced corruption levels. Recent years have also seen the industry enjoying many positive

regulatory changes and government support.

At the moment, there are 133 Asian REITs with a total market capitalisation of US$100 billion,

representing 12% of the global REIT market. The post-GFC period has seen robust recovery of

the Asian REIT industry, outperforming those in the US and UK. Asian REIT markets provide

institutional investors effective channels to tap into the huge potential of the commercial

property markets in the region. In addition, special products such as Islamic REITs and RMB-

denominated REITs also create new investment opportunities and diversification tools for global

investors. As the prospective REIT markets, in particular those of India and China, are expected

to be launched in the next couple of years, the medium and long-term outlook for Asian REITs

remains very optimistic.

Given the significant development and prospect of Asian REITs markets, it is important to assess

the significance and performance of these REIT markets. Subsequent sections of this PhD thesis

will empirically assess the performance of the Asian REIT markets, as well as considering them

in a pan-Asia REIT context. In addition, the role of Asian REITs in domestic and international

mixed-asset portfolios will be examined. Further regression analysis on the inter-dependency

between the Asian REIT markets will also be conducted and reported.

78

Table 2.25: Asian REIT Legislation Overview

JAPAN SINGAPORE HONG KONG SOUTH KOREA TAIWAN THAILAND MALAYSIA

Structure Trust or corporate

(listed REITs are all

corporations)

Collective

investment

scheme (Unit

trust) or

corporate

Unit trust Corporate-

Restructuring,

Entrusted

Management,

Development-

Specialised, Self-

Managed

Trust (Real estate asset

trust or investment

trust)

Closed-end

mutual fund Unit trust

Management

structure External External Internal/ External Internal/ External Internal/ External External External

% invested in

real estate For listed J-REIT, at

least 75% of assets

must be invested in

real estate

At least 70%

of deposited

property

should be

invested in

real estate or

real estate-

related assets

Only invest in real

estate At least 70% in

real estates or

corporate

restructuring

related properties

Cash, government

bonds, property,

property-related rights,

beneficiary securities or

ABS issued under Real

Estate Securitization

Act/ Financial Asset

Securitization Act

(RESA/ FASA) must

form at least 75% of the

NAV

Must invest

at least 75%

of NAV in

property

At least 50% of a

fund’s total asset

value must be

invested in real

estate and/or

single- purpose

companies at all

times

Geographical

restrictions No restriction under

the Investment Trust

and Investment

Company Act, but

no overseas

acquisitions have

been made as the

requirements on real

estate appraisal of

overseas properties

are ambiguous.

No No No No restriction under the

RESA; subject to

approval

Thailand

only No restriction

basically, subject

to approval from

SC and relevant

authorities

79

Table 2.25: Asian REIT Legislation Overview

Property

developments Restricted – at least

50% of total assets

are income

producing and

unlikely be sold

within one year

Property

developments

and

investments

in

uncompleted

projects

should not

exceed 10%

Prohibited, but H-

REIT may acquire

uncompleted units

comprising less

than10% NAV

Allow Allow for urban

renewal, infrastructure

or public amenities

construction;

investments should not

exceed 15% of asset

value

May acquire

properties

over 80%

completed,

but

prohibited to

invest in

dormant land

(for PFPO)

Prohibited, but

may enter into

conditional

forward purchase

agreement

Leverage No restriction Over 35% of

total assets

permitted

with disclosed

credit rating

(capped at

60%)

Capped at 45% of

gross asset value REITs are

permitted to have

exceptional

borrowing up to

1000% of equity

capital, upon

special approval of

shareholders

Varies based on credit

rating; 50% of total

asset value for REITs

with credit rating of

twAA or above

Not more

than 10% of

NAV

50% of total asset

value (revised

from 35%)

Dividend

payout At least 90% of

distributable income

to qualify for tax

deduction

At least 90%

so as to enjoy

exemption

from paying

corporate tax

At least 90% of

annual net income

after tax

At least 90% At least 90% of

distributable income At least 90%

of net profits Not specified in

the M-REIT

guideline

Source: CBRE (2010)

80

3 CHAPTER 3

LITERATURE REVIEW

3.1 Introduction

This chapter will provide a discussion of the theory and literature related to the performance

and significance of property in an investment portfolio. Although the body of knowledge on

this topic is extensive, this review will primarily focus on the role of direct and indirect

property in investment portfolios and the benefits of international property with special

attention given to the Asian property markets. In addition, a critical review of previous

studies on Asian real estate companies and real estate investment trusts (REITs) will also be

presented.

3.2 Direct Property in an Investment Portfolio

Direct property investments are the types of property that are held directly for investment

purposes such as farmland, commercial, industrial or residential property. Direct property has

always been an essential part of a diversified investment portfolio. Given its significance in

the investment universe, there has been a considerable amount of research on direct property

as an investment asset. This section will provide a critical review on the existing literature

regarding the historical performance, the inflation-hedging characteristics, the diversification

potential and the role of direct property in a portfolio.

3.2.1 The Performance of Direct Property

3.2.1.1 Valuation-based Property Indices

The return and risk characteristics of direct property has received much attention in the

literature. Comprehensive reviews of the literature can be found in Sirmans and Sirmans

(1987); Norman et al. (1995); Seiler et al. (1999) and Benjamin et al. (2001). Previous studies

that used the valuation-based property indices reported a low standard deviation of property

returns. Early studies for the US over the period before 1980 by Wendt and Wong (1965);

Robicheck et al. (1972) ; Roulac (1976); Webb and Sirmans (1980) found that farms and

residential properties outperformed other investment classes as they provide similar or better

returns, but with less variability.

81

Ibbotson and Siegel (1984) compared returns of a composite property index, which

comprised farmland, residential housing and business properties, with those of other

investments for the US over the period 1947-1982. They reported an average annual return

and standard deviation of 8.33% and 3.71% for property; 12.42% and 17.52% for shares and

2.24% and 10.83% for bonds respectively. For the period from the first quarter of 1978

through the first quarter of 1990, Myer and Webb (1994a) conducted a study using the

Russell-NCREIF property index. The average annual return and standard deviation were

9.85% and 7.17% for property; 16.73% and 12.90% for stocks, and 2.78% and 7.37% for

bonds respectively. Similar results were also reported by Coleman et al. (1994) for the period

from the first quarter of 1978 to the fourth quarter of 1989. These results suggested that

property had a lower average return than shares, but lower risk as well. As measured by the

return per unit of risk, it appeared that risk-adjusted returns of property were superior to those

of shares and bonds. However, it is important to note that these results were based on the

valuation-based property indices.

For the UK, Fraser (1985) investigates the returns and risk of property in comparison with

those from ordinary shares and long-term Gilts (British government bonds) for the period

from June 1968 to June 1985. The property returns data was taken from the Jones Lang

Wootton (JLW) index. The annual average returns and standard deviation were 15.2% and

10.5% for property; 10.3% and 13.6% for ordinary shares and 18.1% and 22.1% for long-

term Gilts respectively. In a later study, MacGregor and Nanthakumaran (1992) calculated

the annual real returns and standard deviation of 6.90% and 8.04% for property; 12.80% and

17.41% for equities, and 4.88% and 11.52% for Gilts. These results indicated that

commercial property in the UK, as measured by the JLW index, also provided favourable

return-risk attributes. From a risk-adjusted perspective, these results showed that property

outperformed both equities and bonds in the UK, a result which was in line with that found in

studies in the US.

Newell and Webb (1996) undertook a study over the period from 1985 to1993 for

international property investments in Australia, Canada and New Zealand, among other

countries. The results showed that from a risk-adjusted basis, property returns outperformed

those of shares, but were not superior to those of bonds. Similar results were also reported for

South Africa for the period 1980-1995 by Newell and Webb (1998). The property returns

were computed from the Richard Ellis Property Index for South Africa. The average annual

82

capital return and risk were calculated as 13.0% and 7.83% for property, 16.2% and 21.23%

for stocks and 14.8% and 1.93% for bonds respectively.

There is a potential problem with the aforementioned findings, as property returns and risk

were calculated based on valuation-based indices. The valuation-based values may lag true

market values and true volatility of property returns causing the level of property risk to be

underestimated. In other words, valuation-based indices appeared to be smoothed. This

problem has long been recognised by many authors, including Blundell and Ward (1987),

Giliberto (1988); Geltner (1989a, 1989b, 1991) and Ross and Zisler (1987; 1991), among

others. In order to obtain more reliable data, many researchers have proposed methods to

correct index errors by unsmoothing valuation-based property indices, such as those of

Blundell and Ward (1987), Ross and Zisler (1987), Barkham and Geltner (1994), Fisher et al.

(1994), Newell and MacFarlane (1995a, 1996) Geltner and Goetzman (2000) and Cho et al.

(2003).

When smoothing errors were corrected, the risk level of property appeared to increase

considerably. For example, after unsmoothing the Russell-NCREIF property index for the US

over the period from 1978 through 1985, Ross and Zisler (1991) suggested that the standard

deviation of property should be adjusted upward by a factor of 3 to 5. For the period 1979-92,

Fisher et al. (1994) recommended an adjustment factor of 1.58 to 1.66 for the Russell-

NCREIF series. The standard deviation of real annual capital returns on property increased

from 5.30% to 8.62% or 8.19% respectively, depending on different assumptions about the

level of market efficiency. Also using the Russell-NCREIF index over the period 1980-93,

Newell and MacFarlane (1995a) found that the standard risk estimates of US property needed

to be raised by approximately 80% to be reliable for use by institutional investors and

portfolio managers.

A number of studies have also been conducted for the UK. Brown (1991), for example,

reported that the risk estimates of the Investment Property Databank (IPD) property series

need to be increased by a factor of 3.44 over the period 1987-1990. Similarly, Newell and

MacFarlane (1996) reported the conversion factor of 3.5 for the IPD monthly returns series

over the period 1987-1992. On the other hand, MacGregor and Nanthakumaran (1992)

unsmoothed the JLW property index and showed a rise in the standard deviation from 2.45%

to 4.70%, an increase by a factor of 2.86 and 1.92 for the real and nominal data respectively.

83

Also analysing the JLW index over the period 1970-92, Barkham and Geltner (1994) found

an increase in annual standard deviation of property from 11.8% to 18.6%, an increase by a

factor of 1.57.

Valuation-smoothing analysis has also been undertaken for Canada, Australia and New

Zealand, among other countries, by Newell and Webb (1996). The results showed that the

volatility estimates for property indices in those countries over 1985-1993 needed to increase

by 34% to 47% to adjust for valuation-smoothing and intertemporal correlation. The

volatility adjustment factors ranged from 1.36 to 1.45. It can be seen that the adjustment

multiplier varied across countries, indices and time periods under investigation.

When valuation-based indices were corrected for smoothing errors, several authors have

found that property still appeared to outperformed bonds, but were not necessary superior to

shares. Corgel et al. (1998) studied unsmoothed US data and reported an annual average

return and standard deviation of 12.8% and 7.6% for property, 16.2% and 14.8% for shares

and 10.4% and 8.4% for bonds respectively. MacGregor and Nanthakumaran (1992) found

that unsmoothed property series in the UK generated returns superior to those of bonds, but

not those of equities. Average return and risk were 0.97% and 4.70% for smoothing-adjusted

property, 2.80% and 9.90% for equities and 1.03% and 6.94% for Gilts respectively. In a

recent study, Edelstein and Quan (2006) utilised a sample of arms length market transactions

of 71 large commercial properties between 1979 and 1984. When valuation-smoothing errors

in property indices were corrected, the results revealed that property mean returns and

variance appear to be quite similar to those of stocks.

3.2.1.2 Transaction-based Property Indices

In attempts to overcome the weaknesses of valuation-based data, a number of studies have

employed property series developed from transaction-based data. However, the results on the

risk-adjusted performance of transaction-based property indices varied depending on

different index construction methods, the number of properties included and different time

periods. However, it generally appeared that a transaction-based property index provided a

higher risk level than bonds, but lower than shares. When compared to valuation-based

indices, transaction-based property indices exhibit higher volatility, as measured by a higher

standard deviation.

84

Several studies used hedonic models to construct indices for commercial property. Hoag

(1980), for instance, constructed indices for 463 industrial properties from the first quarter of

1973 through to the fourth quarter of 1978. He reported the quarterly mean return and

standard deviation of 2.11% and 9.53% for the equally-weighted index and 3.38% and 8.61%

for the value-weighted index. The equivalent figures for shares were 0.92% and 10.38% and

those for bonds were 1.55% and 4.02% respectively. According to these results, industrial

property had a risk level between that of shares and bonds. In term of risk-adjusted

performance, property outperformed shares, but not bonds. Cole (1988) reported an average

return of 11.1% and standard deviation of 2.1% for property over the period 1982-86. As for

shares, the average return and standard deviation were 21.1% and 15.4% respectively. The

results suggested that property had a lower mean return and risk than shares. From a risk-

adjusted perspective, results from these studies showed that property outperformed shares.

Fisher et al. (1994) developed a transaction-based index by regressing the ex-post transaction

prices of properties sold out of the Russell-NCREIF database onto a vector of hedonic

variables and time dummy variables for the period 1979-1991. He found the average return

and standard deviation of 0.10% and 9.36% for the property index, and 10.72% and 12.39%

for the S&P 500 index. This suggested that the transaction-based property index also

exhibited lower average return and risk than shares. However, shares perform better

compared to property on a risk-adjusted basis.

Webb et al. (1992) analysed a portfolio of 592 commercial properties over the 1980-1988

period and found that transaction-driven returns exhibited more variance than appraisal-

driven returns. However, most of the individual property risk was idiosyncratic and could be

diversified at the portfolio level and property continues to be a dominant asset class in mean-

variance allocation models. Gatzlaff and Geltner (1998) constructed a repeat sales index for

commercial properties in Florida for the 1982-1996 period. The repeat sales property index

showed higher volatility than the appraisal-based NCREIF Property Index (NPI), with a

standard deviation of 4.07% for the former versus 3.86% for the latter.

3.2.2 The Inflation-hedging Characteristics of Direct Property

Property has been commonly viewed as a good hedge against inflation, as property returns

tend to be correlated with rises in the price level. Early studies in the US provided supportive

evidence of the positive correlation between direct property returns and inflation rates. For

85

example, Fama and Schwert (1977) suggested that residential property was a complete hedge

against both expected and unexpected inflation during the 1953-71 period. On average, the

nominal property returns moved in one-to-one correspondence with inflation. Hartzell et al.

(1987) also documented that property had been an effective hedge against both types of

inflation over the period 1973-1983.

However, early studies on the hedging ability of property used valuation-based property

series and some authors have cast doubts on the data and suggested that transaction-based

data would provide more accurate results. When transaction-based data was employed,

property only provided a hedge against expected inflation, but not an effective hedge against

unexpected inflation. For instance, using transaction-based data, Rubens et al. (1989) found

that property over the period 1960-86 only provided hedging ability against expected

inflation. They also found that not all properties are a perfect hedge against inflation and

different types of property provided different levels of hedging. Similarly, Wurtzebach et al.

(1991) found that office and industrial real estate returns offered no significant hedging

possibility against unexpected inflation. The weak hedging power of US property against

unexpected inflation might be attributed to different economic cycles, as pointed out by

Dokko et al. (1991). They found that commercial property in the US during 1975-1984 kept

up with expected inflation, but did not out-perform it and the impact of expected inflation on

property returns is likely to depend on the macroeconomic environment and local market

conditions. During booming periods with high excess demand, increases in expected inflation

would lead to unanticipated increases in property returns; but this may not occur during

recession periods.

In the UK, property has also been reported to provide partial protection against inflation,

particularly expected inflation (Brown 1991; Brown 1990b; Hoesli et al. 1997; Limmack &

Ward 1988; Miles 1996). Limmack and Ward (1988) indicated that UK property in general

provided a hedge against inflation, particularly in its expected component. In addition, the

industrial property sector had provided a better hedge against inflation than offices and shops.

In a later study, Hoesli et al. (1997) showed that the capital appreciation component of

property offered protection against unexpected inflation, but the income component did not.

The inflation-hedging characteristics of property in Australia (Brown 1990a; Newell 1995a,

1996), Canada (Le Moigne & Viveiros 2008; Newell 1995b), New Zealand (Newell & Boyd

1995), Switzerland (Hamelink & Hoesli 1996; Hoesli 1994) and Ireland (Stevenson &

86

Murray 1999) have also been studied. The conclusions from these studies were similar with

what had been found in the US and UK.

3.2.3 The Diversification Potential of Direct Property

Property returns have been observed to have negative or weakly positive correlation with

those of shares and bonds in the US, UK and Australia. The low correlation between property

and other asset classes suggests diversification benefits when property is included in a mixed-

asset portfolio. Hartzell et al. (1986) analysed commercial property in the US over the period

1973-1983 and reported a correlation coefficient of -0.12 between property returns and stock

returns and of -0.39 between property returns and bond returns. Over the period 1978-1990,

Gyourko and Keim (1992) found the Russell-NCREIF index had a correlation coefficient of

-0.04 with the S&P 500 index. Similar results were also obtained by Fisher et al. (1994). For

the UK, MacGregor and Nanthakumaran (1992) reported a correlation coefficient of -0.19

between annual real returns on property and equities and a correlation coefficient of 0.01

between property and Gilts over the period 1977-1990. For Australia, Eves (2002) found a

correlation coefficient of -0.22 between property returns and share returns and of -0.43

between property returns and bond returns during the period 1990-2000. In addition, capital

returns on rural property had correlation coefficients of -0.08 and -0.15 with those of shares

and bonds respectively.

As discussed above, many authors were critical on the use of valuation-based property series

because of the smoothing effects (Fisher et al. 1994; Geltner 1993). They argued that

valuation-based data might cause the covariance and correlation coefficients between

property indices and other financial indices to be biased towards zero. Therefore, valuation-

based property indices should be adjusted to get a more appropriate level of correlation.

When unsmoothed property returns were used, correlation coefficients between property and

other assets appeared to increase slightly, but are still relatively low. Fisher et al. (1994) used

unsmoothed property indices and calculated the correlation coefficient between property

returns and stock returns at between -0.11 and -0.07, depending on the adjustment parameter.

In the same way, Ziobrowski and Ziobrowski (1997) used the unsmoothed data for the US

during 1970-1995 and computed the correlation between property and stocks to be between

-0.14 and -0.10 and the correlation between property and bonds to be between -0.31 and

-0.28.

87

For the UK, MacGregor and Nanthakumaran (1992) demonstrated that the correlation

coefficients between property and stocks and property and bonds were similarly at -0.11.

Barkham and Geltner (1994) reported an increase in correlation between property and stocks

when the JLW index is unsmoothed, but still relatively low. With a middle value for the

quarterly valuation-smoothing parameter, the correlation between property and shares

increased from 0.22 to 0.38 for the period 1970-1992. These results indicated that even when

unsmooth property data was used, property still appeared to preserve diversification potential

because it had low correlation with shares and bonds.

Several authors have also used transaction-based property indices as a proxy for property. For

example, Fisher et al. (1994) computed a property price index based on transaction prices in

the US and reported a correlation of 0.30 between property series and the S&P 500 index.

This result suggested that transaction-based indices generally had positive but moderate

levels of correlation between property returns and the returns on shares and bonds, suggesting

diversification benefits from including property in mixed-asset portfolios.

As Ball et al. (1998) and Hoesli and MacGregor (2000) have pointed out, the observed low

correlation between property, both in the valuation-based and transaction-base indices, with

other financial assets could result from the temporal aggregation remaining in the series or to

supply cycle effects. Due to the high inertia of the property market or the slow reaction of

property supply to changes in demand, new supply might still be constructed when the

market has started declining. As a consequence, returns are further depressed and the impact

may last for a long period of time. On the other hand, new construction may not commence in

a rising market, causing a supply constraint which in turn can lead to a substantial rise in

returns.

3.2.4 The Role of Direct Property in a Portfolio

As discussed in previous sections, property has a number of favourable characteristics that

warrant inclusion in an optimum mixed-asset portfolio. Property has been observed to have

relatively lower standard deviations as compared to shares. Moreover, property returns

appeared to have weak correlation coefficients with shares and bonds. Lastly, the addition of

property in a mixed-asset portfolio would enhance the portfolio’s inflation hedging ability.

Since property plays an important role in improving the portfolio’s performance, the

88

optimum weight which should be allocated to property is a critical issue that should be

assessed.

Despite a significant volume of research, there has not been a consensus conclusion on the

optimum weight of property in mixed-asset portfolios. A number studies reported very high

percentages of allocation to property for a mixed-asset portfolio. For example, Webb and

Rubens (1986; 1987) studied the optimal compositions of multi-asset portfolios in the US

during the period 1974-1984. They argued that the optimum portfolio allocation to residential

property ranged from 0% to 22% and to business property ranged from 49% to 83%. In a

later study, Webb et al. (1988) recommended that approximately two-thirds of the investment

assets should be allocated to property and one-third to other financial assets. For the period

1978-1983, Coleman et al. (1994) suggested that the optimum weight to property should be

84.7% for the quarterly required return of 3.0% and 15.7% for the quarterly required return of

4%. For the UK, MacGregor and Nanthakumaran (1992), for the period 1977-90, computed

the optimum weight should have been attributed to property in mixed-asset portfolios to be

11% to 88% across the risk spectrum.

Other studies suggested more moderate levels of allocation to property. Ziobrowski and

Ziobrowski (1997) concluded that 20% to 30% should be the optimal weight for property. In

a different way, Fogler (1984) suggested an optimum allocation of 15% to 20% for property

in mixed-asset portfolio. This range was supported by many other authors; such as

Cooperman et al. (1984), Gold (1985), Brinson et al. (1986), Irwin and Landa (1987),

Fristenberg et al. (1988), Ennis and Burik (1991) and Ziobrowski and Ziobrowski (1997).

Similar results were also reported for the UK by MacGregor and Nanthakumaran (1992),

Brown and Schuck (1996) and Lee et al. (1996).

At the lower end, Giliberto (1992) suggested 5% to 15% optimal allocation to property.

Kallberg (1996) found that 9% should be allocated to property in a mixed-asset portfolio,

while Hartzell (1986) ascertained that 3-11% should be optimum. These results were

supported by a survey of Bajtelsmit and Worzala (1995), which calculated the percentage of

property allocation for 159 pension funds in the US and found an average allocation of 4.48%

across firms.

89

3.3 Indirect Property in an Investment Portfolio

As data based on direct property indices possesses a number of problems, such as valuation-

smoothing, illiquidity, lack of transparency and high transaction costs, etc., the use of indirect

property data as an indicator for property performance seems more appealing. There are

several indirect means, by which investments in property can be made; through purchasing of

shares in property companies, property commingled funds or real estate investment trusts

(REITs). There are certain advantages of including securitised property in a mixed-asset

portfolio, as they exhibit unique characteristics. This section however will mainly focus on

the investment characteristics of REITs and their role in a portfolio.

3.3.1 The Relationship between Indirect and Direct Property

One of the most important issues for investors is the relationship between indirect and direct

property investment. Indirect property appears to have a low contemporaneous correlation

with direct property returns, as measured by unsmoothed property indices. Despite having a

low correlation coefficient with direct property (Corgel et al. 1998), many studies have

provided evidence that indirect and direct property markets are integrated over the long-run.

Giliberto (1990) was among the first researchers to report that equity REITs and direct

property were associated by a common factor, with lagged values of REITs explaining direct

property performance. Gyourko and Keim (1992) reported that lagged REIT returns can

predict direct property returns after controlling for persistence in the valuation-based Russell-

NCREIF series. Myer and Webb (1993; 1994b) examined REITs and commercial property in

the US and reported that REIT returns Granger-caused commercial property returns, but the

reverse did not hold. Glascock et al. (2000) found that REITs were co-integrated with the

direct property market.

Barkham and Geltner (1995) examined the linkages between unsecuritised and securitised

property markets in the US and UK for the evidence of price discovery. After adjusting for

valuation-smoothing, direct property returns appeared to lag securitised property returns of up

to two years in the US and one year in the UK. For Australia, Newell and MacFarlane

(1995b) and Jones Lang Wootton (1995) reported a one or two year lead by property trusts

over the property market. These results to some extent suggest that the indirect property

market was more informationally efficient than the direct property market.

90

Chiang et al. (2006) performed REIT-based mimicking portfolio analysis for US REITs for

the period from January 1993 to December 2003. The authors found that the market beta of

REIT portfolios appeared to converge to the market beta of the NCREIF Index, suggesting a

strong linkage between REIT returns and the underlying property value. MacKinnon and Al

Zaman (2009) found that REIT and direct property returns were both reverting to the mean,

which was caused by a tendency of property transaction prices to overshoot inflation.

Oikarinen et al. (2011) also presented evidence of the cointegration between indirect

(NAREIT) and direct (NCREIF) property indices. Many researchers (Clayton & MacKinnon

2003; Ghosh et al. 1996; Ziering et al. 1997) have attributed the reason for the integration of

indirect and direct property markets in recent years to the increased participation by

institutional investors in indirect property investment.

3.3.2 The Performance of Real Estate Investment Trusts (REITs)

Research on the performance of REITs in the US generally suggested that REITs have similar

risk and return characteristics to those of stocks over the period prior to the 1990s. Studies

over the period 1963-1974 by Smith and Schuman (1976) and Smith (1980) found that REITs

provided returns comparable to those of diversified portfolios of common stocks. In a later

study, Mueller et al. (1994) computed the average monthly return and standard deviation of

equity REITs, stocks and small-value stocks in the US for the period from January 1976 to

June 1993. REITs also appeared to have a similar return and risk structure to that of shares

and small-value stocks. The average monthly return and standard deviation were 1.46% and

0.0385% for REITs, 1.31% and 0.0444% for shares and 1.75% and 0.0634% for small-value

shares respectively. Han and Liang (1995), Liang et al. (1996) and Brueggeman and Fisher

(1997) also reported that REIT return and risk behaved far closer to those of shares than those

of direct property.

Some authors hypothesised that the observed high correlation between REITs and shares is a

result of the use of leverage in REITs. However, when REIT returns were unlevered, equity

REITs still had a similar risk level to that of shares and higher than that of unsmoothed direct

property. Fisher et al. (1994) ungeared REIT returns over the period from 1982 to 1992 and

reported the standard deviation of 14.25% for REITs, 12.39% for shares and 8.19% or 8.62%

(depending on desmoothing parameter used) for unsmoothed property respectively. These

91

results suggested that REITs and shares might be subject to similar market risk factors. These

common factors might come from macroeconomic events and monetary policy shocks.

Over the long-run, empirical studies on the association between REITs and shares provided

inconsistent results. This was due to several institutional changes in the REIT sector in the

early 1990s. Mueller and Mueller (2003) noted that since 1997, the composition of REIT

portfolios has become more like that of the private market. In addition, the tax changes in the

Revenue Reconciliation Act of 1993 and the REIT Modernisation Act of 1999 had made

REITs more attractive to institutional investors (Cannon & Vogt 1995).

Glascock et al (2000) employed cointegration, Vector Autoregessive (VAR) models and the

Engle-Granger procedure to explore the causality and long-run economic linkages between

REIT and stock returns over the period 1972-1997. The results revealed that REITs behaved

more like stocks and less like bonds after the structural change in the early 1990s. This

conclusion, however, was challenged by a recent study conducted by Lee and Chiang (2010).

They used the Johansen procedure, which is more robust than the Engle-Granger procedure,

to test the co-integration relation between REITs and shares over the period 1978-2008. The

results showed that REITs behaved like common stocks during the period 1978-1993, but

become less like stocks and more like direct property after the early 1990s structural break.

The view that REITs have become more like property and less like stocks was shared by

Clayton and MacKinnon (2003) and Ghosh et al. (1996). Clayton and MacKinnon (2003)

found that the sensitivity of REIT returns to direct property showed a significant increase in

the 1990s, indicating that REITs had been becoming more integrated with direct property

than financial assets. Ghosh et al. (1996) suggested REITs becoming more like property and

less like stocks for two reasons; firstly, REITs appeared to be less liquid than comparable size

stocks and secondly, institutional investment in REITs was growing. Han et al. (1980) also

documented that institutional investors prior to 1990 invested more of their funds in other

stocks than in REITs, whereas after 1990 they invested more of their funds in REITs than in

other stocks in the market. He also noted that the percentage of institutional holdings was

positively correlated with the performance of the REITs.

The risk and return characteristics of REITs were also found to vary greatly by property

types. Mueller and Laposa (1996) examined historical return series for REITs by different

92

property types over the period 1972-1995. They found that property-type REIT groups (i.e.

apartment, retail, hotel, self-storage, office, industrial, healthcare and manufactured homes)

originally moved closely from 1972 to 1985, but since then diverged substantially between

property groups. On the other hand, the co-movement of equity REITs within the same

property type have strengthened after the new REIT era (1990-2004) as suggested by Chiang

(2010). He further suggested that a high institutional participation, a low insider ownership

and a large market capitalisation were associated with a high inter-property-type price

synchronicity.

The determinants of REIT return performance have also been analysed in the past. Among

the early researchers, Chan et al. (1990) employed the capital asset pricing model (CAPM)

and arbitrage pricing theory (APT) approach to examine equity REIT returns. Chen et al.

(1997) tested the factor loading model (FLM) and the macrovariable model (MVM) for the

ability to explain REIT returns. The authors concluded that the MVM was better than the

FLM. Ling and Naranjo (1997; 1999) and Ling et al. (2000), among others, have all

supported the macro-factor approach. Chui et al. (2003a, 2003b) investigated the

determinants of REIT returns in a multifactor framework and also found a structural break in

the early 1990s. Prior to the 1990s, REIT returns were associated with four factors:

momentum, size, turnover and analyst coverage. However, after the 1990s, only momentum

was a significant factor.

3.3.3 The Inflation-hedging Characteristics of REITs

The empirical evidence on the inflation-hedging effectiveness of REITs has been

inconclusive. However, it was generally observed that REITs were not a good hedge against

inflation, especially against unexpected inflation. Gyourko and Linneman (1988) studied the

inflation hedging ability of REITs in the US over the period from the second quarter of 1972

to 1986 and found that REITs provided partial protection against expected inflation, but not

against unexpected inflation. Similar observations were reported over various periods by

Goebel and Kim (1989), Chan et al. (1990), Yobaccio et al. (1994) and Larsen and McQueen

(1995).

Chatrath and Liang (1998) examined long-run inflation hedging effectiveness of REITs over

the period from January 1972 to December 1995. The analysis showed no evidence that REIT

returns provided hedging ability against temporary or permanent components of inflation.

93

The evidence of a positive relationship between REIT returns and inflation over the long-run

was also weak.

REIT returns have also been observed to provide a perverse hedge against inflation by a

number studies. For example, Murphy and Kleinman (1989) found that REIT returns were

negatively correlated with both expected and unexpected inflation over monthly holding

periods from 1972 to 1985. Over annual holding periods, there was no significant relationship

between REIT returns and either component of inflation. Evidence of REITs as a perverse

inflation hedge were also reported by Park et al. (1990). However, Glascock et al. (2002)

ascertained that the observed negative relationship between REIT returns and inflation were

spurious. It was in fact a manifestation of the effects of changes in monetary policies.

Simpson et al (2009) applied a pooled estimation methodology to 195 listed REITs for the

period 1981-2002 and found asymmetric responses of REIT returns to inflation. When

expected and unexpected inflation were separated into positive and negative changes, REIT

returns rose in response to both increases and decreases in inflation. In a recent study, Lee

and Lee (2011) analysed long-run inflation property of US REITs over the period 1971-2007.

The results indicated that REITs had served as a positive, long-run hedge against expected

inflation following the 1990s structural break.

3.3.4 The Diversification Potential of REITs

It has been observed that the correlation between REITs and financial assets have been found

to be dynamic over time. The diversification potentials of REITs in a mixed-asset portfolio

therefore also appeared to be time-dependent over the short-run. For the period 1978-94,

Brueggeman and Fisher (1997) reported a correlation coefficient of 0.69 between equity

REITs and shares and 0.41 between REITs and bonds. Mull and Soenen (1997) analysed the

correlations between US REITs, domestic stocks and domestic bonds for the time period

1985-94. They found that REITs had a positive correlation coefficient with stocks of 0.61 and

a low correlation coefficient with bonds of 0.28. However, these results depend heavily on

the time period under consideration. It has been argued that the high correlation between

equity REITs and shares was due to leverage effects. When Fisher et al. (1994) ungeared

REIT returns, the correlation coefficient between REITs and shares reduced to 0.44 for the

period 1978-92.

94

Khoo et al. (1993) found that the correlation coefficients between US REITs and common

stocks have declined from the 1970s to the 1980s. Ghosh et al. (1996) also reported

decreasing correlation coefficients between REITs and shares over the period 1985-96. The

monthly correlation coefficients were 0.77 for 1985-7, 0.71 for 1988-90, 0.38 for 1991-3 and

0.40 for 1994-June 1996 respectively. Similar results of declining correlations were reported

by Chandrashekaran (1999), Clayton and MacKinnon (2001), Conover et al. (2002), Bley and

Olson (2005) and Westerheide (2006) over various periods from 1972 to 2001. Khoo (1993)

attributed the declining association between REITs and common stocks to the increasing

levels of information about REITs, as measured by the number of analysts following the

REIT industry. In addition, Clayton and MacKinnon (2001) and Chiang et al. (2006)

suggested that the increase in institutional investment in REITs had the effects of shifting

downward the correlation between REIT and non-REIT stocks, since institutional investors

were more equipped to reveal the value of underlying properties backing the securities.

On the other hand, a number of studies indicated that the correlation coefficients between

REITs and stocks had increased in some periods. Westerheide (2006), Cotter and Stevenson

(2006) and Huang and Zhong (2006) found REIT and stock correlations generally increased

during various study periods from as early as January 1999 to as late as December 2005.

Feng et al. (2006) and Ambrose et al. (2007) suggested that REITs’ betas relative to stocks

had increased following the inclusion of REIT in the S&P 500 and other broad stock market

indices starting in October 2001.

Case et al. (2010) reconfirmed the dynamic correlations between REIT and stock returns over

the period from January 1972 through September 2008. The results suggested that REIT-

stock correlations formed three distinct periods; July 1978 to August 1991, August 1991 to

September 2001 and September 2001 to late 2008. The correlation coefficients fluctuated

above 0.59, falling to 0.30 and increased steady to 0.59 during the three periods respectively.

Fei et al. (2010) also examined the dynamics of correlation between REITs and stocks over

the period from January 1987 to May 2008. They documented that time-varying conditional

correlations could be explained by macroeconomic variables, such as term and credit spreads,

inflation and the unemployment rate. Furthermore, they found a strong relationship between

correlations and REIT returns and that the patterns were distinguishable for different types of

REITs. Peng and Schulz (2012) found that portfolios selected with a forecast dynamic

covariance matrix were less risky than portfolios constructed with the static matrix. However,

95

benefits of using a dynamic covariance matrix for active portfolio management are mostly

offset by transaction costs.

Over the long-run, as the REIT market was observed to be cointegrated with the direct

property market, but not with the share market, REITs therefore are likely to have long-term

diversification benefits in a share portfolio. Westerheide (2006) studied the relationship

between REITs and property securities with other financial assets over the period 1990-2004

for eight countries; US, Canada, Australia, Japan, the Netherlands, Belgium, France and

Germany. He concluded that despite relatively high short-term correlation coefficients with

the stock markets in most countries, there was usually no cointegration between the REIT and

the stock markets. On the other hand, Oikarinen et al. (2011), among others, provided

evidence on the long-term integration between the REIT and direct property markets.

3.3.5 The Role of REITs in a Portfolio

The role of REITs in a mixed-asset portfolio has also been studied in the past. Despite

relatively high short-term correlation with shares, REITs had a positive role in diversifying

mixed-asset portfolios. Mueller et al. (1994) examined the benefits of including REITs in

multi-asset portfolios for the period from January 1976 to June 1993. They pointed out that

including REITs to mixed-asset portfolios would have yielded an additional return between 1

and 14 basis point per month or 12 and 168 basis point per annum (without compounding) for

the same level of risk. The efficient allocations of REITs were 14% for low risk, 82% for

medium risk and 3% for high risk portfolios respectively. For the period from the first quarter

of 1982 to the fourth quarter of 1993, Liang et al. (1996) reported weights in the 15-20%

range for equity and apartment REITs.

In line with evidence on the time-varying correlations between REITs and stocks, the value

of REITs as a component in a mixed-asset portfolio was also found to be time-dependent.

Mull and Soenen (1997), for instance, found that REITs offered diversification benefits as a

portfolio component for the period 1985-94. Nevertheless, sub-period analyses indicated

significant temporal differences in REIT efficiency as a portfolio component. Similarly, Case

et al. (2010) reported that optimal REIT allocations in mixed-asset portfolios fluctuated

around 17%-42% over different periods from January 1972 through September 2008.

96

3.4 International Property in an Investment Portfolio

The growth and integration of global financial markets as well as changes in international

politics and economic policies have resulted in increased global investment opportunities

(Newell & Worzala 1995). The benefits and potential of international property investments

have also received ample attention from investors and academics. Sirmans and Worzala

(2003a; 2003b) provided critical reviews of the growing body of literature on the benefits of

international diversification using direct and indirect property investments. The following

sections summarise the findings from previous studies on this issue.

3.4.1 International Direct Property Investment

Previous studies on the inclusion of international direct property with other financial

investments in a mixed-asset portfolio context have indicated that an international property

diversification strategy certainly provided benefits. Among the first studies was one

conducted by Ross and Webb (1985) over the period 1958-79. The authors examined

property rental indices in 14 countries and found that property had less systematic risk than

other financial assets and could be an appropriate portfolio enhancer. On the other hand,

Marks (1986) was concerned about how international investors might benefit from investing

in US direct property, as measured by the Prudential Property Investment Separate Account

(PRISA), over the period 1978-84. He found that US property outperformed other countries’

equity indices except Japan.

Webb and Rubens (1989) examined the potential diversification benefits from including UK

direct property in a US mixed-asset portfolio over the period 1926-86. The correlations

suggested that UK real estate should provide additional diversification benefits if added to a

domestic-only portfolio. However, the role of UK direct property in an optimum portfolio

was limited, due to the high volatility of UK during the study period. A similar study of the

US and UK property markets was also conducted by Giliberto (1989). The author ascertained

that US and UK markets were on different property and economic cycles and suggested that

benefits might be gained by holding investments in both countries. For the period 1970-90,

Gordon (1991) found that UK property returns has no or negative correlation with all of the

US financial assets except inflation. He also reported that investors would gain from holding

both US and UK property in one portfolio.

97

Chua (1999) conducted an analysis on the role of international property in investment

portfolios from five countries; France, Germany, Japan, the UK and the US. The data was

adjusted for the high taxes, transaction costs and asset management fees incurred in property

investments. Using a mean-variance optimization, the author concluded that international

property indeed had a significant role in a global mixed-asset investment portfolio. Using the

bootstrapping technique for property data, Cheng et al. (1999) also investigated the benefits

of international property investments from the perspective of the British, Japanese and

American investors. They, however, recommended that only the high risk-tolerant investors

should allocate into international property in the range of 5-10 percent.

Hoesli et al. (2002) extended the investigation of an international property investment

strategy to seven different countries; the US, the UK, France, the Netherlands, Sweden,

Switzerland and Australia over the period 1987-2001. The study showed low correlation

coefficients of direct and indirect property, suggesting the potential for diversification

benefits on an international investment strategy. Nevertheless, the authors noted that the

results might vary upon different countries.

A number of authors have also advocated for a global property investment strategy. Sweeney

(1988) examined property rent growth rates for 16 countries over the period 1970-86. The

author advised that investors would have earned superior returns for a global investment

strategy. In a later study, Sweeney (1989) computed the efficient frontier and optimum asset

allocations for 11 office markets from 1978 to 1988. The results indicated that for the

maximum return and risk portfolio, the entire portfolio should be allocated to Sydney.

Nevertheless, for the minimum return and risk level, international investments should be

made to 7 out of the 11 countries. Similar conclusions were also reported by Reid (1989).

Arnold and Kavanaugh (1988) compared the JLW property indices for London, Paris and

Sydney with the NCREIF index for the US over the period 1978-87. The authors reported

that Paris and Sydney significantly outperformed the US over this period, and London and

Paris property markets provided high returns with relatively low risk. In addition, US

property returns were positively correlated with Sydney and London property returns, but

negatively correlated with returns in the London property market.

Wurtzebach (1991) looked at office investments in five major metropolitan areas and

compared individual markets with a global portfolio over the period 1978-89. The author

98

found that adding global property investments reduced volatility of the US-only portfolio and

the global portfolio had lower volatility than the single-country portfolios. D’Arcy and Lee

(1998) reviewed the development of a property portfolio strategy for Europe over the period

1990-96 by three options: country, property type and city type. The authors saw a need for

portfolios to be constructed on a country basis. In addition, a portfolio concentrated in

secondary cities outperformed a similar one in capital cities and the composition of the

optimal portfolio exhibited considerable inter-temporal instability.

Hasting and Nordby (2007) examined the benefits of global diversification in a property

portfolio using historical data for 18 international office markets over the period from the first

quarter of 1991 to the fourth quarter of 2005. The authors reported that the optimal global

portfolio, in local currency, would have outperformed a US-only portfolio by 171%. Even

when the volatility of exchange rates was accounted for, the global portfolio would still

outperform a US-only portfolio by 92%. In a recent study, Lim et al. (2008) utilised the

International Property Databank data for 15 countries over the period 1971-2005. The results

indicated that there were significant diversification benefits from combining property

investments from certain countries. Nonetheless, other combinations provided no significant

benefits.

In summary, the general conclusions from studies regarding the role of international direct

property in an investment portfolio suggested that international property investments

certainly provided benefits both in a mixed-asset and inter-property portfolio context.

International efficient portfolios outperformed domestic portfolios, and portfolios that

included international property outperformed those that did not. However, caution should be

given to the treatment of exchange rate fluctuations as it might offset the benefits of

international diversification.

3.4.2 International Indirect Property Investment

A significant amount of research has also been conducted to examine the role of international

indirect property in a mixed-asset portfolio. This section summarises previous findings on the

benefits of international diversification in indirect property. Most of the studies however used

international property company stocks rather than REITs as a proxy for indirect global

property investments. This was due to the lack of available data on global REITs in the past.

99

Asabere et al. (1991) analysed a mixed-asset portfolio containing international property

companies (as measured by the Capital Perspectives Index, CPI), US REITs, US shares and

bonds over the period 1980-88. They reported that the optimal portfolio would have 5.6% in

international property companies and 40% in US REITs. Also using the CPI indices, Kleiman

and Farragher (1992) compared the international indirect property returns with US REIT

returns and found that the global property index showed higher price earnings multiples, but

US REITs outperformed when dividend yields were taken into account.

Barry et al. (1996) studied property companies in emerging markets using the Emerging

Market Database from 1989 to 1995. The author concluded that emerging market property

should be included in the efficient portfolios. The minimum risk portfolio of both developed

and developing property would have an 11% allocation in emerging markets. Eichholtz

(1996) compared international property, stock and bond returns over the period 1985-94 for

nine countries. The author found that correlation coefficient between international property

markets was lower than those between stock and bond markets. Furthermore, he reported that

an international property portfolio outperformed domestic-only portfolios in the UK, Japan,

US and France, as well as both an international stock and bond portfolio.

Eichholtz and Koedjik (1996) investigated property securities in 25 countries and noted low

correlation coefficients between US REITs and foreign property markets. The results also

suggested that diversification into global property companies would enhance portfolio

performance. However, they noted that hedging the exchange rate risk would negatively

impact the diversification benefits. Eichholtz (1997) calculated the correlation coefficient

between property indices and stock indices in 19 countries and found that Asian property

markets were highly correlated, while European markets were not. He suggested that the high

correlation among Asian property securities was a manifestation of the stock market rather

than pure property market, since Asian listed property markets were dominated by property

development companies.

Gordon et al. (1998) examined the role of international property securities and equities to a

mixed-asset portfolio of shares, bond and domestic property for 14 countries over the period

1984-97. The authors reported the domestic asst correlation coefficients ranged from 0.24 to

0.96, suggesting that for some countries, diversification benefits could be achieved from an

international investment strategy. In addition, they found that for individual countries, the

100

cross-country property stocks were not as highly correlated as the cross-country equities.

When efficient portfolios for the US were constructed, the minimum variance portfolio

standard deviation decreased by 46 basis points when international property stocks were

included. According to Liu and Mei (1998), international property would provide

diversification benefits, although the majority of the benefits resulted from the currency

fluctuations and the results were time-dependent.

By examining the Global Property Research (GPR) indices for property securities in 14

countries from 1984 to 1997 and comparing them with the US NAREIT index for US

property investment, Gordon and Canter (1999) found that the efficient portfolios that

included international property securities beat both a US-only portfolio and a portfolio with

US investments and international shares. Stevenson’s (2000) analysis of both the hedged and

unhedged property indices for 10 countries over the period from 1978 to 1997 revealed that

the correlation coefficients between property indices were relatively low (0.008 to 0.461), but

even lower if the hedged indices were used (-0.124 to 0.189). The author also found that both

international property stocks and shares were dominant in the efficient portfolios. However,

when currency risk was adjusted, the position of international property was reduced. These

countries still benefited from holding international indirect property investments; e.g. Japan,

The Netherlands and Singapore.

Maurer and Reiner (2002) analysed mixed-asset portfolios consisting of shares, bonds and

international property companies from five countries. Efficient portfolios were constructed

from both the US and German investors’ perspective and the results suggested that

combining international property improved the performance of mixed-asset portfolios

particularly for the low to medium-risk levels for investors of both countries. They also

indicated that the adoption of a hedging strategy using forwards also helped to mitigate the

currency risk for international investors. Conover et al. (2002) investigated the risk-return

characteristics of foreign property stocks in comparison to US shares, US property and

foreign shares over the period 1986-95. Their results indicated that the foreign property

stocks had a significant role in the mixed-asset portfolios. If international shares and property

stocks were added to a US domestic portfolio, the risk level for the maximum-return portfolio

fell from 8.57% to 5.73%. As for the minimum-risk portfolio, the return would increase from

0.67% to 0.71%, while the standard deviation would fall from 3.07% to 2.92% when

international indirect property was included.

101

Studies on inter-property diversification strategy have also been conducted for global indirect

property. Wilson and Okunev (1999) , for instance, analysed the role of property stocks in a

diversified portfolio for US, UK and Australia over the period 1980-87. Their results

indicated that the share and property stock markets were not co-integrated over the long-run;

therefore, a property-share diversification strategy should be beneficial. The efficient

frontiers showed that efficient portfolios, in most cases, were diversified across all three

markets. In later studies, both Pierzak (2001) and Bigman (2002) confirmed that an

internationally diversified property portfolio outperformed a domestic property portfolio from

a US investor’s perspective.

3.5 Asian Property in an Investment Portfolio

In the past few years, Asian property markets have attracted substantial attention from global

investors, especially institutional investment funds, seeking international diversification

opportunities. Capital investment into Asian properties has been rising steadily. However, the

volume of research on the Asian property markets is still thin in comparison to those on the

US or UK property markets. This section will summarise the previous findings concerning

Asian property markets. The studies will be organised into three major categories: (1) Asian

direct property, (2) Asian property companies and (3) Asian real estate investment trusts.

3.5.1 Asian Direct Property

3.5.1.1 The Performance of Asian Direct Property

Newell et al. (1996) and Chiang and Ganesan (1996) were among the first researchers who

studied the performance of direct property investment in Hong Kong. Newell et al. (1996)

corrected the valuation-smoothing problem and adjusted the volatility of the Hong Kong

direct property index. Chiang and Genesan (1996) examined the risk and return

characteristics of Hong Kong direct property market using transaction-based data collected

during the 1980s and 1990s. They reported that, in terms of risk and return, direct property

investment outperformed property stocks. Based on the security market line, property

investment in the office, retail and industrial sectors all yielded superior returns to those

required by the market portfolio, comprising both property and shares.

102

Brown and Chau (1997) tested for the existence of excess returns for the commercial property

sectors in Hong Kong over the period from 1980 to 1995. The analysis of both transaction-

and valuation-based property series revealed that investors could possibly earn excess returns

in the commercial property market. However, when transaction costs were taken into

account, the ability of earning excess return did not persist for long periods. The office sector

offered the greatest opportunity for earning excess returns. Webb et al. (1997) found that

historical returns of commercial property in Hong Kong were attributed to three main factors;

initial current yield, growth in net operating income and pricing movements.

The effects of political uncertainty in Hong Kong property market associated with the 1997

transfer of sovereignty to the People's Republic of China has also been investigated by Chau

(1997). The analysis suggested that there was a discrete jump in the risk premiums when the

information about the repossession became public in 1983. The increase in the risk premium

was more prominent in the non-residential than the residential sector. However, the risk

premium has declined gradually leading to 1997, suggesting improved investors’ confidence

levels, but still higher than the pre-1983 level.

Tse et al. (1999) examined the performance of the office property sector in the three major

Chinese cities of Shanghai, Guagzhou and Shenzhen. The results showed that the Shanghai

office market outperformed other cities due to its superior risk-adjusted returns. In addition,

the Shanghai office market also experienced the lowest systematic risk. Webb and Tse (2000)

further reported that office property prices in Guangzhou and Shenzhen consistently contain

highly significant information about the future movements of office property prices in

Shanghai. On the other hand, there was a lead-lag relationship between the prices in

Shenzhen and Guangzhou.

For India, Newell and Kamineni (2007) reviewed the performance of property markets

(office, retail and residential) of the two largest cities; New Delhi and Mumbai, over the

period from the second quarter of 1998 to the fourth quarter of 2005. The authors reported

that the property markets underperformed the stock market over the sample period, with most

markets improving their performance in later years. They further suggested that the

deregulation of the capital markets and international investment in India was necessary to

improve the growth of real estate investment in India.

103

Recently, Newell et al. (2009a) provided an extensive study on the performance of

commercial property across seven major markets in Asia; Shanghai (China), Hong Kong

(China), Singapore (Singapore), Bangkok (Thailand), Jakarta (Indonesia), Kuala Lumpur

(Malaysia) and Manila (The Philippines), over the period from the fourth quarter of 1998 to

the first quarter of 2007. On a risk-adjusted basis, Bangkok and Hong Kong were the best

office and retail market respectively. Kuala Lumpur underperformed all the other Asian

markets for both office and retail property.

There has also been evidence of linkages between property prices and the general stock

markets in some Asian markets. Ong (1994) suggested that there was a cointegrating

relationship between property and stock prices. Similar phenomenon was also found for

Hong Kong (Cheung et al. 1995; Fu 1993), with the stock market leading the property market

in price changes. On the other hand, Tse (2001) reported that unexpected changes in both

Hong Kong residential and office property prices were also important determinants of the

change in stock prices.

3.5.1.2 The Inflation-hedging Characteristics of Asian Direct Property

Ganesan and Chiang (1998) examined whether real properties in Hong Kong were a good

hedge against inflation over the period from 1984 to 1994. According to the co-integration

tests, there was no long-term equilibrium between inflation and property returns. Therefore,

the authors concluded that direct property in Hong Kong was not a good hedge against

inflation, because rents in Hong Kong were often fixed for lengthy periods, therefore not

generally indexed to price levels.

Chu and Sing (2004) investigated the inflation hedging characteristics of four major cities in

China; Beijing, Chengdu, Shanghai and Shenzhen for the periods 1988-2002 and 1996-2002

for annual and quarterly data respectively. The results revealed that all property sectors in the

four Chinese cities were a poor hedge against both expected and unexpected inflation. The

authors offered two possible explanations. Firstly, the periods under investigation were

periods with low and negative inflation in China. Secondly, property prices in China over the

sample periods have experienced rapid appreciation due to a massive influx of foreign capital

investment. Similarly, Zhou and Clements (2010) also found that Chinese property was an

ineffective hedge against inflation.

104

For Taiwan, Fang et al. (2008) examined the relationships between housing returns and

inflation over the period from 1991 to 1996. The empirical findings revealed that property

returns had negative relationships with both expected and unexpected inflation. The authors

concluded that direct property investments in Taiwan offered no inflation protection to

investors. Not surprisingly, the results were in line with those found in China and Hong

Kong, given the strong economic linkages between the three markets.

On the contrary, Singapore property investment was reported to be a good hedge against

inflation. Sing and Low (2001) tested for the inflation hedging ability of property and non-

property assets over the period from the first quarter of 1978 to the fourth quarter of 1998.

Their results showed that direct property in Singapore provided a better hedge against

inflation than non-property assets. Industrial property was the most effective hedge against

both expected and unexpected inflation, whilst retail property offered only a significant hedge

against expected inflation.

3.5.1.3 The Diversification Potential of Asian Direct Property

The diversification benefits and the role of Asian direct property in a portfolio have also been

investigated in a number of studies. Asian commercial property has shown to offer

diversification benefits for both international and local investors. In an early work, Addae-

Dapaah and Yong (1998) examined the diversification strategy from a Singaporean investor’s

perspective across nine office markets in the Asia Pacific region over the period 1984-97.

The authors reported that without currency adjustments, the correlation coefficients were

weakly positive or negative, suggesting diversification gains. Accounting for exchange rate

fluctuations, they found that the increased currency risk was not statistically significant.

Therefore, investors would have significant gains from diversification across Asian property

markets and the currency risk did not substantially impact returns on the fully diversified

office portfolio.

Brown and Li (2000) explored the effectiveness of diversifying property portfolios

geographically across Hong Kong. Geographic analysis was carried out for Hong Kong

residential and office properties over the period 1984-95. The authors reported that an intra-

city diversification strategy to some extent outperformed naive strategies based on an equal

allocation across all districts. However, the differences in performance were statistically

trivial.

105

For Japan, Maroney and Naka (2006) studied the benefits of diversifying into property by

Japanese investors from September 1957 to March 2004. Their results showed that Japanese

direct property provided significant diversification benefits. However, the portfolio weights

on Japanese property were relatively small compared to their composition in Japanese

household wealth.

3.5.1.4 The Relationship between Asian Indirect and Direct Property

Previous studies on the linkages between indirect and direct property markets in Asia have

provided mixed evidence. Ong (1995) employed the co-integration test to examine the

relationship between property stock and direct property returns from 1976 to 1988. The tests

revealed that there was no long-term contemporaneous relationship between property stock

and direct property prices. Sing and Sng (2003) revisited the issue of market integration

between the securitised and unsecuritised property market in Singapore, using the data from

the first quarter of 1975 to the second quarter of 1999. They concluded that the securitised

and unsecuritised property markets were not perfectly integrated, but partially integrated.

Interestingly, the authors also found that the information of the direct property market had a

significant impact on the returns of the indirect property market.

In contrast, the Hong Kong direct and indirect property markets were found to have strong

linkages. Newell and Chau (1996) assessed the relationship between Hong Kong property

company and property market performance over 1984-94. They found a common pure

property factor that influenced both property company and property market returns. Tse and

Webb (2000, 2001) reported that both the expected and unexpected changes in commercial

property prices were important determinants of the changes in property security prices for

Hong Kong. Chau et al. (2003) and Newell et al. (2004; 2007) showed that the performance

of property companies in Hong Kong has become more associated with direct property and

less with shares.

Lee and Ting (2011) studied the relationship between residential prices and property

companies in Malaysia. The authors adopted a vector-autoregressive model (VAR) over the

study period 1999-2009. The results showed that there is a uni-directional influence from the

direct to indirect property market and changes in the public market Granger caused changes

106

in the housing market. Evidence of Granger-causality between the indirect and direct

property markets was also reported for Indonesia (Newell et al. 2009a).

3.5.2 Asian Property Companies

3.5.2.1 The Performance of Asian Property Companies

The risk-return characteristics and the performance of property stocks in some Asian markets

have been examined in recent years. The performance of Singapore property companies have

been extensively researched by Liow (1997, 1998, 2000, 2001) in various studies. The author

found that property stocks performed worse on a risk-adjusted basis than the market index.

Moreover, there was lack of consistency in the performance ranking of the companies over

time. In comparison with the direct property market, the results showed that Singaporean

property companies also underperformed direct properties.

For Hong Kong, Newell et al. (2007) reported that property companies had higher average

return and risk than other major asset classes; e.g. shares, bonds and direct property

investments. In term of risk-adjusted performance, property companies were superior to other

assets over the period from the second quarter of 1984 to the third quarter of 2004. The

average annual return and risk were reported to be 24.4% and 42.7% for property companies,

18.4% and 30.2% for stocks, 6.9% and 0.9% for bonds and 9.3% and 14.9% for direct

property respectively. However, the risk and return characteristics of property companies

were very similar to those of shares.

The performance of listed property companies in the Asian emerging markets of Vietnam,

The Philippines, Thailand and South Korea, among others, have also been previously

assessed by Nguyen (2010, 2011) and Pham (2011a, 2011b). In general, the findings

indicated that property companies in those countries generated lower average return and

higher risk than shares and bonds. On a risk-adjusted basis, property companies significantly

underperformed the other financial assets.

In a regional context, Newell et al. (2009b) compared the risk-adjusted performance of the

listed property securities in the Asian international financial centres (IFCs) of Tokyo,

Singapore and Hong Kong to six major non-IFC markets in Asia over January 1998 to March

2008. The authors reported significant risk‐adjusted returns for these Asian IFCs, with this

having been enhanced in the later years. The cross-market dynamics of Asian property

107

companies have also been studied by various authors (Liow 2007; Liow 2011; Liow & Adair

2009). In examining the volatility spillover effects of listed companies in 13 Asian markets,

Nguyen (2012) found significant spillovers from the developed property markets to the

emerging property markets in Asia. The results also suggested that Asian property stocks did

not outperform their corresponding general stock markets.

In a global context, Asian property companies were reported to be inferior to the US REITs

and UK property companies. Liow and Sim (2006) investigated property companies in ten

Asian markets; Hong Kong, Malaysia, The Philippines, Thailand, Indonesia, China, Korea,

Taiwan and Japan, for the period from January 1990 to June 2003.The authors indicated that

Asian property security markets produced relatively lower compound returns than the US

REIT and UK property stock markets since the 1990s. They also had higher risk levels than

their US and UK counterparts.

In summary, most of the previous studies have reported that the risk and return characteristics

of Asian property companies resembled those of the general stock markets. This might be

explained by the fact that property companies accounted for a major part of the stock markets

in many Asian markets, especially in the emerging markets. In addition, returns of property

securities had greater volatility than those of shares and bonds. Property securities in Tokyo,

Singapore and Hong Kong seemed to outperform other Asian markets on a risk-adjusted

basis. However, Asian property securities underperformed shares, bonds and their

counterparts in the US and UK.

3.5.2.2 The Inflation-hedging Characteristic of Asian Property Companies

Contrary to findings in the US and UK property markets, returns of property securities in

some Asian markets had not been found to be an effective hedge against inflation. Liow

(1997) reported that Singaporean property stocks failed to provide ex-post inflation

protection for investors. Lee et al. (2011b) investigated the inflation hedging properties of

property stocks in Malaysia, The Philippines and Taiwan. They found that the property stocks

in the three East Asian markets did not provide an effective hedge against inflation over the

long-run.

108

3.5.2.3 The Role of Asian Property Companies in a Portfolio

Studies on international property investments have provided evidence that Asian property

stocks offered significant benefits to international investors. Among the first authors to

examine diversification benefits of securitised property markets in Asia, Addae-Dapaah and

Kion (1996) examined the diversification benefits of international diversification from the

perspective of a Singaporean investor. The authors examined the property stock indices for

four Asian markets; Singapore, Malaysia, Japan, Hong Kong and the three developed markets

of the UK, Canada and Australia. The results indicated that the correlation coefficients

between most countries were relatively low, ranging from 0.03 (Japan – Hong Kong) to 0.79

(Singapore – Malaysia). Based on the efficient frontiers, the authors concluded that an Asian

regional diversification strategy would enhance returns and reduce risk for Singaporean

investors. In addition, Canadian and British investors would also benefit from holding

investments in the Singaporean property stock markets.

Garvey et al. (2001) investigated the linkages between public property markets in Australia,

Hong Kong, Japan and Singapore over the period from January 1975 to March 2001. The

long-term analysis failed to support evidence of co-integration between the markets,

suggesting long-term diversification opportunities in property in some Asian countries. Bond

et al. (2003) examined listed property companies from 14 countries; including the three Asian

markets of Hong Kong, Japan and Singapore, over the period from February 1990 to

December 2001. When the effects of global market risk were controlled, a country-specific

market risk factor was found for Asian property indices with high significance. Therefore, the

authors suggested that diversification strategies were likely to be more effective in the Asian

than the European markets. Liow and Adair (2009) examined the role of Asian property

companies in regional and international mixed-asset portfolios. The data was for all 13 Asian

countries as well as the US and UK property security markets. The authors advised that

international investors would gain more by investing in Asian property security portfolios

rather than mixed-asset portfolios.

The role of property companies in domestic mixed-asset portfolios has also been assessed for

the emerging markets of Malaysia (Lee & Ting 2009), Thailand (Pham 2011a), Vietnam

(Nguyen 2010), South Korea (Pham 2011b) and The Philippines (Nguyen 2011). Contrary to

the diversification benefits offered to international investors, Asian property stocks did not

offer significant diversification benefits or portfolio return enhancement to domestic

109

investors. This was due to the high correlation coefficients between property stocks and the

domestic share markets. In addition, Asian property stocks in the emerging Asian markets

tended to be dominated by property developers rather than property operating companies.

3.5.3 Asian REITs

Despite a short history of development, the last decade has witnessed the strong growth of the

REIT sectors in Asia. The emergence of REITs has improved the liquidity and efficiency of

the Asian real estate market, as well as the corporate governance of real estate organizations

(Ooi et al. 2006). Research studies on Asian REITs, however, have been thin and modest as

compared to those in the US and UK, given that REITs were new market instruments and

data was limited. This section provides a summary of some of the early studies on Asian

REITs.

3.5.3.1 The Performance of Asian REITs

Chiang et al. (2008) analysed the time-varying performance of REITs in Japan (J-REITs) and

Singapore (S-REITs), among others, over the period 2001-5 and 2002-5 respectively. For

Japan, the average return and volatility were 4.21% and 5.02% for J-REITs, 2.42% and

8.44% for shares and 0.22% and 1.78% for bonds respectively. In regard to Singapore, the

average return and volatility were 5.61% and 5.22% for S-REITs, 4.37% and 5.24% for

shares and 0.68% and 3.49% for bonds respectively. For both countries, REITs appeared to

yield higher average returns than shares and bonds, but have a risk level between that of the

other two asset classes. On a risk-adjusted basis, REIT indices in both countries outperformed

their respective domestic share and bond indices.

Lee and Ting (2009) reported the risk and return profile of Malaysian REITs (M-REITs) from

1991 to 2006. The equally-weighted REIT portfolio was found to provide the highest

annualised return of 15.6% which was substantially higher than that of shares (8.4% p.a.) and

bonds (4.8% p.a.). Accordingly, the equally-weighted REIT index also provided the highest

standard deviation of 0.218 as compared to those of 0.080 and 0.002 for shares and bonds

respectively. The performance of Malaysian Islamic REITs has also be assessed by Newell

and Osmadi (2009). They reported that the Islamic M-REIT index underperformed

conventional M-REITs and the stock market index on a risk-adjusted basis. Newell and

Osmadi (2010) further found in their survey that the key factors for the future development of

110

M-REITs include tax issues, provision of professional M-REIT services, availability of

quality properties and strategic property locations.

Newell et al. (2010) assessed the risk-adjusted performance of Hong Kong REITs (HK-

REITs) over the period 2005-8. HK-REITs were observed to deliver strong risk-adjusted

returns and to have been robust during the GFC. Over the sample period, HK-REITs were

reported to yield an annualised average return of 3.48% and standard deviation of 27.59%,

which were higher than those of shares and bonds. HK-REITs outperformed shares, but not

bonds in terms of risk-adjusted return, as measured by the reward-to-risk and Sharpe ratios.

Pham (2011a) examined the development of REITs in Thailand (Thai-REITs) over the period

from October 2003 to September 2010. The author found that Thai-REITs provided the

lowest annualised return of 0.56% p.a., significantly underperforming bonds (4.61% p.a.),

shares (13.10% p.a.) and property companies (3.35% p.a.). In terms of risk level, Thai-REITs

however recorded a lower level of risk (10.42%) than shares (27.54%) and property

companies (35.92%). On a risk-adjusted basis, Thai-REITs were the poorest performing asset

class, with reward-to-risk and Sharpe ratio of 0.05 and -0.20 respectively. The author further

noted that the poor risk-adjusted performance reflects the restrictive regime of Thai-REITs,

as compared to other REIT markets.

Pham (2011b) also investigated the risk and return characteristics of Korean REITs (K-

REITs) over the period from January 2002 to December 2010 for South Korean REITs,

bonds, shares and property companies. The results revealed that K-REITs performed poorly

over the sample period as compared to shares, bonds and property companies. Annualised

return and risk were reported of 0.56% and 10.42% for K-REITs, 13.10% and 27.54% for

shares and 4.61% and 2.49% for bonds respectively. Accordingly, K-REITs provided lower

returns than other assets, but had a risk level that was between those of shares and bonds.

Newell and Peng (2012) assessed the significance, risk-adjusted performance and portfolio

diversification benefits of J-REITs in a mixed-asset portfolio context in Japan over 2001-

2011. The authors reported that amongst the major asset classes, Japanese REITs (J-REITs)

delivered the best risk-adjusted returns and portfolio diversification benefits over 2001-2011,

with enhanced risk-adjusted returns in the post-GFC period. Newell and Peng (2012) also

reported strong diversification benefits of Taiwanese REITs (T-REITs) in investment

111

portfolios. Even though T-REITs underperformed the construction share sector in Taiwan,

they still enhanced portfolio risk-adjusted returns.

Other research topics on Asian REITs include those on market reaction to public news (Sing

et al. 2002), downside risk asset allocation framework (Sing & Ling 2003), the impact of the

global financial crisis (Kim 2009), performance of China-centric REITs (Quek & Ong 2008),

a diversification strategy for Asian REITs (Cheok et al. 2011), issues concerning Asian REIT

IPOs (Kutsuna et al. 2008; Lee et al. 2011a), seasonal equity issuance of J-REITs and S-

REITs (Ong et al. 2011), REIT market risk (Ting & Choi 2011) and insider shareholding in

the emerging Asian REITs (Kundus & Sing 2011)

In summary, research studies on Asian REITs have generated some interesting observations.

The return performance of Asian REITs as compared to that of shares varied across different

markets. For more developed and more mature Asian markets such as Japan, Singapore,

Hong Kong and Malaysia, REITs provided greater average returns than shares. In contrast,

REITs in less mature markets such as Thailand and South Korea underperformed the stock

markets. In terms of risk, previous studies commonly found that REITs had a risk level

between that of shares and bonds. This result is similar to findings of REITs in the US.

3.5.3.2 The Role of Asian REITs in a Portfolio

Chiang et al. (2008) investigated the role of Japanese and Singapore REITs in mixed-asset

portfolios of stocks, bonds and direct property. They reported that REITs could provide

diversification benefits to a multi-asset investment portfolio. However, the authors also noted

that due to the time-varying nature of the correlation, active management was advised and

REITs should not be viewed as a complete substitute for direct property investment.

For Malaysia, Lee and Ting (2009) examined the position of Malaysian real estate securities

in a mixed-asset portfolio from 1991 to 2006. The authors utilised the mean-variance and

downside risk optimisations. The results showed that property shares offered little

diversification benefits, while the equally-weighted M-REIT portfolio did provide portfolio

enhancement under the mean-variance and downside risk frameworks. The results further

revealed that the equally-weighted and value-weighted portfolios behaved differently. Newell

and Osmadi (2009) also reported that the benefits of Islamic M-REITs for a mixed-asset

112

portfolio were evident, especially during the GFC. Islamic M-REITs were seen to display

defensive characteristics and entered the portfolio at low-risk levels.

The diversification potential of Thai-REITs in a mixed-asset portfolio was analysed by Pham

(2011a). The results showed that Thai-REITs only played a minor role in the mixed-asset

portfolio at low risk levels due to their poor risk-adjusted performance. In a later study, Pham

(2011b) examined the risk and return characteristics and diversification potential of K-REITs.

Similar to Thai-REITs, the results reveal that K-REITs performed poorly over the entire

period as compared to shares, bonds and property companies. Consequently, K-REITs did not

represent a substantial weight in a mixed-asset portfolio. However, K-REITs have a low

correlation with shares and property companies, suggesting some level of diversification

potential for share and property investors.

3.5.3.3 The Volatility Dynamics in Asian REIT Markets

Studies on the volatility spillover effects across Asian securitised property markets have also

been investigated by a number of studies. Most studies however, analysed Asian property

companies rather than REITs. For example, Garvey et al. (2001) found minimal evidence of

both short-term and long-term co-integration between the property security markets in

Australia, Hong Kong, Japan and Singapore. On the contrary, Liow et al. (2009) detected a

strong and positive connection between real estate security market correlations and their

conditional volatilities. They also found the international correlation structure of real estate

securities and the broader stock markets are linked to each other.

The only published paper that examined the volatility transmission of REIT returns in Asia

was that of Li and Yung (2007), who investigated the interactions between the Pacific

(Australia, Hong Kong, Japan and Singapore) and the Atlantic REIT markets (U.K. and

U.S.). The authors found a significant international spillover of REIT volatility from the

Atlantic to the Pacific regions. Pham (2012) studied the dynamics of returns and volatility in

the developed and emerging Asian REIT markets. The author reported strong volatility

linkages between the REIT markets of Singapore and Hong Kong. Volatility appeared to

spillover from Singapore, Hong Kong and South Korea to other Asian REIT markets. The

Asian REIT markets formed three clusters, with Hong Kong, Singapore and South Korea

being the main volatility emitters in the region, Japan and Taiwan are volatility receivers,

while Malaysia and Thailand acted as intermediary markets. Tsai and Lee (2012) analysed

113

whether there is a convergent behaviour in the price indexes in Asian REIT markets (Japan,

Singapore, Hong Kong, Malaysia, Thailand, Taiwan and Korea). The results showed that

Asian REIT markets were more connected with the USA, than with the Japan REIT market.

Overall, the literature review has revealed a lack of studies on Asian REITs due to their

recent development. In particular, investment performance of the emerging REIT markets in

Asia has been limited. Within a regional and global context, the role and benefits of Asian

REITs in a mixed-asset portfolio has not been investigated. In addition, few research studies

have employed advanced econometric techniques in examining Asian REIT returns. This

study aims to bridge these research gaps and to provide a rigorous analysis on the

performance of Asian REITs and their role in an investment portfolio.

114

4 CHAPTER 4

DATA AND METHODOLOGY

4.1 Introduction

This chapter describes the data and research methodologies employed in this study. Section

4.2 presents the sources of data and provides descriptions of the variables. Section 4.3

demonstrates the index construction method. Section 4.4 and Section 4.5 propose the research

methodologies for the performance analysis and regression analysis.

4.2 Sources of Data

The quantitative data was obtained primarily through DataStream. In addition, the data was

verified with information from the primary sources such as the local securities exchanges and

the central banks. For the performance analysis of individual countries, all data are collected

in domestic currencies. For regional and global analysis, data are in US dollars to avoid

currency exchange fluctuations and to maintain the consistency of analysis.

4.2.1 Domestic Context

The data employed in the analysis for individual markets consisted of monthly total returns of

the local REITs, shares, property companies and bonds. Table 4.1 reports the descriptions of

the data series utilised in this study. The start of the analysis for each market was chosen

based on the listing date of the first REIT of the market in question. The data time series

commence from September 2001 for Japan, July 2003 for Singapore, November 2003 for

Thailand, July 2004 for Hong Kong, August 2005 for Malaysia, February 2006 for Taiwan

and June 2006 for South Korea. The end of the analysis was at April 2012.

The securitised REITs in Japan, Singapore, Hong Kong and Taiwan were represented by the

S&P REIT Total Return Index for each of those markets respectively. As indexes for listed

REITs were either not sufficient nor available for Malaysia, Thailand and South Korea, the

value-weighted REIT indices for those markets were constructed, namely the UWS REIT

Total Return Index. The index construction methods and lists of constituents included in the

UWS REIT Total Return Indices will be explained and discussed in the later section.

115

Listed property securities in Japan, Singapore and Hong Kong were represented by the S&P

Property Total Return indices, while Malaysia, Taiwan and Thailand were proxied by the

DJTM Real Estate Total Return indices. In addition, the UWS South Korea Real Estate

Return Index was created to represent property companies traded on the Korean Stock

Exchange (KRX). Asian equity markets were measured by the DJTM Total Return indices

for listed stocks on the local security exchanges. Bond returns were calculated by the long-

term 10-Year Government Bond rates, while the cash rates were computed based on short-

term Treasury Bill (T-Bill) rates or the Overnight Call Money rates for each market,

depending on their availability.

4.2.2 Regional and Global Context

In the regional and global analysis, both monthly and daily returns on REITs and shares were

collected for the seven Asian markets, as well as for US and Australia (Table 4.2). In addition

to the seven Asian REIT indices described above, REIT data for the US and Australia was

collected based on the FTSE/NAREIT All REIT Total Return Index and the S&P Australia

REIT Total Return Index respectively. A pan-Asian REIT index, i.e. the UWS Pan-Asia

REIT Total Return Index, was developed based on a value-weighted portfolio of all currently

listed REITs across the seven Asian REIT markets. In addition, three sub-indices for Asian

REITs were also constructed, i.e. the UWS Pan-Asia Excluding Japan REIT Index, the UWS

Developed Asia REIT Index and the UWS Emerging Asia REIT Index. For the this study, the

three REIT markets of Japan, Singapore and Hong Kong were classified as developed

markets, while the less mature REIT markets of Malaysia, Taiwan, Thailand and South Korea

were categorised as emerging markets.

Correspondingly, the DJTM Total Return Indices for seven Asian countries and territories

were used to represent the local equity markets. The US and Australia share markets were

represented by the conventional S&P 500 Composite Total Return Index and the ASX All

Ordinaries Total Return Index respectively. Pan-Asian share indices and sub-indices were

also constructed using the capitalisation-weighted approach. The mathematical formulas for

the index construction methods will be described in the following section.

116

Table 4.1: Data Descriptions – Local Context

Markets Variables Data Series

Japan

REITs S&P JAPAN REIT - TOTAL RETURN INDEX

Property

Companies S&P JAPAN PROPERTY - TOTAL RETURN INDEX

Shares DJTM JAPAN - TOTAL RETURN INDEX

Bonds JP INTEREST-BEARING GOVERNMENT BONDS - 10-YEAR

Cash JP 3 MONTH INTERBANK RATE (MONTH AVERAGE

Singapore

REITs S&P SINGAPORE REIT - TOTAL RETURN INDEX

Property

Companies S&P SINGAPORE PROPERTY - TOTAL RETURN INDEX

Shares DJTM SINGAPORE - TOTAL RETURN INDEX

Bonds SP 10 YR GOVERNMENT BOND YIELD

Cash SP SINGAPORE THREE MONTH TREASURY BILL RATE

Hong Kong

REITs S&P HONG KONG REIT - TOTAL RETURN INDEX

Property

Companies S&P HONG KONG PROPERTY - TOTAL RETURN INDEX

Shares DJTM HONG KONG - TOTAL RETURN INDEX

Bonds HK EXCHANGE FUND NOTE YIELD - 10 YEAR (EP)

Cash HK TREASURY BILL RATE - 3 MONTH

Malaysia

REITs UWS MALAYSIA REIT - TOTAL RETURN INDEX

Property

Companies DJTM MALAYSIA REAL ESTATE - TOTAL RETURN INDEX

Shares DJTM MALAYSIA - TOTAL RETURN INDEX

Bonds MY GOVERNMENT BOND YIELD - 10 YEAR

Cash MY TREASURY BILL DISCOUNT RATE - 3 MONTH

Taiwan

REITs S&P TAIWAN REIT - TOTAL RETURN INDEX Property

Companies DJTM TAIWAN REAL ESTATE - TOTAL RETURN INDEX

Shares DJTM TAIWAN - TOTAL RETURN INDEX

Bonds TW TAIWAN GOVERNMENT BOND YIELD 10 YEAR (EP)

Cash TW INTERBANK RATE - 2 TO 6 MONTHS

Thailand

REITs UWS THAILAND REIT - TOTAL RETURN INDEX Property

Companies DJTM THAILAND REAL ESTATE - TOTAL RETURN INDEX

Shares DJTM THAILAND - TOTAL RETURN INDEX

Bonds TH GOVT BOND YIELD - LONGTERM

Cash TH BANK OF THAILAND BONDS 3 MONTH

South Korea

REITs UWS SOUTH KOREA REIT - TOTAL RETURN INDEX Property

Companies UWS SOUTH KOREA REAL ESTATE - TOTAL RETURN INDEX

Shares DJTM SOUTH KOREA - TOTAL RETURN INDEX

Bonds KO YIELD 10-YEAR GOVERNMENT BONDS NADJ

Cash KO CALL MONEY RATE - OVERNIGHT, BROKERED (MTH.AVG.)

117

Table 4.2: Data Descriptions – Regional and Global Context

Sectors Markets Data Series

REITs

Japan S&P JAPAN REIT U$ - TOTAL RETURN INDEX

Singapore S&P SINGAPORE REIT U$ - TOTAL RETURN INDEX

Hong Kong S&P HONG KONG REIT U$ - TOTAL RETURN INDEX

Malaysia UWS MALAYSIA REIT U$ - TOTAL RETURN INDEX

Taiwan S&P TAIWAN REIT U$ - TOTAL RETURN INDEX

Thailand UWS THAILAND REIT U$ - TOTAL RETURN INDEX

South Korea UWS SOUTH KOREA REIT U$ - TOTAL RETURN INDEX

US FTSE/NAREIT ALL REITS $ - TOTAL RETURN INDEX

Australia S&P AUSTRALIA REIT U$ - TOTAL RETURN INDEX

Pan-Asia UWS PAN-ASIA REIT U$ - TOTAL RETURN INDEX

Pan-Asia Ex. Japan UWS PAN-ASIA EX-JP REIT U$ - TOTAL RETURN INDEX

Developed Asia UWS DEVELOPED ASIA REIT U$ - TOTAL RETURN INDEX

Emerging Asia UWS EMERGING ASIA REIT U$ - TOTAL RETURN INDEX

Shares

Japan DJTM JAPAN U$ - TOTAL RETURN INDEX

Singapore DJTM SINGAPORE U$ - TOTAL RETURN INDEX

Hong Kong DJTM HONG KONG U$ - TOTAL RETURN INDEX

Malaysia DJTM MALAYSIA U$ - TOTAL RETURN INDEX

Taiwan DJTM TAIWAN U$ - TOTAL RETURN INDEX

Thailand DJTM THAILAND U$ - TOTAL RETURN INDEX

South Korea DJTM SOUTH KOREA U$ - TOTAL RETURN INDEX

US S&P 500 COMPOSITE - TOTAL RETURN INDEX

Australia ASX ALL ORDINARIES - TOTAL RETURN INDEX (~U$)

Pan-Asia UWS PAN-ASIA SHARE U$ - TOTAL RETURN INDEX

Pan-Asia Ex. Japan UWS PAN-ASIA EX-JP SHARE U$ - TOTAL RETURN INDEX

Developed Asia UWS DEVELOPED ASIA SHARE U$ - TOTAL RETURN IND.

Emerging Asia UWS EMERGING ASIA SHARE U$ - TOTAL RETURN INDEX

4.3 Index Construction

This study employs the total return index as the performance measure rather than the price

index. A total return index considers dividends and other distributions realised, in addition to

price movements over a given period of time, while a price index only tracks price

movements. As most REITs are required to distribute their net profits, the dividend

components in REIT total returns can be substantial. A total return index is a more accurate

measure of performance.

Data used for the computation of the UWS total return indices was sourced from Datastream.

The market capitalisation-weighted total return indices were based on free float outstanding

shares and units. They were computed using the standard total return pricing methods based

on the standard industry compilation and pricing methods employed by Thomson Reuters

(2011). The total returns include capital gains and dividend gains over a given holding period.

118

When a dividend was issued, the price of the constituent falls in the exact amount of the

dividend per share or unit amount. The UWS total return (TR) index is computed as follows:

(1)

where,

is the price of security at time and is the original

time period,

is the number of securities in the index,

is the number of shares or units in security at time ,

is per share or unit divided on ex-date.

4.4 Performance Analysis

4.4.1 Return Measurements

The performance was measured by the total returns. Total returns are continuously

compounded and calculated as the difference in natural logarithms of the total return index

value for two consecutive periods. The total return for each period is calculated as follows:

(2)

where is the end of period value of the total return index, which includes both the

price and dividends during the holding period.

The mean return is the average value of returns across time represents expected performance

(ex-ante) or achieved performance (ex-post). The formula for computing the mean of a series

of periodic returns is as follows:

(3)

119

where is the ex-post mean return. While it is acknowledged that the ex-post mean return

from a given historical sample may not be equal to the ex-ante (expected) mean return,

denoted E(R), it is a reasonably good guidance of the expected return in the near future.

4.4.2 Risk Measurements

The volatility of an asset is measured by the standard deviation of a total return index. The

standard deviation represents the magnitude of dispersion around the mean or the expected

value. It is estimated from a historical sample of periodic returns, by the following formula:

(4)

where is the standard deviation from a historical series of periodic return. In this study,

the sample standard deviation , is used as an estimator for the population standard deviation

of .

4.4.3 Risk-adjusted Return Measurements

4.4.3.1 Risk/return Ratio (Coefficient of Variation)

The risk-adjusted returns of an asset are measured by the risk/return ratio or the coefficient of

variation (CV). CV is a normalised measure of dispersion of a probability distribution also

known as unitised risk or the coefficient of variation. It is defined as the ratio of the amount

of risk (standard deviation) to the average return (risk/return ratio). The CV can be estimated

as:

(5)

where is the standard deviation of returns and is the average return.

4.4.3.2 Sharpe Ratio

Another measurement for the risk-adjusted return of an asset is the Sharpe ratio. It is

measured by the risk premium of the risky asset divided by its volatility, written as follows:

120

(6)

where is the average return of the risky asset, is the average return of the risk-free asset

and is the standard deviation. The risk-free rates were proxied by the cash rates of the

individual markets under analysis.

The popularity of the Sharpe ratio over other risk-adjusted performance measures, e.g. beta,

Treynor and Sortino ratio, is due to its mathematical simplicity and theoretical intuition.

However, like the coefficient of variation one of the limitations of the Sharp ratio is that it

relies on the standard deviation which might not be an accurate reflection of the investors’

perception of risk. This is because the standard deviation treats both the upside and downside

potential of returns as risk. However, investors may differentiate between these outcomes.

Some investors are mainly concerned about downside returns and only treat the downside

potential as risk. In addition, the standard deviation is calculated based on the assumption that

returns are normally distributed. Abnormalities such as skewness and excess kurtosis can

make standard deviation biased.

4.4.4 Correlation Coefficient

The diversification potential between two assets in this study is measured by the correlation

coefficient between them. The correlation coefficient ranges from -1 to 1. A value of less

than or close to zero imply negative or no linear correlation between the two assets. As a

result, the combination of negatively or lowly correlated assets will help to reduce the

portfolio volatility. The formula for the correlation coefficient for asset and is:

(7)

where is the correlation coefficient for asset and over periods. The sample statistic

is used to estimate the corresponding population correlation coefficient .

121

It is important to note that despite being a popular tool to measure the association between

two assets, the correlation coefficient suffers from the limitation that it can only measure a

linear relationship. As a resultsm non-linear dependencies between two time series cannot be

captured. In addition, the correlation analysis relies on the condition that the random

variables follow a normal or a multivariate t distribution. Embrechts et al. (2002) have

provided a detailed discussion on the limitations and pitfalls of correlation coefficient as a

measure of dependency.

4.4.5 Portfolio Theory and Diversification

In the 1950s, Markowitz (1952) developed a rigorous approach to mathematically quantify

the optimum allocation of assets in a portfolio. The theory was called the Mean-variance or

Modern Portfolio Theory (MPT), which earned him a Nobel memorial prize in 1990. The

model has since been widely adopted among professional investors and is the fundamental

discipline of strategic investment decision making.

The purpose of MPT is to find the optimum proportion invested in individual assets in order

to either to maximise portfolio return for a given level of risk or minimise risk for a given

level of return. Suppose that there are N assets in a mixed-asset portfolio, the portfolio

expected return is a weighted average return of each asset’s expected return, estimated as:

(8)

where is the expected return on a multi-asset portfolio, is the proportion of the

portfolio invested in the ith

asset and is the expected return of the ith

asset.

The expected portfolio standard deviation (risk) is a combination of the different assets’

variances, their proportion in the portfolio and the correlation of one asset’s variance with

that of another. The expected portfolio return standard deviation is expressed as follows:

(9)

122

where is the portfolio risk, and are individual asset standard deviations and is the

correlation coefficient for asset and .

An investor can reduce portfolio risk by holding combinations of assets which are not

perfectly positive correlated . A combination of risky assets is efficient if it

has the minimum possible level of risk (standard deviation) for a given level of expected

return (Figure 4.1). Given a target value for the mean return of the portfolio, the

efficient frontier is found by minimising the portfolio risk , subject to:

(10)

where short selling is not allowed and is the maximum weight constraint for asset . The

optimal weight ( ) of each asset in the efficient portfolio can then be solved via quadratic

programming. In this study, a portfolio optimiser model has been built using Visual Basic for

Applications (VBA) programming in Microsoft Excel.

123

Figure 4.1: Hypothetical Mean-variance Efficient Frontier

4.5 Regression Analysis

All regression analyses in this study was conducted using the EViews 7 software programme.

The regression analysis follows the standard econometric procedure. First of all, the time

series data for Asian REITs will be examined for basic descriptive characteristics including

mean, median, maximum, minimum, standard deviation, skewness, kurtosis, Jarque-Bera

statistics. Secondly, preliminary statistical models will be estimated in order to perform

specification and hypothesis testing on the coefficients of the models. These tests include the

tests for serial correlation, stationarity, normality, heteroskedasticity and autoregressive

conditional heteroskedasticity. Thirdly, the preliminary specification will then be modified

based on the test results. Finally, additional diagnostics testing will be conducted in order to

ascertain whether the models effectively capture the original biases in the dataset.

4.5.1 Serial Correlation Testing

Time series are often observed to have serial correlations, meaning that the residuals are

correlated with their own lagged values. Serial correlation presents a number of serious

problems for regression analysis and ordinary least squares (OLS) estimation is no longer

efficient since estimates are biased and inconsistent. In addition, the standard errors computed

based on the simple OLS formula are not correct and generally understated.

Portfolio Risk (Standard Deviation)

Po

rtfo

lio

Ret

urn

Efficient Frontier

Individual Assets

High risk/

high return

Medium risk/

medium return

Low risk/

low return

124

This study uses the Ljung-Box (LB) -statistic tests for autocorrelation. If there is no serial

correlation in the residuals, the autocorrelations and partial autocorrelations at all lags should

be nearly zero and all -statistics should be insignificant with large p-values. The LB Q-

statistic tests of up to 24 orders were performed and reported. The -statistics at lag is

computed as:

(11)

where is the -th autocorrelation and is the number of observations.

4.5.2 Unit Root and Stationarity Testing

Another important test for regression series is the unit root and stationarity test. Detailed

discussion on the basis unit root theory can be found in Davidson and MacKinnon (1993),

Hamilton (1994) and Hayashi (2000) among others. Stationary testing is necessary because

financial time series tend to exhibit non-stationary behaviour meaning that they have variant

mean, variance and autocorrelation. The use of non-stationary variables in regression analysis

can lead to spurious estimation as the standard errors are biased. Therefore, it is important

that the time series data must be transformed to stationarity form prior to analysis. Unit root

tests can be employed to assess whether a time series is non-stationary or stationary using an

autoregressive model. In this study, three unit root tests are carried out; the Augmented

Dickey-Fuller (ADF) test, the Phillips-Perron (PP) test and the Kwiatkowski, Phillips,

Schmidt and Shin (KPSS) test.

4.5.3 The Augmented Dickey-Fuller (ADF) Test

The standard Dickey-Fuller (DF) for REIT return series (second moment) was carried out

following the following specification:

(12)

125

where are optional exogenous repressors which may consist of a constant or a constant and

trend, and are the estimated parameters and the are assumed to be white noise. The

null and alternative hypotheses can be stated as:

(13)

and evaluated using the conventional for :

(14)

where is the sample estimate of and is the standard error coefficient.

If , is a non-stationary time series and the variance of increased with time and

approaches infinity. If , is a stationary series or trend-stationary series, depending on

whether the model consists of a trend variable. In other words, if the test results fail to reject

the null hypothesis of a unit root , then the return series is statistically stationary.

However, the basic DF procedure is only valid if the series is an AR(1) process. The

assumption of white noise disturbances would be violated if the return series is correlated

at higher order lags. This problem could be solved by the Augmented Dickey-Fuller (ADF)

test which corrects for higher-order correlation. The series is assumed to follow an AR( )

process and the test regression equation (12) becomes:

(15)

In the analysis, the ADF tests have been performed using both a constant and a linear trend.

In addition, the lag length was chosen, based on the automatic bandwidth and lag length

selection in EViews, which was sufficient to remove serial correlation in the residuals. The

financial data series will also be tested for jumps or financial breaks.

126

4.5.4 The Phillips-Perron (PP) Test

An alternative unit root testing process is the Phillips-Perron (PP) test developed by Phillips

and Perron (1988) .The PP procedure follows the standard DF regression model (12). The

of the coefficient was modified to correct for the serial correlation in the time

series. The is calculated by the following formula:

(16)

where is the estimate, is the of , is the coefficient standard error and s

is the standard error of the test regression and is the estimator of the residual spectrum at

frequency zero. is a consistent estimate of the error variance in equation (12) based on:

(17)

where is the number of regressors. Similar to the ADF test, the PP procedure was executed

using both a constant and a linear trend and the optimum lag length. The series is assumed to

be non-stationary under the null hypothesis.

4.5.5 The Kwiatkowski, Phillips, Schmidt and Shin (KPSS) Test

In addition to the ADF and PP tests, the Kwiatkowski, Phillips, Schmidt and Shin (KPSS)

unit root test was also employed. Unlike the other two tests, the KPSS test proposes that the

data series is stationary or trend-stationary under the null hypothesis. The KPSS is based on

the residuals obtain from the regression of on the exogenous variable :

(18)

The Lagrange Multiplier (LM) statistic is defined as:

(19)

127

where is an estimator of the residual spectrum at frequency zero and is a cumulative

residual function:

(20)

based on the residuals .

4.5.6 Heteroskedasticity Testing

Financial time series have often been observed where the volatility of residuals tends to be

related to the volatility of recent residuals. These are also known as volatility clustering

effects or autoregressive conditional heteroskedasticity (ARCH) effects. If the ARCH effects

are not treated, it might result in loss of efficiency. The LM test for ARCH in the residuals is

performed based on the regression:

(21)

where is the residual and the null hypothesis is that there is no ARCH up to order in the

residuals. The LM test statistic is asymptotically distributed as under general

conditions.

4.5.7 Volatility Spillover Analysis

To investigate the dynamic transmission of returns and volatility in these Asian REIT

markets, the Exponential Generalized Autoregressive Conditional Heteroskedastic

(EGARCH) model is utilised. The EGARCH model, which was developed by Nelson (1991),

is an extended form of the generalized ARCH model. ARCH-type models have been

commonly used in modelling changes in the volatility of financial time series. A complete

survey of the theory and application of ARCH models in finance can be found in Bollerslev

et al. (1992), Bera and Higgins (1993) and Pagan (1996), among others. Importantly, Engle

and Ng (1993) and Stevenson (2002) have provided evidence in favour of the EGARCH

model over other GARCH-type models. The advantage of the EGARCH specification is that

it allows for the testing of an asymmetry impact from positive and negative changes in REIT

128

returns. In addition, the EGARCH model can also capture the leverage effect of market

volatility on REITs. The leverage effect arises as REIT price fall the debt-equity ratio

increases causing REIT price to be more volatile. This phenomenon is better measured by the

EGACRH model.

The EGARCH specification employed in this study is written as follows:

(22)

(23)

where is the natural logarithm difference of the total return price indices and = 1, 2, 3,

4, 5, 6, 7 (1 = Japan, 2 = Singapore, 3 = Hong Kong, 4 = Malaysia, 5 = Taiwan, 6 = Thailand,

7 = South Korea). For market , is the information set at time and is the

conditional variance. Conditional on , is assumed to be normal distributed with a

zero mean and variance .

Equation (22) describes the conditional mean returns of seven Asian REIT markets, where

they can be affected by historical own innovations (own-mean spillovers) when and

cross-market returns (cross-mean spillovers) when . The term measures the direct

effects that a change in return to the market would have on the market. In order to

resolve the problem with serial autocorrelation, an MA(1) term is introduced to capture the

dependency of current returns ( ) on immediate past own errors ( ).

The dummy variable is incorporated in the model to evaluate the effects of the GFC on

each market. The dummy variable takes the value one in the period from March 2008 to

March 2009 and zero otherwise. While it is still unclear about the exact dates of the start and

end of the GFC, March 2008 is used as the beginning of the GFC in this study, since it was

129

the month that Bear Stearns was taken over by J.P. Morgan. March 2009 is chosen as the end

for the GFC, because it was the month that the Asian REIT index rebounded strongly from its

trough since the start of the GFC.

Equation (23) reflects the EGARCH representation of the variance of . According to the

EGARCH specification, the conditional variance of the returns in each market is an

exponential function of past own, cross-market standardized innovations and past own

conditional variance. After considering the Akaike Information Criterion (AIC) and

diagnostics on standardized residuals of different parsimonious models, an EGARCH(2,2)

model is fitted for the analysis. The parameter measures the asymmetric volatility

transmission mechanism. When , then a positive shock (good news) has the same

effects as the negative shock (bad news) of the same magnitude; when , then a

negative shock increases volatility more than a positive shock; and when , a negative

shock increases volatility, while a positive shock reduces volatility. The persistence of

volatility is measured by .

Volatility spillovers from the market to the market are measured by . Statistically

significant values of suggest that past volatility shocks in cross-markets influence the

current volatility of the market. The term measures the impact of the GFC on REIT

volatility. If is positive and statistically significant, it indicates that the crisis affects Asian

REIT returns, making them more volatile.

130

5 CHAPTER 5

THE PERFORMANCE AND SIGNIFICANCE OF ASIAN

REITs IN DOMESTIC MIXED-ASSET PORTFOLIOS

5.1 Introduction

As REITs have become important investment vehicles for listed property in Asia, it is crucial

for investors and market participants to understand the risk and return profile of Asian REITs.

In particular, investors who are considering investing in Asian REITs are primarily concerned

with the following questions. Firstly, how do Asian REITs perform relative to the local stock

markets? Secondly, what is the risk-adjusted performance of Asian REITs as compared to

property companies? Because both are indirect property investment vehicles, it is important

for investors to understand how they differ from each other in terms of risk or returns. Given

this, investors can decide which one is a better option to gain exposure to the Asian property

market. The third and most important question would be whether the inclusion of Asian

REITs in a mixed-asset portfolio would provide diversification benefits. If REITs should be

included in investors’ portfolios, then what should be the optimum allocation to REITs? Last

but not least, what was the impact of the GFC on Asian REIT performance and its role in a

mixed-asset portfolio and how have they performed post-GFC? This chapter will address

these key investment issues.

The analysis presented in the following sections was conducted based on monthly total return

data for REIT, bond, share and property company indices in the seven REIT markets in Asia;

Japan, Singapore, Hong Kong Malaysia, Thailand, Taiwan and South Korea. The risk-

adjusted performance for local REITs, bonds, shares and property companies was assessed.

The statistics on monthly standard deviation and returns were annualized to represent the

annual risk and average return. Both the Sharpe ratio and reward-to-risk ratio were calculated

to further reflect the risk-adjusted performance of all asset classes.

In addition, the correlation between REITs and the other investment assets were calculated to

assess the possible diversification benefits that can be gained from including REITs in a

domestically mixed-asset portfolio. Efficient frontiers and asset allocation diagrams were

constructed to investigate the importance of REITs in an efficient portfolio. All datasets were

131

measured in local currency to avoid currency exchange fluctuations and to maintain

consistency.

To analyse the impact of the global financial crisis (GFC) on Asian REIT performance, the

timeframe was divided into three sub-periods: pre-GFC, GFC and post-GFC. While it is still

unclear about the exact dates of the start and end of the GFC, March 2008 was used as the

beginning of the GFC in this study, since it was the month Bear Stearns was taken over by

J.P. Morgan. March 2009 was chosen as the end for the GFC, because it was the month that

the Asian REIT index rebounded strongly from its trough since the start of the GFC. Risk-

return, correlation, efficient frontiers and asset allocation diagrams were determined for each

sub-period.

The remainder of this chapter is structured as follows. Section 5.2 presents the performance

analysis of Japan REITs. Section 5.3 provides the results from the performance analysis of

Singapore REITs. Section 5.4 exhibits the performance of Hong Kong REITs. Section 5.5

discusses the performance of Malaysia REITs, while section 5.6 shows that of Thailand

REITs. Section 5.7 and 5.8 demonstrate the performance analysis of Taiwan and Korean

REITs respectively. Finally, Section 5.9 provides the summary and strategic implications of

this chapter.

5.2 Performance of Japan REITs

5.2.1 Risk-adjusted Performance Analysis

Table 5.1 presents the risk-adjusted performance of J-REITs over October 2001-April 2012.

According to the results, J-REITs were superior to the other investment classes, generating an

average annual return of 5.6% as compared to that of 0.2% for shares and 3.6% for property

companies. The J-REIT risk level (26.31%), as measured by the standard deviation, was

higher than that of shares (21.45%), but less than that of property companies (30.54%). On a

risk-adjusted basis, measured by return/risk and Sharpe ratio, J-REITs (ranked #2)

significantly outperformed both property companies (#3) and shares (#4). The analysis has

revealed the superior risk-adjusted performance of J-REITs over shares and property

companies over the study period.

132

Table 5.1: J-REIT Risk-adjusted Performance Analysis: Oct 2001-Apr 2012

REITs Shares Property

Companies Bonds

Average Annual Return 5.57% 0.25% 3.62% 0.11%

Annual Risk 26.31% 21.45% 30.54% 0.08%

Risk/return Ratio 4.72 86.68 8.43 0.72

Sharpe Ratio 0.21 0.01 0.12 1.07

Rank #2 #4 #3 #1

Results on the correlation between J-REITs and other assets are shown in Table 5.2. It was

obvious that J-REITs could provide better diversification benefits to share holders than

property companies. The correlation between J-REITs and shares (r=0.70) was significant

and lower than that between property companies and shares (r=0.86). As a result, J-REITs are

seen to provide better diversification benefits to share holders than property companies. The

results also support that J-REITs are a distinct asset class to property companies. J-REITs

focus primarily in owning and managing income-producing properties, while property

companies often engage in property development and non-property activities. J-REITs

therefore provide investors pure exposure to the property market without taking on the

additional risks. The negative correlation between J-REITs and bonds (r=-0.09) also suggests

possible diversification potential between J-REITs and bonds.

Table 5.2: J-REIT Correlation Matrix: Oct 2001-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.70* 1.00

Property Companies 0.73* 0.86* 1.00 Bonds -0.09 -0.03 -0.03 1.00

* Significant at 5% level

5.2.2 Mixed-asset Portfolio Analysis

Table 5.3 and Figure 5.1 show the respective asset allocations at various portfolio risk levels.

It can be seen that J-REITs and bonds dominated the mixed-asset portfolio across the risk

spectrum. On the other hand, shares and property companies failed to enter the portfolio. The

allocation to J-REITs increases as the risk level rises, reflecting the strong risk-adjusted

performance of J-REITs. Furthermore, this highlights the important role of J-REITs in Japan

for investors seeking securitised property exposure in Japan. When REITs and property

133

companies were included in the share-bond portfolio, REITs consistently outdid property

companies in the portfolio at all levels.

Table 5.3: J-REIT Asset Allocations: Oct 2001-Apr 2012

Property Portfolio Portfolio

Shares Bonds REITs Companies Return Risk

0.00% 99.97% 0.03% 0.00% 0.11% 0.08%

0.00% 89.70% 10.30% 0.00% 0.67% 2.69%

0.00% 79.73% 20.27% 0.00% 1.22% 5.31%

0.00% 69.77% 30.23% 0.00% 1.76% 7.92%

0.00% 59.80% 40.20% 0.00% 2.31% 10.53%

0.00% 49.83% 50.17% 0.00% 2.85% 13.14%

0.00% 39.87% 60.13% 0.00% 3.40% 15.75%

0.00% 29.90% 70.10% 0.00% 3.94% 18.37%

0.00% 19.93% 80.07% 0.00% 4.48% 20.98%

0.00% 9.97% 90.03% 0.00% 5.03% 23.59%

0.00% 0.00% 100.00% 0.00% 5.57% 26.20%

Figure 5.1: J-REIT Asset Allocation Diagram: Oct 2001-Apr 2012

Figure 5.2 shows the efficient frontiers for the mixed-asset portfolios for Japan. The results

clearly indicated that the inclusion of J-REITs in the portfolio enhances returns and reduces

risk. When J-REITs are included in the bond-share and the bond-share-property company

portfolios, the efficient frontiers shifted upward suggesting a positive J-REIT effect. As a

result, the portfolios that include J-REITs outperformed the portfolios without J-REITs across

all risk levels.

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.1% 2.7% 5.3% 7.9% 10.5% 13.1% 15.8% 18.4% 21.0% 23.6% 26.2%

Allo

cati

on

Risk

Shares Bonds REITs Property Companies

134

Figure 5.2: J-REIT Efficient Frontiers: Oct 2001-Apr 2012

In order to avoid the disproportionally large exposure to a single asset, a maximum allocation

of 10% of the total portfolio value to REITs and property companies has been enforced.

Table 5.4 reports the optimal asset allocations with these constraints for Japan. These

constraints have resulted in more diversified portfolio allocations across all asset classes.

However, it is noticeable that the maximum constrained portfolio return (1.12%) was

substantial less than that of the unconstrained portfolio (5.57%) due to the constraints on J-

REITs.

The portfolio weight to shares increased with the increased level of risk, from 0.15% at the

minimum risk level to 80.00% at the aggressive risk level. On the other hand, bonds

gradually exited the portfolio, while risk increased. The allocation to REITs reached the 10%

maximum limit at around the 4.41% risk level, while the weight to property companies was

capped at around the 6.58% risk level. The asset allocation diagrams were depicted in Figure

5.3. It can be seen that bonds dominate the optimal portfolio at the low end of the risk

spectrum, while shares dominate at the high end. REITs and property companies were capped

at 10% level across the risk range. J-REITs were still in the efficient portfolio across the risk

profile; in most cases at the 10% maximum level.

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

Re

turn

Risk

Bonds & Shares Bond, Shares & REITs

Bonds, Shares & Property Companies Bonds, Shares, REITs & Property Companies

135

Table 5.4: J-REIT Asset Allocations with Constraints: Oct 2001-Apr 2012



0.00% 99.97% 0.03% 0.00% 0.11% 0.08%

0.00% 91.41% 8.59% 0.00% 0.58% 2.25%

0.00% 83.04% 10.00% 6.96% 0.90% 4.41%

6.96% 73.04% 10.00% 10.00% 1.02% 6.58%

17.87% 62.13% 10.00% 10.00% 1.03% 8.74%

28.46% 51.54% 10.00% 10.00% 1.05% 10.91%

38.88% 41.12% 10.00% 10.00% 1.06% 13.08%

49.23% 30.77% 10.00% 10.00% 1.08% 15.24%

59.52% 20.48% 10.00% 10.00% 1.09% 17.41%

69.77% 10.23% 10.00% 10.00% 1.10% 19.57%

80.00% 0.00% 10.00% 10.00% 1.12% 21.74%

Figure 5.3: J-REIT Asset Allocation Diagrams with Constraints: Oct 2001-Apr 2012

5.2.3 Sub-period Analysis

To assess the dynamic and time-changing nature of J-REIT performance and especially the

effects of the GFC on J-REIT performance, the full period of October 2001-April 2012 was

broken down to the three sub-periods of October 2001-February 2008 (pre-GFC), March

2008-March 2009 (GFC) and April 2009-April 2012 (post-GFC). The sub-period

performance analysis is displayed in Table 5.5. J-REITs consistently showed stronger risk-

adjusted performance (rank #2) than shares (#4) and property companies (#3) across all three

periods. Bonds (#1) were the best performing asset class, reflecting their defensive nature

during the periods of high volatility.

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.08% 2.25% 4.41% 6.58% 8.74% 10.91% 13.08% 15.24% 17.41% 19.57% 21.74%

Allo

cati

on

Risk


136

Before the GFC (panel A), J-REITs delivered the strongest average return (13.7%),

significantly above other sectors, while having marginally higher risk (16.58%) than shares

(16.48%), but significantly below that of property companies (25.34%). During the GFC, J-

REIT average return dropped significantly to -37.39%, while the risk level rose to 62.53%.

However, the post-GFC period saw substantial improvement in J-REIT risk-adjusted

performance. The J-REIT Sharpe ratio was 0.40, outperforming both shares (0.18) and

property companies (0.38).

Table 5.5: J-REIT Sub-period Performance Analysis Panel A: Pre-GFC: Oct 2001-Feb 2008


Companies Bonds


Annual Risk 16.58% 16.48% 25.34% 0.09%


Sharpe Ratio 0.82 0.38 0.44 1.12

Rank #2 #4 #3 #1

Panel B: GFC: Mar 2008-Mar 2009


Companies Bonds

Average Annual Return -37.39% -35.71% -42.68% 0.12%

Annual Risk 62.53% 41.49% 57.72% 0.05%

Risk/return Ratio -1.67 -1.16 -1.35 0.41

Sharpe Ratio -0.60 -0.86 -0.74 0.96

Rank #2 #4 #3 #1

Panel C: Post-GFC: Apr 2009-Apr 2012


Companies Bonds


Annual Risk 21.01% 19.94% 26.03% 0.05%


Sharpe Ratio 0.40 0.18 0.38 1.64

Rank #2 #4 #3 #1

Figure 5.4 illustrates the dynamic nature of the volatility over time for J-REITs, shares and

property companies using rolling 12-month volatility analysis. It was noticeable that the

volatility across all asset classes has increased dramatically during the GFC period, due to

greater instability in the market. From September 2002 to April 2007, the J-REITs were less

volatile than shares and property companies. However, from April 2007 to October 2009 the

J-REIT risk level has increased significantly, exceeding the risk level of shares and property

137

companies. Since October 2009, the J-REIT risk declined substantially below the risk of

property companies, but still relatively close to that of shares.

Figure 5.4: J-REIT 12-Month Rolling Volatility Analysis: Sep 2002-Apr 2012

The sub-period correlation analysis for J-REITs and other asset classes is presented in Table

5.6. Before the GFC, J-REITs had a low correlation with shares (r=0.33) and property

companies (r=0.46). During the GFC, REITs, shares and property companies have become

much more integrated. Correlation between REITs and shares increased to r=0.95, while that

between REITs and property companies rose to r=0.91. This suggests that the diversification

benefits are limited during the GFC due to the increased level of correlation. After the GFC,

correlation levels have been reduced, but still higher than the pre-GFC levels.

The correlation coefficients between REITs and shares were 0.33, 0.95 and 0.72 for the three

periods; pre-GFC, GFC and post-GFC respectively. These figures indicate that shares

investors can get significant diversification by including REITs in a share portfolio during the

periods before and after the GFC. Diversification benefits have been lost during the GFC, but

then regained to some degree after the GFC.

Between REITs and property companies, the correlation statistics were 0.46, 0.91 and 0.72

for the three periods respectively. These also suggest that it is possible to achieve inner-

property diversification if property company shareholders invest in REITs before and after

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

Sep

-20

02

Feb

-20

03

Jul-

20

03

De

c-2

00

3

May

-20

04

Oct

-20

04

Mar

-20

05

Au

g-2

00

5

Jan

-20

06

Jun

-20

06

No

v-2

00

6

Ap

r-2

00

7

Sep

-20

07

Feb

-20

08

Jul-

20

08

De

c-2

00

8

May

-20

09

Oct

-20

09

Mar

-20

10

Au

g-2

01

0

Jan

-20

11

Jun

-20

11

No

v-2

01

1

Ap

r-2

01

2

REITs Property Companies Shares

138

the GFC. On the other hand, there are relatively high correlations between property

companies and shares throughout all three periods (r=0.76, r=0.93 and r=0.90). These results

imply that diversification opportunities between property companies and shares are limited.

Share investors should therefore invest in REITs rather than property companies.

Table 5.6: J-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Oct 2001-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.33* 1.00




Companies Bonds

REITs 1.00 Shares 0.95* 1.00

Property Companies 0.91* 0.93* 1.00 Bonds -0.11 -0.02 0.00 1.00



Companies Bonds

REITs 1.00 Shares 0.72* 1.00

Property Companies 0.84* 0.90* 1.00 Bonds -0.01 0.08 0.01 1.00


Figure 5.5 shows the 12-month rolling correlation analysis for J-REITs. The correlation

between J-REITs and shares has fluctuated dramatically over the period from September

2002 to April 2012. Prior to the start of the GFC in February 2008, the correlation between

REITs and shares has been fluctuation around the -0.50/+0.50 range. This indicates a strong

diversification benefits from J-REITs to share investors. However, since February 2008, the

correlation started to rise rapidly to reach the highest level of 0.94 in October 2008. During

the GFC, J-REITs were highly correlated with shares, thus reducing their diversification

benefits. However, the correlation has been falling back to around 0.50-0.70 by the first

quarter of 2012, signalling a recovery of diversification benefit of the J-REIT market from

the GFC.

139

The rolling correlation between REITs and property companies was similar to those between

REITs and shares. This is due to the high association between property companies and shares.

As depicted in the chart, the correlation between property companies and shares were very

high, hovering above 0.60 over the entire sample period. On the other hand, the correlation

between REIT and bonds has always been low or negative, indicating substantial

diversification opportunities between the two asset classes.

Figure 5.5: J-REIT 12-Month Rolling Correlation Analysis: Sep 2002-Apr 2012

The sub-period asset allocations and efficient frontiers for the pre-GFC and post-GFC are

presented in Figure 5.6 and Figure 5.7 respectively. For the pre-GFC period, the mixed-asset

portfolio was dominated by REITs and bonds. Shares only had a minor role in the portfolio,

while property companies were not included. The efficient frontiers clearly shows that J-

REITs enhanced the portfolio performance significantly. The bond-share-REITs portfolio

outperformed both the bond-share and the bond-share-property company portfolios.

For the post-GFC period, the mixed-asset portfolio was dominated by REITs, bonds and

property companies, while shares were not included. This reflects the weak performance of

the Japanese share market after the GFC. The post-GFC efficient frontiers also show the

important role of J-REITs in the portfolio. The bond-share-REIT efficient frontier was above

the bond-share and the bond-share-property efficient frontier, indicating that the inclusion of

-1.00

-0.80

-0.60

-0.40

-0.20

-

0.20

0.40

0.60

0.80

1.00

Sep

-20

02

Feb

-20

03

Jul-

20

03

De

c-2

00

3

May

-20

04

Oct

-20

04

Mar

-20

05

Au

g-2

00

5

Jan

-20

06

Jun

-20

06

No

v-2

00

6

Ap

r-2

00

7

Sep

-20

07

Feb

-20

08

Jul-

20

08

De

c-2

00

8

May

-20

09

Oct

-20

09

Mar

-20

10

Au

g-2

01

0

Jan

-20

11

Jun

-20

11

No

v-2

01

1

Ap

r-2

01

2

REITs-Shares REITs-Bonds

REITs-Property Companies Shares-Property Companies

140

REITs has enhanced portfolio returns. In addition, J-REITs were better portfolio enhancers

than property companies.

Overall, the sub-period analysis highlights the benefit of J-REITs in a mixed-asset portfolio.

Importantly, the post-GFC period has seen J-REITs registering reduced risk and enhanced

diversification benefits with other asset classes. This is in contrast to the weaker performance

and poor diversification potential of property companies post-GFC. This reaffirms the

investment credentials of J-REITs compared to listed properties for investors seeking indirect

property exposure in Japan.

141

Figure 5.6: J-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Oct 2001-Feb 2008

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.1% 1.7% 3.4% 5.0% 6.6% 8.3% 9.9% 11.6% 13.2% 14.8% 16.5%

Allo

cati

on

Risk


0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



142

Figure 5.7: J-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.0% 2.6% 5.2% 7.7% 10.3% 12.9% 15.4% 18.0% 20.5% 23.1% 25.7%

Allo

cati

on

Risk


0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



143

5.3 Performance of Singapore REITs


Table 5.7 highlights the performance statistics for S-REITs, equities, property companies and

government bonds between the period of August 2003 and April 2012. During this period, S-

REITs registered the best historical mean return at 14.24% p.a., followed by property

companies at 13.43% p.a., equities at 11.54% and finally government bonds at 2.76% p.a. In

terms of volatility, property companies was the most risky asset, recording the highest

standard deviation of 28.28%, followed by S-REITs of 26.02%, equities of 22.90% and bonds

of 2.06%.

The Sharpe ratio was calculated using the average 3-month Treasury bill rate of 1.21% as an

estimate of the risk-free rate over the study period. Bonds generated the highest Sharpe ratio

0.75, followed by S-REITs of 0.50, shares of 0.45 and finally property companies of 0.43.

From the risk-adjusted perspective, S-REITs (#2) appear to provide superior risk-adjusted

returns than both equities (#3) and listed property companies (#4) over this period. This result

is consistent with findings in the previous section for the Japanese REIT market.

Table 5.7: S-REIT Risk-adjusted Performance Analysis: Aug 2003-Apr 2012


Companies Bonds


Annual Risk 26.02% 22.90% 28.28% 2.06%


Sharpe Ratio 0.50 0.45 0.43 0.75

Rank #2 #3 #4 #1

Table 5.8 highlights the correlation coefficients between the four asset classes in Singapore.

These statistics revealed that both property companies and REITs are positively and

significantly correlated with the general stock market. However, stock market returns are

more closely correlated with property companies (r=0.92) than with S-REITs (r=0.88). These

results imply that S-REITs as an asset class provides marginally better diversification benefits

to multi-asset portfolios as compared to those achieved by property companies in Singapore.

The returns of the S-REITs are also positively correlated to those of property companies with

a correlation coefficient of r=0.88, the same as the correlation between S-REITs and shares.

This is not a surprising result, as returns of property companies closely track the returns of

144

the share market. Finally, government bonds register negative or marginally positive

correlation with REITs (-0.01), shares (-0.01) and property companies (0.01).

Table 5.8: S-REIT Correlation Matrix: Aug 2003-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.88* 1.00

Property Companies 0.88* 0.93* 1.00 Bonds -0.01 -0.01 0.01 1.00



The results from the previous section have highlighted the potential of S-REITs offering

portfolio risk reduction and financial performance improvement. Table 5.9 and Figure 5.8

show the historical asset allocation of the efficient portfolios consisting of shares, bonds, S-

REITs and property companies. The optimal portfolio annualised returns were between

2.84% and 14.24%, while the portfolio standard deviations were between 2.04% and 25.89%.

Importantly, the efficient portfolios were dominated by bonds and S-REITs across all risk

levels. The allocation to shares was negligible, while no weight has been given to property

companies. These results underline the important and significant role of S-REITs in multi-

asset portfolios for Singapore.

Table 5.9: S-REIT Asset Allocations: Aug 2003-Apr 2012



0.91% 99.09% 0.00% 0.00% 2.84% 2.04%

0.07% 84.18% 15.75% 0.00% 4.57% 4.43%

0.00% 74.30% 25.70% 0.00% 5.71% 6.81%

0.00% 64.80% 35.20% 0.00% 6.80% 9.20%

0.00% 55.44% 44.56% 0.00% 7.87% 11.58%

0.00% 46.15% 53.85% 0.00% 8.94% 13.97%

0.00% 36.88% 63.12% 0.00% 10.00% 16.35%

0.00% 27.65% 72.35% 0.00% 11.06% 18.74%

0.00% 18.42% 81.58% 0.00% 12.12% 21.12%

0.00% 9.21% 90.79% 0.00% 13.18% 23.51%

0.00% 0.00% 100.00% 0.00% 14.24% 25.89%

145

Figure 5.8: S-REIT Asset Allocation Diagram: Aug 2003-Apr 2012

Figure 5.9 shows the efficient frontiers of various mixed-asset portfolios comprising of

bonds, shares, S-REITs and property companies. Graphical examination shows that the bond-

share-REIT and bond-share-REIT-property company portfolio outperformed the bond-share

and the bond-share-property company portfolio at all levels of the risk spectrum. This means

that the inclusion of S-REITs in mixed-asset portfolios helps improve performance, whilst

reducing the overall risk level.

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2.0% 4.4% 6.8% 9.2% 11.6% 14.0% 16.4% 18.7% 21.1% 23.5% 25.9%

Allo

cati

on

Risk


146

Figure 5.9: S-REIT Efficient Frontiers: Aug 2003-Apr 2012

Table 5.10 and Figure 5.10 report the efficient portfolio allocations subject to a 10%

maximum constraint on S-REITs and property companies. Consequently, the dominance of J-

REITs in the portfolio was reduced, while shares gained ground. It can be observed that the

increase in the volatility from 2.04% (minimum) to 23.19% (maximum) lead to the rise in the

equity allocation from 0.91% to 80.00% at the expense of bonds, falling from 99.09% to

0.00%. S-REITs and property shares were limited at the 10% threshold across the risk range.

S-REITs maintained a stable role in the portfolio remaining at the maximum 10% level across

most of the risk levels.

Table 5.10: S-REIT Asset Allocations with Constraints: Aug 2003-Apr 2012


Shares Bonds REITs Companies Return Risk 0.91% 99.09% 0.00% 0.00% 2.84% 2.04%

5.78% 84.22% 10.00% 0.00% 4.41% 4.16%

12.72% 74.34% 10.00% 2.94% 5.34% 6.27%

18.49% 65.20% 10.00% 6.31% 6.20% 8.39%

24.27% 56.24% 10.00% 9.50% 7.05% 10.50%

33.18% 46.82% 10.00% 10.00% 7.89% 12.62%

42.63% 37.37% 10.00% 10.00% 8.72% 14.73%

52.01% 27.99% 10.00% 10.00% 9.54% 16.85%

61.36% 18.64% 10.00% 10.00% 10.36% 18.96%

70.69% 9.31% 10.00% 10.00% 11.18% 21.08%

80.00% 0.00% 10.00% 10.00% 12.00% 23.19%

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



147

Figure 5.10: S-REIT Asset Allocation Diagrams with Constraints: Aug 2003-Apr 2012


Table 5.11 presents the risk-adjusted performance analysis for S-REITs for the three sub-

periods. The impact of the GFC to the financial market was clearly evident in Singapore.

S-REITs, shares and property companies went from delivering positive average returns of

28.20%, 20.38% and 32.11% pre-GFC respectively, to returns of -51.44%, -42.17% and

-55.74% during the GFC. The risk for S-REITs increased by 153% (19.45% to 49.23%),

while that for shares and property companies rose by 184% (14.63% to 41.62%) and 154%

(18.83% to 47.84%) respectively.

The period after the GFC saw significant improvement across all asset classes. Importantly,

S-REITs delivered the strongest revival both in terms of enhanced performance and reduced

risk. S-REITs registered the best average return (28.46%) in comparison to shares (24.56%)

and property companies (24.05%). On the other hand, S-REIT risk reduced to 18.46%, much

lower than the volatility of shares of 21.76% and property companies of 26.93%. This saw S-

REITs delivering the best risk-adjusted returns after the GFC.

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2.04% 4.16% 6.27% 8.39% 10.50% 12.62% 14.73% 16.85% 18.96% 21.08% 23.19%

Allo

cati

on

Risk


148

Table 5.11: S-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Aug 2003-Feb 2008


Companies Bonds


Annual Risk 19.45% 14.63% 18.83% 1.33%


Sharpe Ratio 1.35 1.26 1.60 0.90

Rank #2 #3 #1 #4



Companies Bonds


Annual Risk 49.23% 41.62% 47.84% 2.09%


Sharpe Ratio -1.06 -1.03 -1.18 0.95

Rank #3 #2 #4 #1



Companies Bonds


Annual Risk 18.46% 21.76% 26.93% 1.45%


Sharpe Ratio 1.52 1.11 0.88 1.32

Rank #1 #3 #4 #2

Figure 5.11 further highlights the dynamics of S-REIT volatility over time. S-REITs were

relatively more volatile than the equity market and property company stocks prior to and

during the GFC. However, after the GFC, S-REIT risk has been reduced significantly and

below share and property company volatility levels.

149

Figure 5.11: S-REIT 12-Month Rolling Volatility Analysis: Aug 2004-Apr 2012

Table 5.12 presents the inter-asset correlation analysis over these three sub-periods. During

the GFC, diversification benefits across multi-assets in Singapore have been impaired as asset

correlations intensified. The correlation coefficient between S-REITs and shares increased

from r=0.69 to r=0.98, while that between S-REITs and property companies jumped from

r=0.72 to r=0.96 during the GFC. This resulted in S-REITs losing their portfolio benefits with

shares and property companies. In contrast, the correlation between bonds and other listed

sectors declined significantly implying the increased role of fixed-interest investments in the

portfolios during the GFC. After the GFC, all asset classes have become moderately less

correlated, leading to the improved benefits of diversification. However, inter-asset

correlations are yet to revert to their pre-crisis levels.

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

Au

g-2

00

4

De

c-2

00

4

Ap

r-2

00

5

Au

g-2

00

5

De

c-2

00

5

Ap

r-2

00

6

Au

g-2

00

6

De

c-2

00

6

Ap

r-2

00

7

Au

g-2

00

7

De

c-2

00

7

Ap

r-2

00

8

Au

g-2

00

8

De

c-2

00

8

Ap

r-2

00

9

Au

g-2

00

9

De

c-2

00

9

Ap

r-2

01

0

Au

g-2

01

0

De

c-2

01

0

Ap

r-2

01

1

Au

g-2

01

1

De

c-2

01

1

Ap

r-2

01

2


150

Table 5.12: S-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Aug 2003-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.69* 1.00

Property Companies 0.72* 0.84* 1.00 Bonds 0.10 0.13 0.12 1.00



Companies Bonds

REITs 1.00 Shares 0.98* 1.00




Companies Bonds

REITs 1.00 Shares 0.90* 1.00



Figure 5.12 plots the rolling correlations between S-REIT returns and the returns of the other

asset classes. As seen, the correlation between REITs and shares spiked upward from around

r=-0.40 in 2004 to r=0.80 by mid-2006 to as high as r=0.90 level when the GFC hit the

market during 2008-2009. The REIT-share correlation has reduced slightly to hover at the

r=0.80 level in recent periods. Similar patterns were also observed for the relationship

between property companies and shares. Property companies however are more closely

related to the general stock market than S-REITs. These results further support the

proposition that investors receive greater diversification benefits by investing to S-REITs

than listed property companies. However, the diversification benefits of S-REITs have been

reduced recently as S-REITs have become more correlated to stocks.

151

Figure 5.12: S-REIT 12-Month Rolling Correlation Analysis: Aug 2004-Apr 2012

Figure 5.13 and Figure 5.14 present the asset allocation diagrams and efficient frontiers for

the pre-GFC and post-GFC respectively. The analysis shows S-REITs gaining a significant

position in the mixed-asset portfolio, especially after the GFC. Prior to the GFC, the efficient

portfolio consisted of S-REITs, bonds and property companies, with S-REITs representing

around 10% of the total value with the increasing weight at the high risk level. After the

GFC, the allocation to S-REITs increased substantially, as the efficient portfolio was

dominated by S-REITs and bonds, while shares and property companies were excluded. The

efficient frontiers for both periods were seen moving upward with the inclusion of S-REITs.

The portfolio performance was improved significantly with the inclusion of S-REITs.

Overall, the sub-period analysis has shown a robust recovery of S-REITs over the post-GFC

period, outperforming the other assets of shares, property companies and bonds. The GFC has

made S-REITs more correlated, resulting in a loss of diversification benefits. Even though,

the correlation levels have not been reduced to the pre-GFC levels, S-REITs still represented

a significant position in the asset mix post-GFC due to their strong risk-adjusted returns. The

mixed-asset portfolio post GFC was dominated by S-REITs and bonds.

-1.00

-0.80

-0.60

-0.40

-0.20

-

0.20

0.40

0.60

0.80

1.00

Au

g-2

00

4

De

c-2

00

4

Ap

r-2

00

5

Au

g-2

00

5

De

c-2

00

5

Ap

r-2

00

6

Au

g-2

00

6

De

c-2

00

6

Ap

r-2

00

7

Au

g-2

00

7

De

c-2

00

7

Ap

r-2

00

8

Au

g-2

00

8

De

c-2

00

8

Ap

r-2

00

9

Au

g-2

00

9

De

c-2

00

9

Ap

r-2

01

0

Au

g-2

01

0

De

c-2

01

0

Ap

r-2

01

1

Au

g-2

01

1

De

c-2

01

1

Ap

r-2

01

2



152

Figure 5.13: S-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Aug 2003-Feb 2008

BondsREITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.3% 3.1% 4.8% 6.5% 8.3% 10.0% 11.7% 13.5% 15.2% 16.9% 18.7%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0%

Re

turn

Risk



153

Figure 5.14: S-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.4% 3.1% 4.8% 6.5% 8.1% 9.8% 11.5% 13.2% 14.9% 16.5% 18.2%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0%

Re

turn

Risk



154

5.4 Performance of Hong Kong REITs


Table 5.13 exhibits the descriptive statistics of the historical returns of various investment

options in Hong Kong over August 2004-April 2012. On average, HK-REIT total returns

appreciated by 13.60% p.a., outperforming both the general stock market (11.69%) and the

listed property companies (11.21%). In addition, the risk level of HK-REITs was also lower

than that of shares (26.95%) and property companies (28.25%). On a risk-adjusted basis, as

measured by the risk/return ratio and Sharpe ratio, HK-REITs were the best performing asset

class (#1) beating bonds (#2), shares (#3) and property companies (#4). This reflects Hong

Kong investors’ strong appetite for HK-REITs in a volatile investment environment, due to

their stable yields.

Table 5.13: HK-REIT Risk-adjusted Performance Analysis: Aug 2004-Apr 2012


Companies Bonds


Annual Risk 22.39% 26.95% 28.25% 3.64%


Sharpe Ratio 0.55 0.38 0.35 0.47

Rank #1 #3 #4 #2

The HK-REIT correlation analysis over August 2004-April 2012 was presented in Table

5.14. HK-REITs were observed to be less correlated with the equity market (r=0.66) than

property companies (r=0.95), indicating greater portfolio diversification benefits by HK-

REITs than listed property companies. HK-REITs and property companies exhibit relatively

low co-movement (r=0.65), suggesting some degree of within-property diversification

benefits. This also indicates that HK-REITs and property companies are subject to different

market factors. HK-REITs also saw diversification potential with bonds (r=-0.10), which was

slightly better than that seen by shares (r=0.10) and property companies (r=0.09).

Table 5.14: HK-REIT Correlation Matrix: Aug 2004-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.66* 1.00



155


The unconstraint portfolios obtained by the standardised Markowitz model for different

levels of risk are displayed in Table 5.15 and Figure 5.15. Different combinations of assets

were simulated to achieve the maximum level of return for any given level of risk. It is

readily observed that the optimal portfolio with the minimum risk (3.51%) consisted of

95.95% in bonds and 4.05% in HK-REITs. However, as the risk and expected returns of the

portfolio increased, the proportion invested in bonds reduces, while the allocation to HK-

REITs increased to as high as 100% at the maximum risk level. Allocation to equities was

limited, at a maximum at 2.3%, while property companies did not contribute to the optimum

portfolios. The large proportion allocated to HK-REITs reflects their importance in the

portfolio. It is consistent with the evidence that HK-REITs, having the highest Sharpe ratio,

outperformed bonds, shares and property companies, historically. The low correlations

between HK-REITs and the other assets also played an important role in reducing portfolio

risk.

Table 5.15: HK-REIT Asset Allocations: Aug 2004-Apr 2012



0.00% 95.95% 4.05% 0.00% 3.51% 3.50%

0.00% 78.14% 21.86% 0.00% 5.39% 5.38%

0.00% 68.23% 31.77% 0.00% 6.43% 7.26%

0.00% 59.15% 40.85% 0.00% 7.38% 9.13%

0.00% 50.41% 49.59% 0.00% 8.30% 11.01%

0.08% 41.83% 58.09% 0.00% 9.20% 12.89%

0.65% 33.26% 66.09% 0.00% 10.09% 14.76%

1.25% 24.74% 74.00% 0.00% 10.98% 16.64%

1.70% 16.30% 82.00% 0.00% 11.86% 18.51%

2.30% 7.86% 89.84% 0.00% 12.73% 20.39%

0.00% 0.00% 100.00% 0.00% 13.60% 22.27%

156

Figure 5.15: HK-REIT Asset Allocation Diagram: Aug 2004-Apr 2012

Figure 5.16 illustrates the efficient frontiers from the multi-asset portfolio sets. It is clear

from the efficient frontier graph that the returns of a portfolio that includes HK-REITs are

always greater than those without HK-REITs for every level of risk. The return enhancement

is more significant as the expected return and risk levels become higher. This observation is

consistent with the results found in the other developed Asian REIT markets of Japan and

Singapore.

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

3.5% 5.4% 7.3% 9.1% 11.0% 12.9% 14.8% 16.6% 18.5% 20.4% 22.3%

Allo

cati

on

Risk


157

Figure 5.16: HK-REIT Efficient Frontiers: Aug 2004-Apr 2012

The optimal weights in the efficient portfolios with constraints on HK-REITs and property

companies were computed and reported in Table 5.16. The portfolio returns ranged from

3.51% to 11.88%, while the portfolio risk varied from 11.88% to 25.64%. The limitation on

HK-REITs in the portfolio has allowed shares to enter the mix at an increasing rate as the risk

increased. At the minimum risk level, the efficient portfolio consisted of 95.95% bonds and

4.05% REITs. However, when the risk increased, the exposure to shares gradually took place

at the expense of bonds and HK-REITs. At the maximum risk level of 25.64%, the portfolio

was dominated by shares (90.00%), while HK-REITs reached their 10% ceiling. On the other

hand, no allocations were made to property companies, reflecting their lack of performance in

a mixed-asset portfolio. In contrast, HK-REITs were allocated at the maximum level of 10%

for most of the risk spectrum.

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



158

Table 5.16: HK-REIT Asset Allocations with Constraints: Aug 2004-Apr 2012



11.73% 78.27% 10.00% 0.00% 5.15% 5.72%

21.45% 68.55% 10.00% 0.00% 5.98% 7.93%

30.46% 59.54% 10.00% 0.00% 6.76% 10.14%

39.20% 50.80% 10.00% 0.00% 7.51% 12.36%

47.80% 42.20% 10.00% 0.00% 8.25% 14.57%

56.31% 33.69% 10.00% 0.00% 8.98% 16.79%

64.78% 25.22% 10.00% 0.00% 9.71% 19.00%

73.21% 16.79% 10.00% 0.00% 10.44% 21.21%

81.61% 8.39% 10.00% 0.00% 11.16% 23.43%

90.00% 0.00% 10.00% 0.00% 11.88% 25.64%

Figure 5.17: HK-REIT Asset Allocation Diagrams with Constraints: Aug 2004-Apr 2012


The performance of HK-REITs was observed over three sub periods; pre-GFC, GFC and

post-GFC. As seen in Table 5.17, the property and stock market in Hong Kong were badly

affected during the GFC period. All sectors, except bonds, registered negative annualised

returns. However, HK-REITs (-18.66%) were less affected than shares (-42.07%) and

property companies (-42.19%). HK-REIT risk (46.46%) during the GFC was marginally

higher than property companies (45.87%), but lower than equities (47.55%). Consequently,

HK-REITs (ranked #2) outperformed both shares (#3) and property companies (#4). The

post-GFC saw the robust recovery of the REIT sector in Hong Kong. HK-REITs delivered

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

3.50% 5.72% 7.93% 10.14% 12.36% 14.57% 16.79% 19.00% 21.21% 23.43% 25.64%

Allo

cati

on

Risk


159

the highest return (30.21%) and the lowest level of risk (14.24%), except bonds. As a result,

HK-REITs registered the best performance (#1) in the post-GFC period.

Table 5.17: HK-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Aug 2004-Feb 2008


Companies Bonds


Annual Risk 16.26% 17.34% 17.91% 1.76%


Sharpe Ratio 0.54 1.44 1.35 0.75

Rank #4 #1 #2 #3



Companies Bonds


Annual Risk 46.46% 47.55% 45.87% 2.66%


Sharpe Ratio -0.41 -0.90 -0.93 0.74

Rank #2 #3 #4 #1



Companies Bonds


Annual Risk 14.24% 24.57% 28.60% 1.88%


Sharpe Ratio 2.11 0.79 0.67 1.09

Rank #1 #3 #4 #2

The correlations between HK-REITs and the other investment options over the three sub-

periods were assessed. The results in Table 5.18 indicate that the GFC has significantly

impacted on the correlations between HK-REITs and other assets, in particular shares and

property companies. These results confirm that the GFC has caused a significant increase in

cross-asset correlations, resulting in the loss of diversification benefits. The correlation

between HK-REITs and shares spiked from r=0.22 to r=0.90, while that between HK-REITs

and property companies jumped from r=0.36 to r=0.86. For the post-GFC period, the degree

of correlation has declined substantially. The correlation coefficient of HK-REIT with shares

and property companies fell to r=0.61 and r=0.60 respectively. This has helped HK-REITs

regaining some degree of diversification benefits. However, the level of correlation was still

much higher than the pre-GFC levels.

160

Table 5.18: HK-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Aug 2004-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.22* 1.00




Companies Bonds

REITs 1.00 Shares 0.90* 1.00




Companies Bonds

REITs 1.00 Shares 0.61* 1.00



The asset allocations and efficient frontiers for the pre- and post-GFC period were computed

and presented in Figure 5.18 and Figure 5.19 respectively. For the pre-GFC period, the results

from the optimisation model show on average about 6.79% of the optimal portfolio should be

allocated to HK-REITs. After the GFC, HK-REITs gained a much more significant role in the

optimum portfolio, reflecting their strong risk-adjusted performance and improved

diversification benefits. The post-GFC efficient frontiers further demonstrated the benefits of

HK-REITs in mixed-asset portfolios, as portfolios with HK-REITs outperformed those

without. In contrast, shares and property companies failed to enhance portfolio performance.

161

Figure 5.18: HK-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-

GFC: Aug 2004-Feb 2008

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.7% 3.2% 4.8% 6.3% 7.9% 9.4% 10.9% 12.5% 14.0% 15.6% 17.1%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0%

Re

turn

Risk



162

Figure 5.19: HK-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-

GFC: Apr 2009-Apr 2012

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.9% 3.1% 4.3% 5.5% 6.7% 8.0% 9.2% 10.4% 11.6% 12.8% 14.0%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



163

5.5 Performance of Malaysia REITs


The risk-adjusted performance analysis for M-REITs over September 2005-April 2012 is

presented in Table 5.19. M-REITs delivered an annualised average return of 9.00%,

underperforming shares (13.03% p.a.) and listed property stocks (12.42% p.a.) over this

period. On the other hand, M-REITs (11.94%) were considerably less volatile than shares

(16.29%) and listed property companies (21.68%). This saw M-REITs only having 73% of

the level of equity market risk and only having 55% of the property company risk. Not

surprisingly, bonds offered the lowest risk (1.29%) along with the lowest historical returns

(4.00%). This is somehow consistent with the convention that low levels of uncertainty are

associated with low potential returns and vice versa. On a risk-adjusted perspective, as

measured by the Sharpe ratio, M-REITs ranked #3, exceeded by bonds (#1) and shares (#2),

but outperformed listed property securities (#4).

Table 5.19: M-REIT Risk-adjusted Performance Analysis: Sep 2005-Apr 2012


Companies Bonds


Annual Risk 11.94% 16.29% 21.68% 1.29%


Sharpe Ratio 0.51 0.62 0.44 0.82

Rank #3 #2 #4 #1

The inter-asset correlation matrix over the seven year period is detailed in Table 5.20. The

returns of M-REITs were lowly correlated with the overall stock market (r=0.65) and listed

property companies (r=0.59), while being negatively correlated with bonds (r=-0.27). These

relatively low associations suggest that substantial diversification benefits might be achieved

if M-REITs were combined in an investment portfolio. Shares and property companies, on

the other hand, show a high level of dependence (r=0.84) resulting in limited diversification

potential. Overall, M-REITs are seen to provide better portfolio enhancement than property

companies.

164

Table 5.20: M-REIT Correlation Matrix: Sep 2005-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.65* 1.00

Property Companies 0.59* 0.84* 1.00 Bonds -0.27* -0.22* -0.20 1.00



Table 5.21 and Figure 5.20 show the respective asset allocations at various portfolio risk

levels from 1.19% to 16.19% (corresponding to the average returns from 4.23% to 13.03%).

The four-asset efficient portfolio is dominated by bonds and shares. M-REITs, however, are

also included in the mixed-asset portfolio. The optimum portfolios see an average allocation

to M-REITs of around 11.11% across the risk spectrum. No allocations were made to

property companies as a consequence of their poor risk-adjusted performance and high

correlation with shares.

Table 5.21: M-REIT Asset Allocations: Sep 2005-Apr 2012



13.81% 80.11% 6.08% 0.00% 5.55% 2.69%

22.48% 69.45% 8.07% 0.00% 6.43% 4.19%

30.81% 59.16% 10.03% 0.00% 7.28% 5.69%

39.18% 49.11% 11.71% 0.00% 8.12% 7.19%

47.51% 39.14% 13.35% 0.00% 8.96% 8.69%

55.66% 29.07% 15.27% 0.00% 9.79% 10.19%

63.96% 19.17% 16.87% 0.00% 10.62% 11.69%

72.02% 9.09% 18.89% 0.00% 11.45% 13.19%

81.27% 0.00% 18.73% 0.00% 12.28% 14.69%

100.00% 0.00% 0.00% 0.00% 13.03% 16.19%

165

Figure 5.20: M-REIT Asset Allocation Diagram: Sep 2005-Apr 2012

The efficient frontiers are shown in Figure 5.21. The bond-share efficient frontier line

superimposed on the bond-share-REITs and the bond-share-property companies efficient

frontier lines. This is because shares and bonds dominated the three-asset portfolios. As

observed, the four-asset efficient frontier was below the other efficient frontier lines. This

means that M-REITs and property companies failed to improve the portfolio returns, further

verifying the dominant role of shares and bonds in Malaysian domestic portfolios.

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.2% 2.7% 4.2% 5.7% 7.2% 8.7% 10.2% 11.7% 13.2% 14.7% 16.2%

Allo

cati

on

Risk


166

Figure 5.21: M-REIT Efficient Frontiers: Sep 2005-Apr 2012

The 10% maximum allocation constraint on M-REITs and property companies was imposed

on the mixed-asset portfolio. The results were reported in Table 5.22 and Figure 5.22. The

figures are not significantly different from the unconstrained optimisation. This is because M-

REITs did not represent a significant portion in the unconstraint portfolio. M-REITs reach the

10% maximum allocation at the medium-to-low risk level (around 7.19%). Similar to the

unconstrained case, the portfolio was dominated by bonds at low risk and completely taken

over by shares at the maximum risk.

Table 5.22: M-REIT Asset Allocations with Constraints: Sep 2005-Apr 2012



13.80% 80.10% 6.10% 0.00% 5.55% 2.69%

22.50% 69.47% 8.02% 0.00% 6.43% 4.19%

30.99% 59.30% 9.71% 0.00% 7.28% 5.69%

40.11% 49.89% 10.00% 0.00% 8.12% 7.19%

49.33% 40.67% 10.00% 0.00% 8.96% 8.69%

58.50% 31.50% 10.00% 0.00% 9.78% 10.19%

67.65% 22.35% 10.00% 0.00% 10.61% 11.69%

76.78% 13.22% 10.00% 0.00% 11.44% 13.19%

85.91% 4.09% 10.00% 0.00% 12.26% 14.69%

100.00% 0.00% 0.00% 0.00% 13.03% 16.19%

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0%

Re

turn

Risk



167

Figure 5.22: M-REIT Asset Allocation Diagrams with Constraints: Sep 2005-Apr 2012


As the period under study involves the GFC, it is important to conduct the sub-period

analysis to assess the investment dynamics and structural breaks of M-REITs under different

market conditions. The three sub-periods of September 2005-February 2008 (pre-GFC),

March 2008-March 2009 (GFC) and April 2009-April 2012 (post-GFC) were assessed.

Table 5.23 presents the risk-adjusted performance of M-REITs over these three sub-periods.

Notably, M-REITs have shown more robustness than both equities and property companies

which were significantly impacted during the GFC. Apart from bonds, the M-REIT sector has

been the least affected during the GFC, with the average return falling to only -15.00% versus

the equivalent figure of -33.31% for shares and -37.48% for property companies. M-REITs

also experienced less volatility than the other two listed sectors. M-REIT risk (13.02%)

during the GFC was only half the level of stock market risk (24.39%) and property security

risk (26.14%).

The risk-adjusted analysis saw M-REITs gradually improve their performance over these

three periods (#4 to #2 to #1), taking the top position post-GFC against shares (#2), bonds

(#3) and property stocks (#4). Importantly, M-REITs were the only asset class, other than

bonds, that become less risky post-GFC than it has been pre-GFC. On the contrary, the risk-

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.19% 2.69% 4.19% 5.69% 7.19% 8.69% 10.19% 11.69% 13.19% 14.69% 16.19%

Allo

cati

on

Risk


168

adjusted performance of property companies has gradually worsened over time, moving from

second-best asset class pre-GFC to the worst position (#4) during and after the GFC.

Table 5.23: M-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Sep 2005-Feb 2008


Companies Bonds


Annual Risk 13.03% 11.29% 21.16% 1.40%


Sharpe Ratio 0.09 1.88 0.97 0.53

Rank #4 #1 #2 #3



Companies Bonds


Annual Risk 13.02% 24.39% 26.14% 1.99%


Sharpe Ratio -1.39 -1.50 -1.55 0.46

Rank #2 #3 #4 #1



Companies Bonds


Annual Risk 9.22% 13.46% 17.82% 0.84%


Sharpe Ratio 2.25 1.71 1.40 1.66

Rank #1 #2 #4 #3

The dynamic volatility of M-REITs, shares and property companies were shown in Figure

5.23 below. Importantly, this chart depicts M-REITs having lower volatility than shares and

property companies over most of the period under investigation. The M-REIT risk level was

seen to be less affected than shares and property companies both during the GFC and the

ongoing European debt crisis, which caused significant rises in market volatility. The analysis

further shows that M-REITs were the only sector that had a lower risk level post-GFC than

pre-GFC. This might have a strong implication for investors seeking a low-risk investment

option post-GFC in Malaysia.

169

Figure 5.23: M-REIT 12-Month Rolling Volatility Analysis: Sep 2006-Apr 2012

The portfolio diversification benefits of M-REITs over these three sub-periods are shown in

Table 2.24. From this table, it can be seen that the GFC has significantly reduced the

diversification benefits of M-REITs with shares (r=0.44 rising to r=0.80) and property

companies (r=0.42 increasing to r=0.72). The post-GFC period has seen these diversification

benefits improving, as the correlation coefficient between REITs and shares dropped to

r=0.74 and that between REITs and property companies felt to r=0.64. In contrast, the

correlation between property companies and shares has been progressively increasing (from

r=0.78 to r=0.81 to r=0.91) over these three sub-periods, hence reducing the benefits of

holding property companies in a share portfolio. M-REITs have also shown consistent

diversification benefits with bonds over these three sub-periods.

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

Sep

-20

06

De

c-2

00

6

Mar

-20

07

Jun

-20

07

Sep

-20

07

De

c-2

00

7

Mar

-20

08

Jun

-20

08

Sep

-20

08

De

c-2

00

8

Mar

-20

09

Jun

-20

09

Sep

-20

09

De

c-2

00

9

Mar

-20

10

Jun

-20

10

Sep

-20

10

De

c-2

01

0

Mar

-20

11

Jun

-20

11

Sep

-20

11

De

c-2

01

1

Mar

-20

12


170

Table 5.24: M-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Sep 2005-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.44* 1.00




Companies Bonds

REITs 1.00 Shares 0.80* 1.00




Companies Bonds

REITs 1.00 Shares 0.74* 1.00



Figure 5.24 illustrates the changing dynamics of the correlations between M-REITs and the

other financial assets. The impact of the GFC on the relationship of M-REITs and the general

stock market is clearly seen in the chart. M-REITs have become more co-integrated with

shares during the GFC, limiting their diversification benefits. The correlation has since been

reduced gradually after the GFC, but is yet to get back to the pre-GFC level. Property

companies on the other hand, maintained a high degree of association with shares across the

whole period. This suggests limited diversification potential for investing in listed property

companies.

171

Figure 5.24: M-REIT 12-Month Rolling Correlation Analysis: Sep 2006-Apr 2012

The analysis of M-REITs in mixed-asset portfolios was undertaken for the three sub-periods.

Before the GFC, M-REITs only had limited exposure in the mixed-asset portfolio at the

lower end of the risk spectrum (Figure 5.25). The average allocation to M-REITs was only

0.28% before the GFC, but increased to 38.60% post-GFC (Figure 5.26). In particular, the

allocation to M-REITs increased as the risk level increased from the low to the medium level,

but then started to reduce as the risk increased to the higher levels. This reflects the low

volatility level and strong risk-adjusted performance of M-REITs post-GFC. The efficient

frontiers in Figure 5.26 further confirmed that the inclusion of M-REITs enhances the

portfolio return for any given level of risk.

-1.00 -0.80 -0.60 -0.40 -0.20

-0.20 0.40 0.60 0.80 1.00

Sep

-20

06

De

c-2

00

6

Mar

-20

07

Jun

-20

07

Sep

-20

07

De

c-2

00

7

Mar

-20

08

Jun

-20

08

Sep

-20

08

De

c-2

00

8

Mar

-20

09

Jun

-20

09

Sep

-20

09

De

c-2

00

9

Mar

-20

10

Jun

-20

10

Sep

-20

10

De

c-2

01

0

Mar

-20

11

Jun

-20

11

Sep

-20

11

De

c-2

01

1

Mar

-20

12



172

Figure 5.25: M-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Sep 2005-Feb 2008

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.2% 2.2% 3.2% 4.2% 5.2% 6.2% 7.1% 8.1% 9.1% 10.1% 11.1%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0%

Re

turn

Risk



173

Figure 5.26: M-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.8% 2.5% 4.2% 5.9% 7.5% 9.2% 10.9% 12.5% 14.2% 15.9% 17.6%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0%

Re

turn

Risk



174

5.6 Performance of Thailand REITs


Table 5.25 exhibits the performance analysis of Thai-REITs, bonds, shares and property

companies over the period of December 2003-April 2012. The analysis reveals that Thai-

REITs behaved distinctively different to shares and property companies. In particular, Thai-

REITs provided the lowest annualised return of 3.50% p.a., underperforming shares (13.55%)

and property companies (5.07%). On the other hand, Thai-REITs recorded a significantly

lower level of risk (10.44%) than property companies (36.41%) and shares (26.74%). On a

risk-adjusted basis, Thai-REITs ranked 3rd

, above only property companies, with a risk-to-

return and Sharpe ratio of 2.98 and 0.07 respectively.

Table 5.25: Thai-REIT Risk-adjusted Performance Analysis: Dec 2003-Apr 2012


Companies Bonds


Annual Risk 10.44% 26.74% 36.41% 2.66%


Sharpe Ratio 0.07 0.40 0.06 0.60

Rank #3 #2 #4 #1

The correlation matrix for REITs, bonds, shares and property company returns over the entire

sample is presented in Table 5.26. The data shows all correlations being statistically

significant at the 5% level of significance. The strongest correlation was observed between

shares and property companies (r=0.86), indicating that investors receive little diversification

potential by diversifying from shares to property stocks during this period. This saw Thai-

REIT correlations with shares (r=0.68) and property stocks (r=0.70). This suggests that Thai-

REITs could be good diversifiers for both share and listed property portfolios. In contrast, the

relatively high correlation between property companies and shares suggested limited

diversification benefits between the two asset classes. Bonds were negatively correlated to

shares (r=-0.34), REITs (r=-0.21) and property stocks (r=-0.20), reflecting their defensive

nature.

175

Figure 5.27 depicts the 12-month rolling volatility analysis for the different asset classes in

Thailand over December 2003-April 2012. The risk levels have increased considerably for all

asset classes since mid-2008 through to the end of 2009. This was due to the impact of the

GFC which took place during 2007-2008. It is important to note that the volatility of Thai-

REITs was substantially lower than those of shares and property stocks over the entire sample

period. This implies that Thai-REITs are lower risk investment options for investors.

The rolling 12-month correlations were assessed over December 2003-April 2012 and

illustrated in Figure 5.28. Over the entire period, it is noticeable that the correlation between

Thai-REITs and shares was lower than those between property companies and shares, except

for a brief period by the end of 2005. This means that Thai-REITs are not as highly correlated

to the stock market as property companies. For this reason, they are useful for the

diversification of portfolios. The correlation between Thai-REITs and shares has only risen

Table 5.26: Thai-REIT Correlation Matrix: Dec 2003-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.68* 1.00



Figure 5.27: Thai-REIT 12-Month Rolling Volatility Analysis: Dec 2004-Apr 2012

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

De

c-2

00

4

Ap

r-2

00

5

Au

g-2

00

5

De

c-2

00

5

Ap

r-2

00

6

Au

g-2

00

6

De

c-2

00

6

Ap

r-2

00

7

Au

g-2

00

7

De

c-2

00

7

Ap

r-2

00

8

Au

g-2

00

8

De

c-2

00

8

Ap

r-2

00

9

Au

g-2

00

9

De

c-2

00

9

Ap

r-2

01

0

Au

g-2

01

0

De

c-2

01

0

Ap

r-2

01

1

Au

g-2

01

1

De

c-2

01

1

Ap

r-2

01

2


176

for a short period during 2009, but has since decreased significantly signifying the improving

diversification benefits of Thai-REITs in the recent period. On the other hand, the correlation

between property companies and shares remained at a high level. Correlations of Thai-REITs

and bonds mostly stayed in negative territory.


Table 5.27 and Figure 5.29 show the allocations of mixed-asset portfolios consisting of Thai-

REITs, bonds, shares and property companies over December 2003-April 2012. During this

period, shares and bonds were dominant across the risk spectrum in the portfolio. The

proportion of property companies in the portfolio increased as the risk level increased,

indicating their high volatility. Thai-REITs only marginally entered the portfolio at the low

risk level, suggesting them as a low-risk asset class. Property companies did not appear in the

portfolio due their poor performance, high risk and high correlation with shares. The efficient

frontiers (Figure 5.29) also show that Thai-REITs and property companies failed to enhance

the share and bond portfolio.

Figure 5.28: Thai-REIT 12-Month Rolling Correlation Analysis: Dec 2004-Apr 2012

-1.00

-0.80

-0.60

-0.40

-0.20

-

0.20

0.40

0.60

0.80

1.00

De

c-2

00

4

Ap

r-2

00

5

Au

g-2

00

5

De

c-2

00

5

Ap

r-2

00

6

Au

g-2

00

6

De

c-2

00

6

Ap

r-2

00

7

Au

g-2

00

7

De

c-2

00

7

Ap

r-2

00

8

Au

g-2

00

8

De

c-2

00

8

Ap

r-2

00

9

Au

g-2

00

9

De

c-2

00

9

Ap

r-2

01

0

Au

g-2

01

0

De

c-2

01

0

Ap

r-2

01

1

Au

g-2

01

1

De

c-2

01

1

Ap

r-2

01

2



177

Table 5.27: Thai-REIT Mixed-Asset Portfolio Allocations: Dec 2003-Apr 2012



0.00% 87.70% 12.30% 0.00% 4.31% 2.23%

17.28% 82.72% 0.00% 0.00% 6.00% 4.67%

27.24% 72.76% 0.00% 0.00% 6.91% 7.10%

36.63% 63.37% 0.00% 0.00% 7.76% 9.54%

45.83% 54.17% 0.00% 0.00% 8.60% 11.98%

54.93% 45.07% 0.00% 0.00% 9.43% 14.42%

63.99% 36.01% 0.00% 0.00% 10.26% 16.86%

73.02% 26.98% 0.00% 0.00% 11.09% 19.30%

82.03% 17.97% 0.00% 0.00% 11.91% 21.73%

91.02% 8.98% 0.00% 0.00% 12.73% 24.17%

100.00% 0.00% 0.00% 0.00% 13.55% 26.61%

Figure 5.29: Thai-REIT Asset Allocation Diagram: Dec 2003-Apr 2012

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2.2% 4.7% 7.1% 9.5% 12.0% 14.4% 16.9% 19.3% 21.7% 24.2% 26.6%

Allo

cati

on

Risk


178


Table 5.28 demonstrates the impact of the GFC on Thai-REITs, bonds, shares and property

companies over the three sub-periods: pre-GFC (December 2003-February 2008), GFC

(March 2008-March 2009) and post-GFC (March 2009-April 2012). The effects of the GFC

was clearly observed across all asset classes. Annualized average returns on Thai-REITs went

down from -2.86% to -15.07%. Share and property stock returns were also severely affected

by the crisis, dropping from 13.28% to -43.85% and -3.68% to -56.61% respectively. The

GFC also lead to the substantially increased level of volatility across all asset classes. Risk

levels of Thai-REITs rose by 148% (from 8.10% to 20.11%), which is slightly higher than

that of shares (133% increase) and property companies (136% increase).

After the GFC, the performance of all asset classes has improved significantly, as revealed in

panel C of Table 5.28. Thai-REIT average returns increased to 20.87% p.a., while shares and

property company returns increased to 44.52% p.a. and 58.60% respectively. Importantly, the

annual risk of Thai-REITs dropped to 6.29%, substantially lower than that of shares (20.63%)

and property companies (31.23%). This has helped to improve the risk-adjusted performance

(measured by reward-to-risk and Sharpe ratios) of Thai-REITs, which was the best

performing asset class, as compared to the other investment options. This highlights the

robustness of Thai-REITs during the post-GFC period, as investors become more confident

Figure 5.30: Thai-REIT Efficient Frontiers: Dec 2003-Apr 2012

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



179

with positive regulatory changes from the government. The post-GFC period saw Thai-REITs

ranked #1 in risk-adjusted performance.

The sub-period correlation analysis of Thai-REITs and the other investment vehicles is

presented in Table 5.29. Before the GFC, Thai-REITs have low correlations with both shares

(r=0.47) and property companies (r=0.48), suggesting diversification benefits of Thai-REITs

for shares and property investors. During the GFC, the diversification benefits for Thai-

REITs has been reduced significantly, as correlations of Thai-REITs with shares and property

companies have increased to r=0.93 and r=0.91 respectively. After the GFC, Thai-REITs

have gained back diversification potential, with correlations with shares and property

companies falling back to low levels of r=0.35 and r=0.47 respectively. Bonds have

maintained mostly low or negative correlations with other asset classes across all three

periods, showing them as a defensive investment.

Table 5.28: Thai-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Dec 2003-Feb 2008


Companies Bonds

Average Annual Return -2.86% 13.28% -3.68% 4.94%

Annual Risk 8.10% 20.66% 26.74% 1.84%

Risk/return Ratio -2.83 1.56 -7.27 0.37

Sharpe Ratio -0.80 0.47 -0.27 0.74

Rank #4 #2 #3 #1



Companies Bonds


Annual Risk 20.11% 48.14% 63.22% 2.99%


Sharpe Ratio -0.89 -0.97 -0.94 0.51

Rank #2 #4 #3 #1



Companies Bonds


Annual Risk 6.29% 20.63% 31.23% 1.17%


Sharpe Ratio 3.00 2.06 1.81 1.46

Rank #1 #2 #3 #4

180

Figure 5.31 and Figure 5.32 exhibit the asset allocation diagrams and efficient frontiers for

the mixed-asset portfolios over the pre- and post-GFC periods. In the pre-GFC period, Thai-

REITs only played a minor role in the portfolio at low-risk levels, while shares and bonds

dominated the portfolio across the risk spectrum. Thai-REITs did offer some diversification

benefits to shares and bonds, while the inclusion of property companies only increased the

level of risk. After the GFC, Thai-REITs have clearly increased their significance in the

mixed-asset portfolio, especially at the low-to-medium risk level. Shares and property

companies dominate the portfolio at high risk levels. This reflects Thai-REITs’ robust

performance post-GFC, enhancing returns while reducing risk to the mixed-asset portfolios.

The post-GFC efficient frontiers clearly show the positive impact of Thai-REITs in the

mixed-asset portfolios.

Table 5.29: Thai-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Dec 2003-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.47* 1.00

Property Companies 0.48* 0.69* 1.00 Bonds -0.27* -0.15* 0.06 1.00



Companies Bonds

REITs 1.00 Shares 0.93* 1.00




Companies Bonds

REITs 1.00 Shares 0.35* 1.00

Property Companies 0.47* 0.83* 1.00 Bonds 0.05* -0.13* -0.06 1.00


181

Figure 5.31: Thai-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-

GFC: Dec 2003-Feb 2008

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.6% 3.5% 5.4% 7.3% 9.2% 11.0% 12.9% 14.8% 16.7% 18.6% 20.5%

Allo

cati

on

Risk


0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0%

Re

turn

Risk



182

Figure 5.32: Thai-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-

GFC: Apr 2009-Apr 2012

SharesBonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.1% 4.1% 7.1% 10.0% 13.0% 16.0% 18.9% 21.9% 24.9% 27.8% 30.8%

Allo

cati

on

Risk


0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

Re

turn

Risk



183

5.7 Performance of Taiwan REITs


The performance summary for T-REITs over March 2006-April 2012 is presented in Table

5.30. Interestingly, T-REITs delivered the strongest average return of 11.09%, out-

performing all other assets; shares (5.76%), property companies (8.07%) and bonds (1.79%),

whilst having lower risk (17.78%) than shares (27.30%) and property companies (45.54%).

This resulted in T-REITs being the best asset class in terms of risk-adjusted performance, as

measured by a Sharpe ratio of 0.56. In contrast, property companies were the worst-

performing asset class (#4), with highest risk level and low risk-adjusted returns. This

highlights the superior characteristics of T-REITs over property companies for investors

seeking low risk investment exposure in the Taiwanese property market.

Table 5.30: T-REIT Risk-adjusted Performance Analysis: Mar 2006-Apr 2012


Companies Bonds


Annual Risk 17.78% 27.30% 45.54% 1.61%


Sharpe Ratio 0.56 0.17 0.15 0.38

Rank #1 #3 #4 #2

The correlation matrix among the four asset classes in Taiwan is presented in Table 5.31. It

is clear that shares and property stocks (r=0.77) had a higher correlation coefficient than

shares and REITs (r=0.58). This indicates that share investors would have received better

diversification benefits by holding T-REITs than listed property stocks. T-REITs were also

moderately correlated with property companies (r=0.54), suggesting the possibility for inter-

property diversification. In addition, T-REITs also offer diversification potential for bond

holders, as they are strongly and negatively correlated (r=-0.26) with bonds. This section has

highlighted the potential role of T-REITs in a mixed-asset portfolio. A detail portfolio

analysis in the mean-variance framework is reported in the following section.

184

Table 5.31: T-REIT Correlation Matrix: Mar 2006-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.58* 1.00

Property Companies 0.54* 0.77* 1.00 Bonds -0.26* -0.16 -0.14 1.00



Figure 5.33 displays the efficient frontiers using the classical mean-variance analysis. The

chart clearly demonstrates that adding T-REITs significantly improved portfolio efficiency.

Property companies, on the other hand, could not provide significant enhancement to the

portfolio. The bond-share and bond-share-property company efficient frontiers almost

overlap each other, suggesting that property companies can only provide minimal

enhancement. When T-REITs are added to both the two-asset (bond-share) and three-asset

(bond-share-property company) portfolios, the efficient frontiers were improved significantly,

shifting upwards.

Figure 5.33: T-REIT Efficient Frontiers: Mar 2006-Apr 2012

The dominant role of T-REITs in the four-asset efficient portfolio was further demonstrated

in Table 5.32 and Figure 5.34. Notably, the best portfolio return is 17.66%, at which T-REITs

dominate the allocation. Across the risk spectrum, bonds and T-REITs heavily weighted the

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% 45.0% 50.0%

Re

turn

Risk



185

portfolios, while shares only represented a minimum role and property companies were

entirely excluded. When T-REITs are included the share-bond portfolio, the average portfolio

return improves by 300 basis points, while the average standard deviation was reduced by

480 basis points.

Table 5.32: T-REIT Asset Allocations: Mar 2006-Apr 2012



0.15% 96.96% 2.89% 0.00% 2.06% 1.51%

0.00% 81.91% 18.09% 0.00% 3.47% 3.12%

0.00% 72.22% 27.78% 0.00% 4.37% 4.74%

0.00% 62.96% 37.04% 0.00% 5.23% 6.35%

0.00% 53.85% 46.15% 0.00% 6.08% 7.97%

0.00% 44.81% 55.19% 0.00% 6.92% 9.58%

0.00% 35.82% 64.18% 0.00% 7.76% 11.20%

0.00% 26.84% 73.16% 0.00% 8.59% 12.81%

0.00% 17.88% 82.12% 0.00% 9.42% 14.43%

0.00% 8.94% 91.06% 0.00% 10.26% 16.05%

0.00% 0.00% 100.00% 0.00% 11.09% 17.66%

Figure 5.34: T-REIT Asset Allocation Diagram: Mar 2006-Apr 2012

As the unconstraint portfolios tend to be over-exposed to one type of investment (i.e. T-

REITs), allocation constraints have been set in order to keep the portfolio in balance. In

particular, the maximum weights to T-REITs and property companies were capped at 10% of

the total value. Table 5.33 and Figure 5.35 report the asset allocation with constraints for

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.5% 3.1% 4.7% 6.4% 8.0% 9.6% 11.2% 12.8% 14.4% 16.0% 17.7%

Allo

cati

on

Risk


186

Taiwanese REITs and property companies. The portfolio returns range from 2.06% to 6.52%,

while the risk ranged from 1.51% to 26.41%. Similar to the results in other markets, the

weightings of shares increased with the level of volatility, while allocations to bonds

decreased. T-REITs and property companies were at their 10% allocation limits across all

risk levels. At the aggressive end of the risk spectrum, the efficient portfolio should be

invested 80% in shares, 10% in REITs and 10% in property companies.

Table 5.33: T-REIT Asset Allocations with Constraints: Mar 2006-Apr 2012



6.60% 80.85% 10.00% 2.54% 3.14% 4.00%

12.76% 72.19% 10.00% 5.05% 3.54% 6.49%

18.52% 63.92% 10.00% 7.56% 3.93% 8.98%

24.37% 55.70% 10.00% 9.94% 4.31% 11.47%

33.75% 46.25% 10.00% 10.00% 4.69% 13.96%

43.10% 36.90% 10.00% 10.00% 5.06% 16.45%

52.39% 27.61% 10.00% 10.00% 5.43% 18.94%

61.62% 18.38% 10.00% 10.00% 5.79% 21.43%

70.82% 9.18% 10.00% 10.00% 6.16% 23.92%

80.00% 0.00% 10.00% 10.00% 6.52% 26.41%

Figure 5.35: T-REIT Asset Allocation Diagrams with Constraints: Mar 2006-Apr 2012

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.51% 4.00% 6.49% 8.98% 11.47% 13.96% 16.45% 18.94% 21.43% 23.92% 26.41%

Allo

cati

on

Risk


187


The sub-period analysis reveals the detrimental impact of the GFC to Taiwan’s investment

environment (Table 5.34). However, the results showed that T-REITs have weathered the

GFC quite well as compared to the other asset classes. In particular, T-REITs were the least

affected sector in terms of average returns (-8.23%), as compared to the general stock market

(-33.31%) and property company (-55.55%), which was the most distressed sector. Moreover,

T-REIT volatility during the GFC (33.55%) was not as high as share (40.94%) and property

company (67.07%) risk levels. This reflects T-REITs’ strong financial status and low gearing

levels prior to and during the GFC. This was because the Taiwanese government imposes a

35% upper limit of T-REITs’ leverage ratios. Only T-REITs with a credit rating of “twAA”

or above can lift the borrowing ceiling to a maximum of 50% of their net worth.

T-REITs further showed superior performance than shares and property companies after the

GFC. They distributed the best post-GFC mean return of 25.23%, against that of shares at

16.98% and property companies at 23.28%. Consistently, the T-REIT risk level (13.73%)

was significantly lower than the equivalent figure for shares (23.67%) and property stocks

(39.81%). This resulted in them moving to the second best performing asset class (from #4

pre-GFC) in risk-adjusted terms, outpacing both shares (#3) and property companies (#4).

188

Table 5.34: T-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Mar 2006-Feb 2008


Companies Bonds


Annual Risk 8.60% 22.08% 36.05% 0.92%


Sharpe Ratio 0.10 0.64 1.07 0.36

Rank #4 #2 #1 #3



Companies Bonds


Annual Risk 33.55% 40.94% 67.07% 1.70%


Sharpe Ratio -0.29 -0.85 -0.85 0.26

Rank #2 #4 #3 #1



Companies Bonds


Annual Risk 13.73% 23.67% 39.81% 0.38%


Sharpe Ratio 1.80 0.70 0.57 2.30

Rank #2 #3 #4 #1

Figure 5.36 highlights the varying nature of T-REIT volatility over the period from March

2006 to April 2012. As depicted, all sectors had become significantly more volatile in

reaction to the GFC. The risk levels however have reduced substantially since the end of

2009 as Taiwan financial market started to recover. It can also be clearly seen that T-REITs

enjoyed a much lower risk than shares and property companies across the entire period. T-

REIT risk was only 60% of the risk level of shares and 38% that of property companies.

189

Figure 5.36: T-REIT 12-Month Rolling Volatility Analysis: Mar 2007-Apr 2012

The sub-period correlation analysis also demonstrates that correlation across all asset classes

have increased tremendously during the GFC (Figure 5.37). The highest correlation during

the GFC was that seen between shares and property companies (r=0.84), thus limiting their

diversification benefits. T-REIT correlations with the other listed sectors also increased

during the GFC, r=0.77 with shares and r=0.63 with property companies, but not to the same

magnitude of that between shares and property companies. After the GFC, while T-REITs’

correlation with shares and property stocks have lowered to the pre-GFC level (r=0.41 with

shares and r=0.49 with property companies), property companies remained highly correlated

with the general stock market (r=0.85). This will subsequently see T-REITs being more

attractive to investors than property companies to be included in a mixed-asset portfolio,

especially for the post-GFC period.

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

Mar

-20

07

Jun

-20

07

Sep

-20

07

De

c-2

00

7

Mar

-20

08

Jun

-20

08

Sep

-20

08

De

c-2

00

8

Mar

-20

09

Jun

-20

09

Sep

-20

09

De

c-2

00

9

Mar

-20

10

Jun

-20

10

Sep

-20

10

De

c-2

01

0

Mar

-20

11

Jun

-20

11

Sep

-20

11

De

c-2

01

1

Mar

-20

12


190

Figure 5.37: T-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Mar 2006-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.38* 1.00




Companies Bonds

REITs 1.00 Shares 0.77* 1.00




Companies Bonds

REITs 1.00 Shares 0.41* 1.00



The 12-month rolling correlation analysis is depicted in Figure 5.38. The graph clearly

reasserts that both T-REITs and property companies have become more like stocks during the

GFC. However, the trailing correlation of T-REITs and shares has been reduced significantly

post-GFC while the correlation between property companies and shares continued to trend

upward. T-REITs also have a low or negative correlation with bonds. These low or negative

correlations of T-REITs with shares and bonds in the recent period mean that they would help

to smooth out fluctuations in returns within a traditional portfolio of stocks and bonds.

191

Figure 5.38: T-REIT 12-Month Rolling Correlation Analysis: Mar 2007-Apr 2012

The sub-period asset allocations in Figure 5.39 and Figure 5.40 depict the changing role of T-

REITs in a mixed-asset portfolio context. Before the GFC, T-REITs only represented a minor

position in the mixed-asset portfolio. Only a small portion of the portfolio is allocated to T-

REITs at low risk levels. The post-GFC period, however, saw the increasingly significant

position of T-REITs in the portfolio. Together with bonds, T-REITs dominated the portfolio

across the risk spectrum. This reflects the stronger risk-adjusted performance of T-REITs

over shares and property companies post-GFC.

-1.00

-0.80

-0.60

-0.40

-0.20

-

0.20

0.40

0.60

0.80

1.00

Mar

-20

07

Jun

-20

07

Sep

-20

07

De

c-2

00

7

Mar

-20

08

Jun

-20

08

Sep

-20

08

De

c-2

00

8

Mar

-20

09

Jun

-20

09

Sep

-20

09

De

c-2

00

9

Mar

-20

10

Jun

-20

10

Sep

-20

10

De

c-2

01

0

Mar

-20

11

Jun

-20

11

Sep

-20

11

De

c-2

01

1

Mar

-20

12



192

Figure 5.39: Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC: Mar

2006-Feb 2008

Shares

Bonds

REITs

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.8% 4.3% 7.7% 11.2% 14.6% 18.0% 21.5% 24.9% 28.4% 31.8% 35.3%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

45.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%

Re

turn

Risk



193

Figure 5.40: Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC: Apr

2009-Apr 2012

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.8% 4.3% 7.7% 11.2% 14.6% 18.0% 21.5% 24.9% 28.4% 31.8% 35.3%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% 45.0%

Re

turn

Risk



194

5.8 Performance of Korean REITs


The average returns and volatility of K-REITs, bonds, shares and property companies for the

sample period are summarized in Table 5.35. The results indicate that K-REITs performed

poorly from July 2006 to April 2012 as compared to the other asset classes. K-REITs

reported an average annual return of 1.51% p.a., which was significantly lower than that of

shares (10.52%), property companies (9.75%) and bonds (4.96% pa). In relation to volatility,

K-REITs have a lower risk level (20.95%) than that of shares (27.29%) and property stocks

(36.38%). In terms of reward-to-risk performance, K-REITs (#4) underperformed all other

asset classes as they have the highest risk-to-return (13.84) ratio and lowest Sharpe ratio (-

0.09). Overall, the results show that K-REITs over sample period provided unfavourable

returns to investors both in absolute and risk-adjusted terms.

Table 5.35: K-REIT Risk-adjusted Performance Analysis: Jul 2006-Apr 2012


Companies Bonds


Annual Risk 20.95% 27.29% 36.38% 2.06%


Sharpe Ratio -0.09 0.26 0.17 0.74

Rank #4 #2 #3 #1

The correlation between K-REITs and the other asset classes during July 2006-April 2012

was examined to gain insight into the diversification benefits that KREITs might bring when

combined with other asset classes. As can be seen from Table 5.36, K-REITs moved with low

correlation to shares (r=0.44) and property companies (r=0.23), implying that K-REITs could

certainly offer diversification benefits to both share and property investors. However,

property companies had a lower correlation with shares (r=0.37) than K-REITs. These results

suggest that property companies in Korea probably offer better portfolio enhancement for

share investors than REITs. Interestingly, these are in contrast with results found in other

Asian countries.

195

Table 5.36: K-REIT Correlation Matrix: Jul 2006-Apr 2012


Companies Bonds

REITs 1.00 Shares 0.44* 1.00

Property Companies 0.23 0.37* 1.00 Bonds 0.03 -0.14 -0.19 1.00


The 12-month rolling volatility analysis of all asset classes in South Korea over the period of

July 2007-April 2012 is shown in Figure 5.41. It is clearly seen that the risk level of K-REITs

changed dramatically during this period. Their volatility doubled from around 4% in 2007 to

around 8% in 2008, due to the negative effects from the GFC. It then declined to the 4% level

by mid-2010, but then increased gradually again recently. This might be a result of the new

amendment that allows K-REITs to increase their gearing from two times to ten times their

NAVs, hence seeing K-REITs becoming riskier. With regard to the volatility of shares and

property companies, both have become more volatile during the GFC, but have then reduced

steadily. Noticeably, K-REITs have become riskier than shares and property companies

towards the end of the sampling period.

Figure 5.41: K-REIT 12-Month Rolling Volatility Analysis: Jul 2007-Apr 2012

The 12-month rolling correlation analysis documented in Figure 5.42 shows the correlation

between K-REITs and the other asset classes over time. The results indicate that the

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

Jul-

20

07

Oct

-20

07

Jan

-20

08

Ap

r-2

00

8

Jul-

20

08

Oct

-20

08

Jan

-20

09

Ap

r-2

00

9

Jul-

20

09

Oct

-20

09

Jan

-20

10

Ap

r-2

01

0

Jul-

20

10

Oct

-20

10

Jan

-20

11

Ap

r-2

01

1

Jul-

20

11

Oct

-20

11

Jan

-20

12

Ap

r-2

01

2


196

correlation between REITs and shares has been tracking that between property companies

and shares prior to the beginning of 2011. The correlation was relatively low at around 0.2

during 2007-2008, increased substantially to around 0.8 during 2008-2009 then reduced

quickly toward the end of 2010. However, there was a clear structural break at the beginning

of 2011 where the correlation between REITs and shares picked up again, while that between

property companies and shares continued to decline. The increasing association between K-

REITs and the equity sectors somewhat hampered their risk reduction potential in the mixed-

asset portfolio.

Figure 5.42: K-REIT 12-Month Rolling Correlation Analysis: Jul 2007-Apr 2012


Asset allocation diagrams and efficient frontiers have been constructed in order to further

examine the return enhancement and risk reduction benefits of K-REITs in a mixed-asset

framework. Table 5.37 and Figure 5.43 illustrate the allocation of K-REITs, shares, property

companies and bonds in an efficient mixed-asset portfolio across all risk levels. It is clearly

seen that K-REITs failed to enter the mixed-asset portfolio during this period, reflecting their

poor risk-adjusted performance. Not surprisingly, shares and bonds dominated the portfolio,

while property companies were included at the medium-to-high risk levels. The efficient

frontiers depicted in Figure 5.45 further demonstrate that only property companies could

slightly improve the bond-share portfolio, while K-REITs were inefficient for portfolio

enhancement.

-1.00

-0.80

-0.60

-0.40

-0.20

-

0.20

0.40

0.60

0.80

1.00

Jul-

20

07

Oct

-20

07

Jan

-20

08

Ap

r-2

00

8

Jul-

20

08

Oct

-20

08

Jan

-20

09

Ap

r-2

00

9

Jul-

20

09

Oct

-20

09

Jan

-20

10

Ap

r-2

01

0

Jul-

20

10

Oct

-20

10

Jan

-20

11

Ap

r-2

01

1

Jul-

20

11

Oct

-20

11

Jan

-20

12

Ap

r-2

01

2



197

Table 5.37: K-REIT Asset Allocations: Jul 2006-Apr 2012



1.00% 97.92% 0.00% 1.08% 5.06% 1.97%

13.45% 82.23% 0.00% 4.32% 5.91% 4.48%

21.76% 71.77% 0.00% 6.47% 6.48% 7.00%

29.78% 61.69% 0.00% 8.53% 7.02% 9.51%

37.69% 51.73% 0.00% 10.58% 7.56% 12.02%

45.49% 41.83% 0.00% 12.68% 8.10% 14.53%

53.31% 31.96% 0.00% 14.72% 8.63% 17.05%

61.22% 22.14% 0.00% 16.64% 9.16% 19.56%

69.00% 12.30% 0.00% 18.69% 9.69% 22.07%

76.80% 2.48% 0.00% 20.72% 10.22% 24.58%

100.00% 0.00% 0.00% 0.00% 10.52% 27.09%

Figure 5.43: K-REIT Asset Allocation Diagram: Jul 2007-Apr 2012

The asset allocations with constraints were reported and displayed in Table 5.38 and Figure

5.44 respectively. The results were quite similar to the unconstrained optimisation, since the

efficient portfolio was dominated by shares and bonds. Property companies only represented

a minor portion in the asset mix, while K-REITs were totally excluded. At the minimum

portfolio risk (1.97%), the asset weightings included 97.92% bonds, 1.00% shares and 1.08%

property stocks. At the maximum risk, the investment should be placed entirely in shares.

This is because bonds were defensive assets while shares were highly risky investments.

Shares

Bonds

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2.0% 4.5% 7.0% 9.5% 12.0% 14.5% 17.0% 19.6% 22.1% 24.6% 27.1%

Allo

cati

on

Risk


198

Table 5.38: K-REIT Asset Allocations with Constraints: Jul 2006-Apr 2012



13.47% 82.23% 0.00% 4.30% 5.91% 4.48%

21.76% 71.77% 0.00% 6.47% 6.48% 7.00%

29.76% 61.69% 0.00% 8.56% 7.02% 9.51%

38.18% 51.82% 0.00% 10.00% 7.56% 12.02%

47.69% 42.31% 0.00% 10.00% 8.09% 14.53%

57.09% 32.91% 0.00% 10.00% 8.61% 17.05%

66.43% 23.57% 0.00% 10.00% 9.13% 19.56%

75.72% 14.28% 0.00% 10.00% 9.65% 22.07%

84.98% 5.02% 0.00% 10.00% 10.17% 24.58%

100.00% 0.00% 0.00% 0.00% 10.52% 27.09%

Figure 5.44: K-REIT Asset Allocation Diagrams with Constraints: Jul 2007-Apr 2012

Shares

Bonds

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.97% 4.48% 7.00% 9.51% 12.02% 14.53% 17.05% 19.56% 22.07% 24.58% 27.09%

Allo

cati

on

Risk


199

Figure 5.45: K-REIT Efficient Frontiers: Jul 2007-Apr 2012


This section analysed the impact of the global financial crisis on the performance of K-

REITs, bonds, shares and property companies. The timeframe was split into three sub

periods: pre-crisis from July 2006 to February 2008, GFC from March 2008 to March 2009

and post-crisis from April 2009 to April 2012. As seen from Table 5.39, all asset classes were

severely affected by the GFC. Shares were the hardest hit, with the average annual return

falling from 23.60% to -25.27%; K-REIT and property company annual average returns

dropped from 8.04% to -4.39% and 3.57% to -25.66% respectively.

Volatility also increased across all investment options during the crisis. The highest increase

was for K-REITs, with the risk level increasing almost three times from 10.99% to 27.35%

(149% increase), much higher than that of shares from 21.01% to 44.71% (113% increase)

and property companies from 32.34% to 53.48% (65% increase). The results indicate that K-

REIT volatility was more sensitive to the global financial crisis than that of shares and

property companies. The crisis also affected the risk-adjusted performance of K-REITs,

shares and property companies, indicated by risk-to-return ratios of -0.31, -0.66 and -0.56

respectively.

In the post-GFC period from April 2009 to April 2012, all asset classes showed strong

improvements, as seen in Panel C of Table 5.39. The average annual return of shares

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk



200

increased significantly to 19.29%, while that of K-REITs and property companies rose to

0.19% and 30.01% respectively. In addition, risk levels were reduced across all asset classes.

However, K-REIT volatility (22.96%) was still higher than the period before the global

financial crisis. This further implies that K-REITs have become riskier and more associated

with equities after the crisis. The period after the crisis also saw considerable improvement in

the risk-adjusted performance of all asset classes. The fixed income sector remained stable

during all three periods, with average yields around 4.63%–5.43% p.a. and volatility of

around 1.05%–1.96%, making it the best performing asset class during and after the GFC.

Table 5.39: K-REIT Impact of the Global Financial Crisis Panel A: Pre-GFC: Jul 2006-Feb 2008


Companies Bonds


Annual Risk 10.99% 21.01% 32.34% 1.05%


Sharpe Ratio 0.31 0.90 -0.03 0.53

Rank #3 #1 #4 #2



Companies Bonds


Annual Risk 27.35% 44.71% 53.48% 1.70%


Sharpe Ratio -0.31 -0.66 -0.56 0.78

Rank #2 #4 #3 #1



Companies Bonds


Annual Risk 22.96% 21.55% 30.83% 1.96%


Sharpe Ratio -0.10 0.78 0.89 1.08

Rank #4 #3 #2 #1

The sub-period correlation analysis of K-REITs with the other asset classes is presented in

Table 5.40. Before the global financial crisis, K-REITs experienced substantially low

correlation with shares (r=0.32) and property companies (r=0.20). This suggests strong

diversification benefits that K-REITs could offer to both shares and property investors.

However, some loss of diversification potential was seen during the crisis, as the K-REIT

correlation coefficient with shares and property companies increased to r=0.69 and r=0.51

201

respectively. After the crisis, K-REIT correlations with shares and property companies fell to

r=0.30 and r=0.07 respectively. This correlation was lower than the pre-crisis levels.

Table 5.40: K-REIT Sub-period Correlation Analysis Panel A: Pre-GFC: Jul 2006-Feb 2008


Companies Bonds

REITs 1.00 Shares 0.32* 1.00

Property Companies 0.20 0.38* 1.00 Bonds -0.38 -0.15 -0.13 1.00



Companies Bonds

REITs 1.00 Shares 0.69* 1.00

Property Companies 0.51 0.75* 1.00 Bonds -0.27 -0.34 -0.22 1.00



Companies Bonds

REITs 1.00 Shares 0.30* 1.00

Property Companies 0.07 -0.14* 1.00 Bonds 0.19 0.08 -0.09 1.00


The asset allocation diagrams and efficient frontiers for the three sub-periods are displayed in

Figure 5.46 and Figure 5.47. For the period before the global financial crisis, K-REITs were

only included in the efficient portfolio at the low risk level while property companies were

excluded. This suggests that K-REITs offer better return enhancement and risk reduction than

property companies for the mixed-asset portfolio. After the crisis, the allocations to K-REITs

substantially reduced in the efficient portfolio due to the loss of diversification benefits. As

seen in the efficient frontiers, K-REITs could not improve returns when being combined in

the share-bond and share-bond-property company portfolios.

202

Figure 5.46: K-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Pre-GFC:

Jul 2006-Feb 2008

Shares

Bonds

REITs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.9% 2.9% 4.8% 6.8% 8.7% 10.7% 12.7% 14.6% 16.6% 18.5% 20.5%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0%

Re

turn

Risk



203

Figure 5.47: K-REIT Sub-Period Asset Allocation Diagrams and Efficient Frontiers: Post-GFC:

Apr 2009-Apr 2012

Shares

Bonds

Property Companies

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1.9% 4.8% 7.6% 10.5% 13.3% 16.2% 19.0% 21.9% 24.7% 27.5% 30.4%

Allo

cati

on

Risk


0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

Re

turn

Risk



204

5.9 Summary and Strategic Implications This chapter investigates the performance and significance of Asian REITs in domestic

mixed-asset portfolios across seven Asian markets (Japan, Singapore, Hong Kong, Malaysia,

Thailand, Taiwan and South Korea). The performance and portfolio analyses are studied

based on monthly data of domestic REIT, bond, share and property company total return

indices. The primary findings are as follows:

The results from this study indicate that most Asian REIT markets (four out of seven) under

investigation appear to outperform their respective stock markets (Table 5.41). In particular,

REITs outperformed the stock markets in Japan, Singapore, Hong Kong and Taiwan

respectively. In terms of risk, Asian REIT returns were observed to have lower volatility

levels than that of the general share markets across five (Hong Kong, Taiwan, Malaysia,

Thailand and South Korea) out of the seven markets, this includes the three major Asian

REIT markets. This might be attributed to the low leverage level and high dividend yields

across Asian REIT regimes. This low leverage coupled with high dividend payout policy has

reduced variation in REIT returns. On average, REIT risk accounts for approximately 82% of

the stock market risk. On a risk-adjusted basis, the analysis shows that the majority of Asian

REIT markets performed better than their share counterparts. Stronger risk-adjusted

performance was seen for Japan, Singapore, Hong Kong and Taiwan, while Malaysia,

Thailand and South Korea REITs performed no better than their domestic stock markets.

REITs in Malaysia, Thailand and South Korea are typically smaller in size, which might

hinder their abilities to achieve income growth through asset diversification.

Table 5.41: Performance Summary of Asian REITs Versus Stocks

Outperform stock market

Lower risk than stock market

Better risk adjusted

performance than stocks

Japan Yes No Yes

Singapore Yes No Yes

Hong Kong Yes Yes Yes

Taiwan Yes Yes Yes

Malaysia No Yes No

Thailand No Yes No

South Korea No Yes No

When Asian REIT performance was compared to listed property companies, the results were

similar to that between REITs and shares. In the markets where REITs outperformed shares

(Japan, Singapore, Hong Kong and Taiwan), the performance of REITs exceeded that of

205

listed property companies (Table 5.42). This was achieved, however, without inordinate risk.

REIT risk was significant lower than property company risk across all markets. The average

standard deviation of Asian REITs was equivalent to only 63% of the figure of property

company risk. This might be due to the fact that many listed property companies in Asia are

engaging in property development, which is subject to higher risk due to excessive leverage

and higher exposure to property development risk. Asian REITs on the other hand, are

structured as operating companies investing in income-producing properties. Consequently,

their income streams are more predictable and less volatile. All REIT indices under study,

except South Korea, outperformed the local listed property stocks in terms of risk-adjusted

returns.

Table 5.42: Performance Summary of Asian REITs Versus Property Companies

Outperform property

companies Lower risk than property

companies

Better risk adjusted

performance than property

companies

Japan Yes Yes Yes

Singapore Yes Yes Yes

Hong Kong Yes Yes Yes

Taiwan Yes Yes Yes

Malaysia No Yes Yes

Thailand No Yes Yes

South Korea No Yes No

Overall, the performance analysis shows that most Asian REIT markets provide superior

returns than both the general stock markets and listed property companies. Due to their low

leverage, high dividend payouts and steady earning streams, REIT returns are observed to be

less volatile than share and property stock returns. This in turn would provide greater stability

to investor portfolios. In addition, investors who are seeking exposure to local property

markets in Asia and a stable income stream, whilst do not wish to be exposed to the high

property market risk, should invest in Asian REITs rather than listed property companies.

The empirical evidence from this study showed that Asian REITs would provide

diversification benefits when included in a mixed-asset portfolio. REITs are found to have

moderate correlations with both shares (average r=0.66) and property companies (average

r=0.62). On the other hand, property companies and stocks were more highly correlated

(average r=0.80). This highlights that Asian REITs’ returns are subject to different

fundamental factors than shares and property companies. As a result, Asian REITs would

206

offer significant portfolio diversification for domestic investors. As REITs have a lower

correlation with shares than property companies, investors should also chose REITs over

property companies when selecting a property asset class to be included in the share

portfolios. In addition, the low correlation between REITs and property companies further

suggests that an inter-property diversification strategy would be possible.

Table 5.43: Asian REIT Diversification Summary

Correlation between

REITs and stocks

Correlation between

REITs and property

companies

Correlation between

property companies and

stocks

Japan 0.70 0.73 0.86

Singapore 0.88 0.88 0.93

Hong Kong 0.66 0.65 0.95

Taiwan 0.58 0.54 0.77

Malaysia 0.65 0.59 0.84

Thailand 0.68 0.70 0.86

South Korea 0.44 0.23 0.37

Average 0.66 0.62 0.80

Because of their strong performance, low risk and low correlations with other asset classes,

REITs represent a significant element of the domestic mixed-asset portfolios across Asia.

According to the asset allocation results, on average about one-third of the total domestic

portfolio should be allocated to REITs. Significant allocations to REITs are targeted for Japan

(50.1%), Singapore (53.0%), Hong Kong (56.2%) and Taiwan (54.3%). This was clearly a

manifestation of the strong risk-adjusted performance of local REITs over shares and

property companies in these four markets; this includes the three main Asian REIT markets.

In contrast, Malaysia (2.9%) and Thailand (1.1%) saw minimal level of REITs in the

domestic portfolios, while South Korean REITs were not included in the efficient frontier at

all. This is a result of poorer performance of REITs in Malaysia, Thailand and South Korea.

Table 5.44: Asset Allocation Summary: Average Allocation in Mixed-Asset Portfolios

REITs Bonds Shares Property

Companies

Japan 50.14% 49.86% 0.00% 0.00%

Singapore 52.99% 46.92% 0.09% 0.00%

Hong Kong 56.19% 43.26% 0.54% 0.00%

Taiwan 54.33% 45.65% 0.01% 0.00%

Malaysia 2.86% 43.03% 54.09% 0.00%

Thailand 1.12% 45.07% 53.81% 0.00%

South Korea 0.00% 43.28% 46.32% 10.40%

Average 31.09% 45.30% 22.12% 1.49%

207

REIT performance has varied significantly during the periods under study due to the GFC.

Returns of REITs, shares and property companies have been negatively impacted by the

GFC. However, Asian REITs (5/7 markets) have shown more robust recovery than shares

(4/7) and property stocks (3/7) post-GFC. Except for Japan and South Korea, all REIT

markets in Asia have enhanced returns post-GFC versus pre-GFC levels. Shares have also

made good comebacks from the GFC, while property companies in most markets have not

recovered back to their pre-GFC levels.

Most of Asian REIT markets (4/7) have experienced lower risk post-GFC than pre-GFC. On

the other hand, the majority of share (6/7) and property company (5/7) markets still have

higher risk levels. This again reasserts the defensive nature of Asian REITs during a volatile

market environment, due to their stable income and high dividend payouts. It is interesting to

note that the Japanese investment market was still struggling to recover from the GFC (lower

returns and higher risk). Part of this might be a result of the devastating effects of the March

2011 tsunami and the subsequent nuclear crisis.

Table 5.45: Return and Risk post-GFC Versus pre-GFC Enhanced Returns Post-GFC

REITs Shares Property Companies

Japan No No No

Singapore Yes Yes No

Hong Kong Yes No No

Malaysia Yes Yes Yes

Thailand Yes Yes Yes

Taiwan Yes Yes No

South Korea No No Yes

Reduced risk post-GFC

REITs Shares Property Companies

Japan No No No

Singapore Yes No No

Hong Kong Yes No No

Malaysia Yes No Yes

Thailand Yes Yes No

Taiwan No No No

South Korea No No Yes

The risk-adjusted performance of Asian REITs relative to shares and property stocks has also

changed considerably over the different sub-periods (Table 5.46). Most Asian REIT markets

did not outperform the corresponding stock markets and listed property companies before the

208

GFC. However, after the GFC, Asian REITs delivered superior returns to shares and property

companies across the board, except in South Korea. Given this strong risk-adjusted

performance of REITs, investors should be allocating more into REITs than shares and

property companies during the post-GFC period.

Table 5.46: Asian REIT Risk-adjusted Performance by Sub-period Did REITs deliver superior risk-adjusted returns to stock?

Pre-GFC Post-GFC

Japan Yes Yes

Singapore Yes Yes

Hong Kong No Yes

Malaysia No Yes

Thailand No Yes

Taiwan No Yes

South Korea No No

Did REITs deliver superior risk-adjusted returns to property companies?

Pre-GFC Post-GFC

Japan Yes Yes

Singapore No Yes

Hong Kong No Yes

Malaysia No Yes

Thailand No Yes

Taiwan No Yes

South Korea Yes No

The position of Asian REITs in domestic portfolios has improved greatly after the GFC for

most countries. On average, the REIT proportion in the efficient portfolio was increased

significantly from 9.6% pre-GFC to 36.2% post-GFC. The allocations to REITs have been

increased in the mixed-asset portfolios for Singapore (from 10.4% to 52.4%), Hong Kong

(from 6.8% to 54.6%), Malaysia (from 0.3% to 38.6%), Thailand (0.9% to 28.2%) and

Taiwan (0.4% to 51.2%). On the other hand, due to their weak performance, the post-GFC

amount allocated to REITs for Japan was reduced from 48.1% to 28.5%, while REITs in

South Korea was excluded from the portfolio.

209

Table 5.47: GFC and Asset Allocation: Average Portfolio Allocation in REITs

Pre-GFC Post-GFC Enhanced allocation

post-GFC

Japan 48.08% 28.45% No

Singapore 10.42% 52.41% Yes

Hong Kong 6.79% 54.56% Yes

Malaysia 0.28% 38.60% Yes

Thailand 0.86% 28.16% Yes

Taiwan 0.36% 51.15% Yes

South Korea 0.69% 0.00% No

Average 9.64% 36.19%

Overall, this chapter has demonstrated the performance of Asian REITs relative to their

domestic share market and listed property companies, as well as their position in a mixed-

asset portfolio. Sub-period analysis has also revealed the changing nature of REIT

performance within a domestic investment framework. These have resulted in several

important implications for investors when they consider adding Asian REITs in their

domestic portfolios. In particular, it has reinforced the role of Asian REITs in the post-GFC

period.

So far, the analysis has been conducted within a local investment framework. It would be

interesting to examine REIT performance within a regional pan-Asia and international

context. The next chapter will extend the analysis further in order to assess the role of Asian

REITs within a pan-Asian and international investment portfolio.

210

6 CHAPTER 6

THE SIGNIFICANCE OF ASIAN REITs IN REGIONAL AND

INTERNATIONAL INVESTMENT PORTFOLIOS

6.1 Introduction

The previous chapter has presented the role and significance of Asian REITs in the domestic

mixed-asset portfolios. This chapter extends the analysis to a regional and global investment

perspective. The risk-adjusted performance of the seven Asian REIT markets is assessed

during the period from October 2001 to April 2012. An inter-REIT portfolio composed of the

seven Asian REIT markets is developed and analysed. In addition, a pan-Asia REIT index

and four sub-indexes are developed to assess the risk-adjusted performance and

diversification benefits of a pan-Asia REIT strategy in a regional and global investment

context. Sub-period analysis is also performed to assess the impact of the GFC on the pan-

Asian performance index.

The remainder of this chapter is organised as follows. The following section provides a

summary of the key findings on the risk-return profiles and diversification benefits of Asian

REIT in a regional inter-property portfolio. This is followed by the empirical findings relating

to the significance of a pan-Asian REIT strategy in a global investment portfolio. The last

section provides the summary and concluding comments.

6.2 Asian REITs in a Regional and Inter-property Investment Strategy

6.2.1 Performance Analysis

Table 6.1 contains the comparison of the average annual returns and risk for the seven REITs

market in Asia over the common period from July 2006 to April 2012. The results featured

the strong performance of REITs in Malaysia (16.65%) and Taiwan (16.61%), with both

markets delivering the best average returns among the seven REIT markets. On the other

hand, the poorest performer was South Korea REITs, with the market index falling by 1.02%

p.a. Japan (2.77%) also showed weak performance as compared to its regional competitors

over this period. In terms of risk, Thailand was the least risky market in Asia, as measured by

the annualised standard deviation of 13.76%, while Singapore had the highest volatility, with

a standard deviation of 33.57%.

211

Interestingly, the results suggest that the emerging markets, with the exception of South

Korea, tended to have lower volatility than the developed REIT markets, while delivering

better average returns. On a risk-adjusted basis, the three emerging markets of Malaysia,

Taiwan and Thailand ranked first, second and third respectively, outperforming the

developed markets of Hong Kong (#4), Singapore (#5) and Japan (#6). The low risk

environments in the three emerging REIT markets might be a result of the legislative

restrictions on borrowing levels which led to less exposure to interest rate volatility. In

addition, the lower trading volumes and the lack of financial sophistication and speculative

trading in those emerging REIT markets may make them relatively more stable during this

period.

Table 6.1: Asian REITs Risk-adjusted Performance Analysis - Common Sample: Jul 2006-Apr

2012

Average Annual Return Annual Risk Risk/Return Ratio Rank

Developed Asia Japan 2.77% 31.20% 11.26 #6

Singapore 12.75% 33.57% 2.63 #5

Hong Kong 15.61% 23.72% 1.52 #4

Average 10.38% 29.49%

Emerging Asia Malaysia 16.65% 14.99% 0.90 #1

Taiwan 16.61% 22.15% 1.33 #3

Thailand 11.36% 13.76% 1.21 #2

South Korea -1.02% 30.83% NA NA

Average 10.90% 20.43%

Figure 6.1 plots the risk and returns of the seven Asian REIT markets in the four risk/return

quadrants. Malaysia, Thailand, Taiwan and Hong Kong are in the most favourable place, the

high return-low risk quadrant. In contrast, Japan and South Korea were in the least desirable

section with high volatility, but negligible returns. On the other hand, Singapore REITs

belonged to the high risk-high return quadrant. Interestingly, the analysis shows that the risk

and return profile of Asian REITs over this period were not consistent with the conventional

high risk-high return trade-off. Less risky markets, such as Malaysia and Taiwan tended to

have higher average returns, while more volatile markets such as Japan and South Korea

generated lower returns. However, as the analysis was based on a short common period, one

might suspect that the ex-post mean returns might not be good estimates of the true potential

212

returns. Therefore, a further analysis using individual samples was conducted in order to

obtain long-term estimates of the ex-ante parameters.

Figure 6.1: Asian REIT Risk and Return Profile - Common Sample: Jul 2006-Apr 2012

Table 6.2 features the descriptive statistics for individual samples of the seven Asian REIT

markets. The longest data set was that of Japan, from August 2001 through April 2012. For

other markets, the data started in August 2003 for Singapore, December 2003 for Thailand,

August 2008 for Hong Kong, September 2005 for Malaysia, March 2006 for Taiwan and July

2006 for South Korea. Over the individual time periods, Singapore delivered the highest

average return of 18.78% p.a., followed by Taiwan at 15.74% p.a. and Hong Kong at 13.68%

p.a. In risk-adjusted terms, Malaysia was the best performer (ranked #1) with a risk/return

ratio of 1.12, followed by Taiwan (#2) and Singapore (#3).

Japan

Singapore

Hong KongMalaysia Taiwan

Thailand

South Korea-2.00%

0.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

16.00%

18.00%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%

Re

turn

Risk

213

Table 6.2: Asian REITs Risk-adjusted Performance Analysis - Individual Samples: Aug 2001-

Apr 2012*

0 Average Annual Return Annual Risk Risk/Return Ratio Rank


Singapore 18.78% 29.30% 1.56 #3

Hong Kong 13.68% 22.30% 1.63 #4

Average 14.00% 25.68%

Emerging Asia Malaysia 13.37% 14.97% 1.12 #1

Taiwan 15.74% 21.77% 1.38 #2

Thailand 6.60% 12.99% 1.97 #5

South Korea -1.02% 30.83% NA NA

Average 8.67% 20.14%

Note: * various individual periods

When the risk and return profile of Asian REITs were graphed in Figure 6.2, risk and returns

appeared to have an inverse relationship. This is different to the results obtained from the

common samples, but more consistent with the conventional risk/return trade-off. Assets that

generated higher average returns tended to have higher risk. The only exception is South

Korea, which appeared to be an outlier. Singapore REITs generated the highest return

potential, but also registered the highest volatility. On the other hand, Thailand REITs had the

lowest level of risk, but provided the lowest level of return. This indicates that the trade-off

between risk and returns seems to hold true over the long term for Asian REITs. High risk

REITs provided higher historical returns over the long term. These results have important

investment implications for the application of modern portfolio theory to develop a

diversification strategy for Asian REITs.

214

Figure 6.2: Asian REIT Risk and Return Profile - Individual Samples: Aug 2001-Apr 2012*

Note: * various individual periods

Table 6.3 depicts the correlation matrix between the seven Asian REIT indexes. Over this

period, the correlation coefficients ranged from r=0.41 between Malaysia and Japan to r=0.71

between South Korea and Singapore. Asian REITs were found to be less interrelated than

stocks. The average correlation coefficient between Asian REITs was r=0.60, which is

relatively lower as compared to the average correlation between stocks of r=0.70. As seen in

Figure 6.3, most of the cross-correlations between Asian REIT markets are lower than those

between the corresponding stock markets. On average, Asian REITs were 20% less correlated

than stocks. As a consequence, investors can receive better diversification benefits from an

inter-REIT investment strategy than an inter-share portfolio.

Table 6.3: Asian REIT Correlation Matrix: Jul 2006-Apr 2012

Japan Singapore Hong

Kong Malaysia Taiwan Thailand South

Korea

Japan 1.00 Singapore 0.70* 1.00

Hong Kong 0.58* 0.67* 1.00 Malaysia 0.41* 0.68* 0.47* 1.00

Taiwan 0.59* 0.69* 0.69* 0.52* 1.00 Thailand 0.54* 0.69* 0.60* 0.66* 0.68* 1.00

South Korea 0.48* 0.71* 0.45* 0.57* 0.59* 0.55* 1.00


Japan

Singapore

Hong KongMalaysia

Taiwan

Thailand

South Korea

-5%

0%

5%

10%

15%

20%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

Re

turn

Risk

215

Figure 6.3: Regional Correlations REITs vs. Shares: Jul 2006-Apr 2012

6.2.2 Portfolio Analysis

Table 6.4 exhibits the optimum asset allocations across the seven Asian REIT markets based

on the mean-variance framework over the common period from July 2006 to April 2012.

Portfolio returns ranged from 13.35% to 16.65%, with the corresponding risk levels ranging

from 12.93% to 14.88%. Importantly, the analysis reveals the dominant roles of the

developing REIT markets (i.e. Malaysia, Thailand and Taiwan) in the regional inter-REIT

portfolio over this period. In particular, Malaysia REITs accounted for the majority of the

investment mix, with an increasing weighting as the risk increases. On average, three quarters

of the optimum portfolio value should be allocated to M-REITs to achieve maximum

efficiency. This is a manifestation of the strong risk-adjusted performance of M-REITs over

this period. Allocations to Thailand and Taiwan REITs are also significant. On the other

hand, the only developed REIT market included in the optimum mix was Hong Kong, but at a

marginal level. There was no exposure to Japan, Singapore and South Korea REITs.

- 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

South Korea-Singapore

Singapore-Japan

Taiwan-Hong Kong

Thailand-Singapore

Taiwan-Singapore

Thailand-Taiwan

Malaysia-Singapore

Hong Kong-Singapore

Thailand-Malaysia

Thailand-Hong Kong

South Korea-Taiwan

Taiwan-Japan

Hong Kong-Japan

South Korea-Malaysia

South Korea-Thailand

Thailand-Japan

Taiwan-Malaysia

South Korea-Japan

Malaysia-Hong Kong

South Korea-Hong Kong

Malaysia-Japan

REITs

Shares

216

Table 6.4: Asian REIT Allocations - Common Samples: Jul 2006-Apr 2012

Portfolio Allocations Portfolio

Sum JP SG HK ML TW TL KO Return Risk 100% 0.00% 0.00% 0.00% 37.53% 0.00% 62.47% 0.00% 13.35% 12.93%

100% 0.00% 0.00% 0.35% 56.39% 0.00% 43.26% 0.00% 14.36% 13.13%

100% 0.00% 0.00% 1.53% 62.67% 1.06% 34.74% 0.00% 14.79% 13.32%

100% 0.00% 0.00% 1.85% 67.11% 3.11% 27.93% 0.00% 15.15% 13.52%

100% 0.00% 0.00% 2.13% 71.03% 4.87% 21.97% 0.00% 15.46% 13.71%

100% 0.00% 0.00% 2.35% 74.50% 6.56% 16.60% 0.00% 15.74% 13.90%

100% 0.00% 0.00% 2.54% 77.77% 8.05% 11.64% 0.00% 16.00% 14.10%

100% 0.00% 0.00% 2.85% 80.81% 9.35% 6.99% 0.00% 16.24% 14.29%

100% 0.00% 0.00% 3.02% 83.70% 10.67% 2.61% 0.00% 16.47% 14.49%

100% 0.00% 0.00% 0.00% 91.67% 8.33% 0.00% 0.00% 16.64% 14.68%

100% 0.00% 0.00% 0.00% 100.00% 0.00% 0.00% 0.00% 16.65% 14.88%

Average 0.00% 0.00% 1.51% 73.02% 4.73% 20.75% 0.00% 15.53% 13.90%

Figure 6.4: Asian REIT Efficient Frontier Compositions - Common Samples: July 2006 to April

2012

Table 6.5 shows the optimal weights of the Asian REITs in the efficient frontier using

individual sample data. The portfolio returns ranged from 9.1% to 18% and portfolio risk

spanned from 12.9% to 33.3%. Compared to the allocations based on the common data, the

“emerging three” (Malaysia, Taiwan and Thailand) still dominated the efficient portfolio at

moderate to low risk spectrum. However, the optimal share in Singapore REITs has increased

dramatically, especially at the moderate to aggressive risk targets (Figure 6.5). Singapore

REITs entered the efficient portfolio at the 17% risk level (15% return) and rise

progressively, first at the expense of Thailand and Malaysia at low risk ranges then at the

Hong Kong

Malaysia

Taiwan

Thailand

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

12.9% 13.1% 13.3% 13.5% 13.7% 13.9% 14.1% 14.3% 14.5% 14.7% 14.9%

Allo

cati

on

Risk

217

expense of Taiwan REITs at more risky ranges. On average, Singapore REITs made up about

43% of the efficient portfolio, while Taiwan (30%), Malaysia (21%) and Thailand (6%)

REITs together accounted for the remaining of the total value. Japan, Hong Kong and South

Korea were not included in the portfolios due to their poor risk-adjusted performance.

Table 6.5: Asian REIT Allocations - Individual Samples


SUM JP SG HK ML TW TL KO Return Risk 100% 0.00% 0.00% 0.00% 37.53% 0.00% 62.47% 0.00% 9.14% 12.93%

100% 0.00% 0.00% 0.00% 69.99% 30.01% 0.00% 0.00% 14.08% 14.97%

100% 0.00% 10.42% 0.00% 49.29% 40.29% 0.00% 0.00% 14.89% 17.01%

100% 0.00% 21.15% 0.00% 35.91% 42.95% 0.00% 0.00% 15.53% 19.05%

100% 0.00% 30.77% 0.00% 23.80% 45.43% 0.00% 0.00% 16.11% 21.09%

100% 0.00% 39.65% 0.00% 12.25% 48.09% 0.00% 0.00% 16.65% 23.13%

100% 0.00% 48.32% 0.00% 1.41% 50.27% 0.00% 0.00% 17.17% 25.17%

100% 0.00% 62.78% 0.00% 0.00% 37.22% 0.00% 0.00% 17.65% 27.21%

100% 0.00% 76.23% 0.00% 0.00% 23.77% 0.00% 0.00% 18.05% 29.25%

100% 0.00% 88.50% 0.00% 0.00% 11.50% 0.00% 0.00% 18.43% 31.29%

100% 0.00% 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% 18.78% 33.32%

Average 0.00% 43.44% 0.00% 20.93% 29.96% 5.68% 0.00% 16.04% 23.13%

Figure 6.5: Asian REIT Efficient Frontier Composition - Individual Samples

Figure 6.6 compares the efficient frontiers of the common and individual samples. As seen

from the graph, the efficient frontier constructed based on common sample returns was above

Singapore

Malaysia

Taiwan

Thailand

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

12.9% 15.0% 17.0% 19.0% 21.1% 23.1% 25.2% 27.2% 29.2% 31.3% 33.3%

Allo

cati

on

Risk

218

that based on individual sample returns. This suggests that the former efficient portfolio

outperformed the latter one at every risk level. However, the returns (13.35% through

16.65%) and risk (12.93% to 14.88%) ranges of the common-sample efficient frontier were

quite limited and the portfolio tended to be overweighted by Malaysia and Thailand REITs.

On the other hand, the efficient frontier based on individual samples spanned over a broader

range of returns (9.14% to 18.78%) and risk targets (12.93% to 33.32%). This was a result of

Singapore REITs entering the portfolio at high risk and high return levels.

Figure 6.6: Efficient Frontiers - Common vs. Individual Samples

In order to reduce the over-exposure of the portfolio to emerging REITs, allocation

restrictions of 10% were imposed on the emerging REIT markets in the optimisation process.

The allocation results with constraints on emerging REITs are shown in Table 6.6 and Figure

6.7. The constraint portfolio saw the additional inclusion of Japan and South Korean REITs.

Hong Kong REITs have become the major asset class in the portfolio, while Singapore

REITs still could not make it to the asset mix. The minimum risk portfolio saw Hong Kong

REITs accounting for more than half (51%) of the value and Japanese REITs taking up 8.3%,

while the emerging REIT markets (Malaysia, Taiwan, Thailand and South Korea) accounting

for 10% each. The maximum return (15.81%) portfolio contained mainly Hong Kong (80%),

Malaysia (10%) and Thailand (10%) REITs.

EF Individual Samples

EF Common Samples

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

Re

turn

Risk

219

Table 6.6: Asian REIT Allocations with Constraints: July 2006 to April 2012


SUM JP SG HK ML TW TL KO Return Risk 100% 8.30% 0.00% 51.70% 10.00% 10.00% 10.00% 10.00% 12.66% 19.15%

100% 0.08% 0.00% 62.39% 10.00% 10.00% 10.00% 7.53% 14.13% 19.35%

100% 0.00% 0.00% 66.55% 10.00% 10.00% 10.00% 3.45% 14.82% 19.56%

100% 0.00% 0.00% 69.72% 10.00% 10.00% 10.00% 0.28% 15.34% 19.76%

100% 0.00% 0.00% 71.34% 10.00% 10.00% 8.66% 0.00% 15.45% 19.97%

100% 0.00% 0.00% 72.81% 10.00% 10.00% 7.19% 0.00% 15.51% 20.17%

100% 0.00% 0.00% 74.27% 10.00% 10.00% 5.73% 0.00% 15.57% 20.38%

100% 0.00% 0.00% 75.72% 10.00% 10.00% 4.28% 0.00% 15.63% 20.58%

100% 0.00% 0.00% 77.16% 10.00% 10.00% 2.84% 0.00% 15.69% 20.79%

100% 0.00% 0.00% 78.58% 10.00% 10.00% 1.42% 0.00% 15.75% 21.00%

100% 0.00% 0.00% 80.00% 10.00% 10.00% 0.00% 0.00% 15.81% 21.20%

Average 0.76% 0.00% 70.93% 10.00% 10.00% 6.37% 1.93% 15.12% 20.17%

Figure 6.7: Asian REIT Efficient Frontier Composition with Constraints: July 2006 to April

2012


Results of the performance analysis over sub-periods: pre-GFC, GFC and post-GFC were

presented in Table 6.7. The risk and returns of Asian REITs have changed dramatically over

these periods. Asian REIT performance has been heavily affected during the GFC. Average

returns suffered major declines, while volatility levels increased drastically. Average risk of

Asian REITs almost tripled during the GFC. The severe impact of the GFC on Asian REIT

risk and returns can be observed more clearly in Figure 6.8. All markets have been falling

Japan

Hong Kong

Malaysia

TaiwanThailand

South Korea

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

19.1% 19.4% 19.6% 19.8% 20.0% 20.2% 20.4% 20.6% 20.8% 21.0% 21.2%

Allo

cati

on

Risk

220

from the top left to the bottom right corner of the risk return plane signifying a severe

reduction in returns coupled with rapidly rising risk.

Singapore REITs suffered the biggest impact, with the average return falling from 39.56%

p.a. pre-GFC to -55.19% p.a. during the GFC. However, it also had the most robust recovery

after the crisis, having the average return rebounding to 36.90% p.a. Performance

improvement post-GFC have also been witnessed across the board due to the implementation

of monetary measures by governments and central banks in the region. Volatility has been

reduced significantly; from 53.31% to 18.85% for developed REITs and from 33.10% to

16.99% for emerging REITs. The majority of Asian REIT markets have improved

performance in the post-GFC as compared to the pre-GFC period. In particular, Japan, Hong

Kong, Taiwan and Thailand all had higher average returns post-GFC versus pre-GFC as

depicted in Figure 6.8. Another important observation is that Malaysia had strong risk-

adjusted performance across all three periods. In particular, it was ranked the best market in

the pre-GFC and GFC period and second in the post-GFC. Significant progress has also been

seen for Thailand, jumping from #3 pre-GFC to #1 post-GFC and Taiwan, rising two places

from #6 to #4.

The sub-period correlation analysis is presented in Table 6.8. In the pre-GFC period, Asian

REIT markets had relatively low and negative correlations, with an average correlation

coefficient of 0.22. Correlation figures are not statistically significant (p>0.05) suggesting a

low interdependency among the regional REIT markets before the GFC. The highest

correlation coefficient was that between Singapore and Japan (r=0.55), while the lowest value

was that between Taiwan and Malaysia (r=-0.17) which moved inversely with each other.

These low and negative correlations indicate strong diversification benefits for investors

holding an Asian REIT portfolio.

221

Table 6.7: Asian REIT Sub-period Performance Analysis:

Pre-GFC: Jul 2006-Feb 2008



Singapore 39.56% 24.49% 0.62 #2

Hong Kong 15.87% 14.67% 0.92 #4

Average 21.32% 20.89% Emerging Asia

Malaysia 21.78% 11.89% 0.55 #1

Taiwan 6.04% 11.10% 1.84 #6

Thailand 11.73% 8.95% 0.76 #3

South Korea 9.48% 12.85% 1.36 #5

Average 12.26% 11.20%

GFC: Mar 2008-Mar 2009


Developed Asia Japan -32.66% 58.31% -1.79 #5

Singapore -55.19% 55.49% -1.01 #2

Hong Kong -18.33% 46.14% -2.52 #7

Average -35.39% 53.31% Emerging Asia

Malaysia -23.92% 16.68% -0.70 #1

Taiwan -15.53% 38.47% -2.48 #6

Thailand -22.59% 23.40% -1.04 #3

South Korea -31.74% 53.87% -1.70 #4

Average -23.45% 33.10%

Post-GFC: Apr 2009-Apr 2012



Singapore 36.90% 22.78% 0.62 #5

Hong Kong 30.16% 14.36% 0.48 #3

Average 27.49% 18.85% Emerging Asia

Malaysia 31.99% 13.93% 0.44 #2

Taiwan 37.09% 17.99% 0.48 #4

Thailand 25.99% 9.25% 0.36 #1

South Korea 6.55% 26.78% 4.09 #7

Average 25.40% 16.99%

222

Figure 6.8: Risk and Return Profile of Asian REITs over Sub-periods

JP Pre-GFC

SG Pre-GFC

HK Pre-GFC

ML Pre-GFC

TW Pre-GFC

TL Pre-GFCKO Pre-GFC

KO Post-GFC

TL Post-GFC

TW Post-GFC

ML Post-GFC

HK Post-GFC

SG Post-GFC

JP Post-GFC

KO GFC

TL GFC

TW GFC

ML GFC

HK GFC

SG GFC

JP GFC

-60.00%

-40.00%

-20.00%

0.00%

20.00%

40.00%

60.00%

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0%

Re

turn

Risk

223

However, the GFC has resulted in significant increases in cross-correlations among Asian

REITs. All correlation coefficients were statistically significant, suggesting the seven Asian

REIT markets have become more correlated. The average correlation coefficient has almost

quadrupled during the GFC, from r=0.22 to r=0.81, resulting in a significant loss of

diversification benefits. The highest increase in correlation was that between Hong Kong and

Japan, which increased by almost forty times from r=0.02 to r=0.73. However, the period

post-GFC has seen some improvements in the diversification benefits of Asian REITs, as

correlations have falling back to around the r=0.37 region. Nevertheless, it might take some

time before they could get back to the pre-GFC levels. All correlation coefficients are

statistically significant at the p=0.05 level. Compared to the pre-GFC figures, these results

show that Asian REITs have become more interdependent since the GFC.

Figure 6.9 presents the optimal allocations among the seven Asian REIT markets over the

pre- and post-GFC period. Prior to the GFC, Singapore and Malaysia dominated the pan-

Asian REIT portfolio due to their strong risk-adjusted performance and low correlations with

other markets. Singapore REITs were more dominant at the high risk ranges, while Malaysia

REITs were more prevailing at the medium risk levels. Over this period, about half of the

portfolio value was placed into Singapore REITs, one-third into Malaysia REITs and the

remaining across the other five REIT markets, primarily at the low risk targets. These results

highlight the vital role of both the Malaysia and Singapore REIT markets in a regional REIT

portfolio. Singapore REITs offered the greatest potential for growth but also the greatest risk,

while M-REITs were generally less volatile and offered moderately good returns. Other REIT

markets offered low risk and low return potential.

The post-GFC period saw the rising role of Thailand and Taiwan REIT markets in the

portfolio. In particular, the optimal shares allocated to Thailand and Taiwan REITs have

increased significantly at the expense of Singapore REITs. At the lower end of the risk

spectrum, the majority of the portfolio was invested in Thailand REITs, while the high risk

ranges were occupied primarily by Taiwan REITs. Hong Kong REITs entered the portfolio at

the defensive to moderate risk levels. Malaysia and Singapore continued to hold substantial

positions in the asset mix, but at much lower levels as compared to the pre-GFC period. On

the other hand, due to their weak performance, South Korea and Japan only accounted for

minor portions in the portfolio across both periods.

224

Table 6.8: Asian REIT Sub-period Correlation Analysis




Korea


Hong Kong 0.02 0.17 1.00 Malaysia -0.11 0.15 0.07 1.00

Taiwan 0.38 0.47* 0.24 -0.17 1.00 Thailand -0.13 0.26 0.25 0.46* 0.37 1.00

South Korea 0.42 0.44* -0.11 0.34 0.37 0.25 1.00

Average Correlation 0.22




Korea


Hong Kong 0.73* 0.81* 1.00 Malaysia 0.78* 0.91* 0.72* 1.00

Taiwan 0.88* 0.87* 0.86* 0.78* 1.00 Thailand 0.77* 0.83* 0.73* 0.91* 0.85* 1.00

South Korea 0.73* 0.91* 0.67* 0.80* 0.79* 0.75* 1.00





Korea


Hong Kong 0.40* 0.56* 1.00 Malaysia 0.23 0.65* 0.37* 1.00

Taiwan 0.05 0.46* 0.46* 0.48* 1.00 Thailand 0.28 0.50* 0.33* 0.44* 0.44* 1.00

South Korea -0.02 0.50* 0.11 0.49* 0.33* 0.27 1.00



225

Figure 6.9: Asset Allocation Sub-period



The efficient frontiers of the portfolios in the two sub-periods; pre-GFC and post-GFC, were

compared and displayed in Figure 6.10. As depicted in the line charts, the post-GFC efficient

frontier was placed above the pre-GFC efficient frontier. This implies that portfolio

performance has been improved post-GFC. In general, the efficient portfolio returns were

increased by about 50 to 60 basis points after the GFC. However, the risk and return ranges

post-GFC were more constrained. Portfolio risk ranged from 8.53% to 17.74%, while returns

Japan

SingaporeHong Kong

MalaysiaTaiwan

Thailand

South Korea

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

6.7% 8.4% 10.1% 11.8% 13.6% 15.3% 17.0% 18.7% 20.4% 22.1% 23.9%

Allo

cati

on

Risk

JapanSingaporeHong Kong

Malaysia

Taiwan

Thailand

South Korea

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

8.5% 9.4% 10.4% 11.3% 12.2% 13.1% 14.1% 15.0% 15.9% 16.8% 17.7%

Allo

cati

on

Risk

226

ranged from 26.67% to 37.09%. Before the crisis, portfolio risk stretched from 6.68% to

23.87% and return potentials varied from 13.21% through 39.56%.

Figure 6.10: Efficient Frontiers Pre-GFC vs. Post-GFC Period

6.3 Asian REITs in a Global Investment Strategy

The previous section has provided a discussion on the role and significance of Asian REITs

in a regional and inter-property context. This section extends the analysis of Asian REITs to a

global context. The investment implications for US and international property investors are

also assessed. A pan-Asia REIT index consisting of all listed REITs across the seven Asian

REIT markets was constructed. In addition, three sub-indexes; a pan-Asia excluding Japan

REIT index, an emerging Asia REIT index and a developed Asia REIT index, were created.

6.3.1 Performance Analysis

Table 6.9 reveals the risk and return profile of REIT and share indexes in Asia, the US and

Australia. A number of interesting observations can be noted from the results. All Asian

REIT indexes outperformed their US and Australian counterparts, as well as the

corresponding share markets. Importantly, this was achieved while having lower volatility

than stocks.

Consequently, the four Asia REIT indexes took up the top 4 positions in terms of risk-

adjusted performance. The pan-Asia ex. Japan index was the best performer in both absolute

EF Post-GFC

EF Pre-GFC

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

45.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0%

Re

turn

Risk

227

and relative terms (#1). It delivered the best average return, 16.04% p.a., while having a

relatively low standard deviation of 22.93%. In contrast, US (#11) and Australia (#12) REITs

delivered lacklustre performance over this period with relative high risk and low growth. US

and Australia REITs produced inferior performance to their corresponding share markets

during the study period.

Table 6.9: Performance Analysis – REITs vs. Shares: Jul 2006-Apr 2012

Average Annual Return Annual Risk Risk/Return

Asset class (%) (%) Ratio Rank

REITs

Pan-Asia 11.32% 22.81% 2.02 #3

Pan-Asia ex. Japan 16.04% 22.93% 1.43 #1

Developed Asia 10.97% 23.80% 2.17 #4

Emerging Asia 12.43% 18.01% 1.45 #2

US 8.57% 29.02% 3.38 #11

Australia 4.25% 31.47% 7.41 #12

Average 10.60% 24.67%

Shares Pan-Asia 10.95% 25.56% 2.33 #7

Pan-Asia ex. Japan 11.85% 26.29% 2.22 #6

Developed Asia 10.38% 25.22% 2.43 #8

Emerging Asia 11.38% 28.43% 2.50 #9

US 5.51% 17.41% 3.16 #10

Australia 13.11% 29.02% 2.21 #5

Average 10.53% 25.32%

The risk/return scatter diagrams of the six REIT indexes were displayed on Figure 6.11. As

seen on the risk/return plane, the Asian REIT markets were superior to the more developed

REIT markets of the US and Australia. The Asian REITs produced greater returns while

having relatively less risk. It can be seen that the four Asian REIT indexes lie on the upper

left portion of the scatter-plot, which is the less risk, higher return quadrant. Quite the

opposite, US and Australia REITs were in lower right corner, meaning that they produce less

returns, but have more volatility. The poorest performer was Australia, which had the lowest

average return of 4.25% p.a., but the highest level of risk, as measured by a standard

deviation of 31.47%.

There might be a number of explanations for the superior risk-adjusted performance of Asian

REITs over US and Australia REITs during this period. Firstly, Asian REIT returns were less

volatile than US and Australia REITs due to their borrowing restrictions which made them

228

less exposed to leverage risk. One exception might be the case of Singapore REITs, which

was the most risky market among the Asian REITs as discussed in the previous section.

Secondly, most Asian REITs were more focused and less aggressive than their counterparts

in the US and Australia in terms of asset acquisitions. Thirdly, their property portfolios are

predominantly domestic, which made them less exposed to volatility in the global financial

market. The GFC, which was caused by the mortgage market crisis in the US during this

period, would have played a major part in the high level of volatility in US and Australia

REIT markets. More analysis of the impact of the GFC will be discussed in a later section.

Figure 6.11: Risk and Return Profile of Asian REITs: Jul 2006-Apr 2012

Table 6.10 depicts the cross-correlation matrix between global REITs and shares. Several

interesting observations have been observed. The analysis shows that Asian REITs had lower

correlation with US and Australia financial markets than Asian shares. The correlation

coefficients of the pan-Asia REIT index with US and Australia share markets were 0.67 and

0.79 respectively. In comparison, the correlation measures of the pan-Asia shares with their

US and Australia equivalents were 0.76 and 0.87 respectively. This signifies that US and

Australia investors would receive better diversification benefits by including Asian REITs

rather than shares in their investment portfolios.

Another important finding is that Asian REITs were more connected with the financial

markets (both REIT and share) in Australia than in the US. This is attributed to the strong

PA REITs

PAexJP REITs

DA REITs

EA REITs

US REITs

AU REITs4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

17.0% 19.0% 21.0% 23.0% 25.0% 27.0% 29.0% 31.0% 33.0%

Re

turn

Risk

229

economic and financial linkages between Australia and Asia. On average, Asian REITs were

30% more correlated with Australia than with the US. Consequently, the risk reduction

effects from investing in Asian REITs would be greater for US-based investors than for

Australia-based investors.

The correlation analysis of sub-indexes further reported that emerging Asia REITs were less

exposed to the global financial markets than developed Asia REITs as indicated by lower

correlation measures. The emerging Asia REITs would therefore play an important role in a

global investment strategy. In addition, there is a low level of correlation between Asian

REITs and US REITs. The correlation coefficients ranged from 0.42 to 0.62, indicating the

advantage of Asian REITs in an international REIT investment strategy.

230

Table 6.10: Correlation Matrix - Whole Period: : Jul 2006-Apr 2012

Asset Class PA

REITs PAexJP

REITs DA

REITs EA

REITs US

REITs AU

REITs PA

Shares PAexJP

Shares DA

Shares EA

Shares US

Shares AU

Shares

PA REITs 1.00 PAexJP REITs 0.86* 1.00

DA REITs 1.00* 0.86* 1.00 EA REITs 0.63* 0.66* 0.61* 1.00

US REITs 0.59* 0.60* 0.58* 0.42* 1.00 AU REITs 0.79* 0.78* 0.78* 0.56* 0.71* 1.00

PA Shares 0.76* 0.86* 0.75* 0.63* 0.54* 0.75* 1.00 PAexJP Shares 0.72* 0.85* 0.71* 0.58* 0.52* 0.75* 0.98* 1.00

DA Shares 0.74* 0.83* 0.74* 0.59* 0.52* 0.72* 0.99* 0.97* 1.00 EA Shares 0.74* 0.85* 0.73* 0.67* 0.57* 0.76* 0.94* 0.92* 0.88* 1.00

US Shares 0.67* 0.74* 0.67* 0.49* 0.80* 0.78* 0.76* 0.73* 0.72* 0.79* 1.00 AU Shares 0.79* 0.84* 0.78* 0.63* 0.66* 0.90* 0.87* 0.86* 0.84* 0.87* 0.84* 1.00


231

6.3.2 Portfolio Analysis

To assess the role of Asian REITs in an international investment context, a global portfolio

that consisted of pan-Asia REITs, US REITs, Pan-Asia shares and US shares was

constructed. The optimal asset allocations into each asset are presented in Table 6.11. The

portfolio returns ranged from 6.23% to 11.32%, while portfolio risk was from 17.20% to

22.70%. Pan-Asia REITs accounted for the largest share in the efficient portfolio (averaging

61%), followed by US shares (31%) and pan-Asia shares (8.2%). Allocations to US REITs

were only minimal due to their poor risk-adjusted performance over this period. As seen in

Figure 6.12, the portfolio portion in US shares was more significant at the prudent risk level,

then gradually reduced as risk levels escalated allowing for the steadily increased share of

pan-Asia REITs and shares. At the high end of the risk spectrum, pan-Asian REITs accounted

for the majority of the portfolio value.

Table 6.11: Global Portfolio Asset Allocations: : Jul 2006-Apr 2012


Sum PA REITs US REITs PA Shares US Shares Return Risk

100% 12.32% 0.00% 0.00% 87.68% 6.23% 17.20%

100% 38.27% 0.00% 0.00% 61.73% 7.73% 17.75%

100% 49.31% 0.00% 0.00% 50.69% 8.37% 18.30%

100% 55.17% 0.00% 3.09% 41.74% 8.88% 18.85%

100% 59.51% 0.00% 6.75% 33.74% 9.33% 19.40%

100% 63.52% 0.00% 10.07% 26.41% 9.75% 19.95%

100% 67.13% 0.00% 13.30% 19.57% 10.13% 20.50%

100% 70.58% 0.00% 16.32% 13.10% 10.50% 21.05%

100% 73.94% 0.00% 19.13% 6.93% 10.84% 21.60%

100% 76.49% 1.40% 21.79% 0.33% 11.18% 22.15%

100% 100.00% 0.00% 0.00% 0.00% 11.32% 22.70%

Average 60.57% 0.13% 8.22% 31.08% 9.48% 19.95%

232

Figure 6.12: Global Asset Allocation: : Jul 2006-Apr 2012

Table 6.12 and Figure 6.13 depict the constrained portfolio allocations. The maximum weight

constraints on Pan-Asia REITs and shares were set at 10%. The results show that the efficient

portfolio was dominated by US shares at the low risk levels and US REITs at high risk. Pan-

Asia REITs and shares were limited at 10%, but maintained their positions across the risk

scale. On average, the efficient portfolios contained 43.21% US REITs, 37.70% US shares,

43.21% Pan-Asian REITs and 9.09% Pan-Asia shares.

Table 6.12: Global Portfolio Asset Allocations with Constraints: : Jul 2006-Apr 2012


Sum PA REITs US REITs PA Shares US Shares Return Risk

100% 10.00% 0.00% 0.00% 90.00% 6.09% 17.21%

100% 10.00% 8.48% 10.00% 71.52% 6.90% 18.09%

100% 10.00% 19.78% 10.00% 60.22% 7.24% 18.98%

100% 10.00% 29.33% 10.00% 50.67% 7.53% 19.87%

100% 10.00% 37.88% 10.00% 42.12% 7.80% 20.75%

100% 10.00% 45.77% 10.00% 34.23% 8.04% 21.64%

100% 10.00% 53.18% 10.00% 26.82% 8.26% 22.52%

100% 10.00% 60.25% 10.00% 19.75% 8.48% 23.41%

100% 10.00% 67.04% 10.00% 12.96% 8.69% 24.30%

100% 10.00% 73.61% 10.00% 6.39% 8.89% 25.18%

100% 10.00% 80.00% 10.00% 0.00% 9.09% 26.07%

Average 10.00% 43.21% 9.09% 37.70% 7.91% 21.64%

Pan-Asia REITs

US REITsPan-Asia Shares

US Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

17.2% 17.8% 18.3% 18.9% 19.4% 19.9% 20.5% 21.0% 21.6% 22.1% 22.7%

Allo

cati

on

Risk

233

Figure 6.13: Global Asset Allocation with Constraints: : Jul 2006-Apr 2012

The previous section highlighted the significant role of Asian REITs in a global portfolio

context. This section assesses Asia REITs within a regional mixed-asset portfolio that

consisted of pan-Asia REITs and shares. The developed and emerging REITs were placed in

a portfolio with the developed and emerging Asia shares. The asset allocations are shown in

Table 6.13 and Figure 6.14. The results depict the dominant role of the emerging Asia (EA)

REITs in the portfolio. EA REITs account for 81.01% to 100.00% of the portfolio across the

risk ranges. The developed Asia (DA) REITs only accounted for a small portion in the asset

mix at the low risk levels, but were reduced as the risk increased. DA shares only appeared at

the low end of the risk spectrum, while the EA shares could not be included at all. These

results were quite consistent with findings in the previous analysis for the individual

countries, highlighting the dominance position of developing Asia REITs in the portfolio due

to their low risk and favourable returns over this period.

PA REITs

US REITs

PA Shares

US Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

17.2% 18.1% 19.0% 19.9% 20.8% 21.6% 22.5% 23.4% 24.3% 25.2% 26.1%

Allo

cati

on

Risk

234

Table 6.13: Regional Asset Allocations: : Jul 2006-Apr 2012


Sum DA REITs EA REITs DA Shares EA Shares Return Risk

100% 13.59% 81.01% 5.40% 0.00% 12.12% 17.60%

100% 12.16% 86.74% 1.11% 0.00% 12.23% 17.63%

100% 10.68% 89.32% 0.00% 0.00% 12.27% 17.66%

100% 8.59% 91.41% 0.00% 0.00% 12.31% 17.69%

100% 6.91% 93.09% 0.00% 0.00% 12.33% 17.72%

100% 5.48% 94.52% 0.00% 0.00% 12.35% 17.76%

100% 4.20% 95.80% 0.00% 0.00% 12.37% 17.79%

100% 3.04% 96.96% 0.00% 0.00% 12.39% 17.82%

100% 1.96% 98.04% 0.00% 0.00% 12.40% 17.85%

100% 0.95% 99.05% 0.00% 0.00% 12.42% 17.88%

100% 0.00% 100.00% 0.00% 0.00% 12.43% 17.92%

Average 6.14% 93.27% 0.59% 0.00% 12.33% 17.76%

Figure 6.14: Regional Asset Allocations: Jul 2006-Apr 2012

To reduce the overweighting in the emerging REITs, maximum constraints were placed on

the emerging markets, so that the allocations to these assets cannot be more than 10% of the

total value. As a result, the portfolio has become more diversified in terms of the number of

assets included. In the unrestricted model above, the portfolio was exposed heavily to EA

REITs. However, when constraints on the emerging markets were put in place, the portfolio

has become more diversified. All four assets were included in the portfolio at different stages.

In the least risky (21.59%) portfolio, there were 55.2% in DA REITs, 10% in EA REITs and

34.8% in DA shares. The increase in risk saw the continued increase in allocation to the DA

Developed Asia REITs

Emerging Asia REITs

Developed Asia Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

17.6% 17.6% 17.7% 17.7% 17.7% 17.8% 17.8% 17.8% 17.9% 17.9% 17.9%

Allo

cati

on

Risk

235

REITs and the reduction in DA shares, while EA REITs maintained their stable position. As

the risk goes up, the EA shares started to enter the portfolio. On average, DA REITs

represented 68.70% of the portfolio, while DA shares, EA REITs and EA shares accounted

for 13.67%, 10.00% and 7.63% respectively.

Table 6.14: Regional Asset Allocations: : Jul 2006-Apr 2012


Sum DA

REITs EA REITs DA Shares EA Shares Return Risk

100% 55.19% 10.00% 34.81% 0.00% 10.91% 21.59%

100% 62.65% 10.00% 25.73% 1.62% 10.97% 21.66%

100% 63.01% 10.00% 22.13% 4.86% 11.00% 21.73%

100% 63.77% 10.00% 18.75% 7.48% 11.03% 21.80%

100% 64.23% 10.00% 15.77% 10.00% 11.06% 21.87%

100% 68.06% 10.00% 11.94% 10.00% 11.08% 21.94%

100% 71.09% 10.00% 8.91% 10.00% 11.10% 22.01%

100% 73.67% 10.00% 6.33% 10.00% 11.12% 22.08%

100% 75.98% 10.00% 4.02% 10.00% 11.13% 22.15%

100% 78.07% 10.00% 1.93% 10.00% 11.14% 22.22%

100% 80.00% 10.00% 0.00% 10.00% 11.16% 22.29%

Average 68.70% 10.00% 13.67% 7.63% 11.07% 21.94%

Figure 6.15: Regional Asset Allocations: Jul 2006-Apr 2012

Developed Asia REITs

Emerging Asia REITs

Developed Asia Shares

Emerging Asia Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

21.59% 21.66% 21.73% 21.80% 21.87% 21.94% 22.01% 22.08% 22.15% 22.22% 22.29%

Allo

cati

on

Risk

236


Table 6.15 exhibits the sub-period performance analysis highlighting the impact of the GFC

on REIT and share performance. As the statistics show, the crisis has taken a heavy toll on

both the global REIT and share markets. Average market returns across the board declined by

about 40% p.a. during the GFC period. This was coupled with the significant increase in

volatility. Australia REITs suffered the heaviest loss during this period, with returns falling

by almost 70% p.a., while risk level increased from 16% to 68%.

The post-GFC has seen significant improvements across all REIT markets. All REIT indexes

had enhanced returns post-GFC as compared to pre-GFC. Even though risk remained

relatively high, the risk-adjusted performance post-GFC has improved significantly. Asian

REITs lead the way, with the emerging Asia and pan-Asia ex. Japan taking up the top and

second position. Importantly, all Asia REIT indexes outperformed their corresponding stock

markets after the GFC.

The sub-period correlation matrix was presented in Table 6.16. The data in the table shows

that the inter-correlations were low or negative in the period before the crisis and very high

and positive during the GFC. The average correlation coefficient jumped from r=0.52 pre-

GFC to r=0.87 during GFC, eroding the diversification benefits between the different asset

classes. The post-GFC period has gained back some diversification benefits as correlations

dropped, but are yet to be back to the pre-GFC levels.

237

Table 6.15: Sub-period Performance Analysis




REITs

Pan-Asia 16.52% 15.53% 0.94 #6

Pan-Asia ex. Japan 23.65% 14.65% 0.62 #1

Developed Asia 16.41% 16.07% 0.98 #7

Emerging Asia 12.87% 18.56% 1.44 #12

US 11.40% 16.40% 1.44 #11

Australia 14.95% 15.74% 1.05 #9

Average 15.97% 16.16%

Shares Pan-Asia 19.96% 16.36% 0.82 #4

Pan-Asia ex. Japan 22.23% 17.76% 0.80 #3

Developed Asia 19.73% 16.36% 0.83 #5

Emerging Asia 19.50% 19.63% 1.01 #8

US 7.09% 10.07% 1.42 #10

Australia 23.90% 16.13% 0.67 #2

Average 18.73% 16.05%




REITs

Pan-Asia -34.25% 48.05% -1.40 #2

Pan-Asia ex. Japan -41.68% 46.05% -1.11 #7

Developed Asia -35.13% 50.20% -1.43 #1

Emerging Asia -22.97% 27.46% -1.20 #4

US -53.80% 57.70% -1.07 #9

Australia -67.92% 68.04% -1.00 #11

Average -42.62% 49.58%

Shares Pan-Asia -41.68% 45.22% -1.09 #8

Pan-Asia ex. Japan -42.60% 47.10% -1.11 #6

Developed Asia -40.85% 43.35% -1.06 #10

Emerging Asia -43.54% 52.37% -1.20 #3

US -36.87% 32.99% -0.89 #12

Australia -49.38% 58.79% -1.19 #5

Average -42.49% 46.64%

238




REITs

Pan-Asia 23.49% 15.68% 0.67 #4

Pan-Asia ex. Japan 32.14% 17.28% 0.54 #2

Developed Asia 22.99% 16.65% 0.72 #6

Emerging Asia 27.76% 10.23% 0.37 #1

US 40.00% 24.15% 0.60 #3

Australia 30.57% 21.13% 0.69 #5

Average 29.49% 17.52%

Shares Pan-Asia 21.29% 25.16% 1.18 #11

Pan-Asia ex. Japan 20.93% 24.68% 1.18 #10

Developed Asia 19.15% 25.52% 1.33 #12

Emerging Asia 25.14% 25.31% 1.01 #9

US 21.54% 15.73% 0.73 #7

Australia 26.65% 25.36% 0.95 #8

Average 22.45% 23.63%

239

Table 6.16: Sub-period Correlation Analysis

Pre-GFC: July 2006-Feb 2008

Asset Class PA

REITs PAexJP

REITs DA

REITs EA

REITs US

REITs AU

REITs PA

Shares PAexJP

Shares DA

Shares EA

Shares US

Shares AU

Shares


DA REITs 1.00* 0.66* 1.00 EA REITs 0.32* 0.30* 0.30* 1.00

US REITs 0.39* 0.34* 0.39* -0.01* 1.00 AU REITs 0.43* 0.45* 0.44* 0.17* 0.46* 1.00*

PA Shares 0.44* 0.59* 0.44* 0.31* 0.27* 0.63* 1.00* PAexJP Shares 0.32* 0.55* 0.32* 0.19* 0.30* 0.67* 0.92* 1.00*

DA Shares 0.46* 0.55* 0.46* 0.26* 0.31* 0.63* 0.98* 0.87* 1.00* EA Shares 0.33* 0.59* 0.33* 0.41* 0.11* 0.53* 0.87* 0.87* 0.76* 1.00*




240


Asset Class PA

REITs PAexJP

REITs DA

REITs EA

REITs US

REITs AU

REITs PA

Shares PAexJP

Shares DA

Shares EA

Shares US

Shares AU

Shares


DA REITs 1.00* 0.93* 1.00 EA REITs 0.94* 0.96* 0.93* 1.00

US REITs 0.71* 0.70* 0.70* 0.71* 1.00 AU REITs 0.90* 0.88* 0.90* 0.86* 0.76* 1.00*






241


Asset Class PA

REITs PAexJP

REITs DA

REITs EA

REITs US

REITs AU

REITs PA

Shares PAexJP

Shares DA

Shares EA

Shares US

Shares AU

Shares


DA REITs 1.00* 0.85* 1.00 EA REITs 0.63* 0.80* 0.59* 1.00

US REITs 0.35* 0.43* 0.33* 0.48* 1.00 AU REITs 0.73* 0.69* 0.73* 0.55* 0.58* 1.00*






242

Figure 6.16 shows the asset allocation diagrams for the pre-GFC and post-GFC periods.

During the period before the GFC, Pan-Asia REITs had a major role in the portfolio. The

optimal allocation to Pan-Asia REITs was consistent over a broad spectrum of relatively safe

to aggressive risk targets. The allocation to pan-Asia shares increased gradually with the risk

levels while US shares dominated the low risk ranges. US REITs entered the portfolio at the

low to medium risk ranges. Across the middle range of the risk spectrum, pan-Asia REITs

make up about a quarter of the optimal portfolio. The portfolio was dominated mostly by US

shares at the lower end and pan-Asia shares at the upper end of the risk scale.

The post-GFC period saw the increased share of US REITs in the portfolio at the expense of

US shares and pan-Asia shares, which were completed eliminated. Pan-Asia REITs however

retained their strong position in the optimal mix, mostly at the conservative risk targets. At

the lower end of the risk preferences, the optimal portfolio contained mostly US shares and

pan-Asia REITs. The average share allocated to pan-REITs across the risk spectrum was

around one-third of the total portfolio value.

Figure 6.16: Sub-period Asset Allocation Diagrams


Pan-Asia REITs

US REITs

Pan-Asia Shares

US Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

9.6% 10.2% 10.9% 11.6% 12.2% 12.9% 13.6% 14.2% 14.9% 15.5% 16.2%

Allo

cati

on

Risk

243


6.4 Summary and Strategic Implications

This chapter set out to investigate the role of Asian REITs in a regional and global investment

context. The risk-adjusted performance was assessed based on the seven Asian REIT

markets; Japan, Singapore, Hong Kong, Malaysia, Taiwan, Thailand and South Korea. A

pan-Asia performance index and three additional sub-indexes; pan-Asia excluding Japan,

emerging Asia and developed Asia, were also constructed. The diversification potentials of a

pan-Asia REIT strategy in both a regional and global context were evaluated. The results

offered several important implications for US and international investors in Asian REITs.

Within a regional context, the three emerging REIT markets of Malaysia, Taiwan and

Thailand offered better returns at a relatively lower level of risk than the developed markets

over the period under study. This was owing to the fact that these emerging markets have

more restrictive leverage rules imposed by the government. The low levels of borrowing have

been beneficial during the period of high uncertainly, in particular during the global financial

crisis. However, when longer time periods were considered, Singapore REITs delivered the

highest return as well as a highest level of risk, particularly before and after the GFC. The

correlations among Asian REITs were lower than those between the regional equity markets.

Therefore, regional investors would receive greater diversification benefits by investing

across Asian REITs than shares. The optimal REIT portfolio would consist of mostly

Thailand and Malaysia REITs at the low end of the risk spectrum, Taiwan and Singapore

Pan-Asia REITs

US REITs

US Shares

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

13.9% 14.9% 15.9% 16.9% 17.9% 18.9% 19.9% 20.9% 21.8% 22.8% 23.8%

Allo

cati

on

Risk

244

REITs at the medium ranges and mostly Singapore REITs at the aggressive end of the risk

preferences. Across all risk ranges, the efficient portfolio on average would be composed of

43% in Singapore, 30% in Taiwan, 21% in Malaysia and about 6% in Thailand REITs.

Within a global context, Asian REITs also delivered strong performance over the study

period. Table 6.17 provides a summary of the performance analysis of Asian REITs in

comparison to stocks. All the four pan-Asian REIT indexes outperformed their corresponding

stock markets. Importantly, these high returns were achieved without having a higher level of

volatility. Consequently, Asian REITs registered better risk-adjusted performance than the

stock markets. This is in contrast to the lacklustre performance of the more developed REIT

markets of the US and Australia during this period. The findings also show that Asian REITs

consistently outperformed US REITs in risk-adjusted terms. The regulatory regimes and

operational structures of Asian REITs might have contributed to their favourable

performance over this period. Asian REITs are restricted in borrowing limits which made

them less exposed to leverage risk over this period. In addition, Asian REIT strategies were

less aggressive, more focused and their properties are predominantly domestic. These unique

characteristics of Asian REITs have played important roles in negating the adverse effects of

the GFC.

Table 6.17: Performance Summary - pan-Asia REITs Versus Stocks

REIT market Outperform

stock market

Lower risk

than stock

market

Better risk-adj.

performance than

stocks

Better risk-adj.

performance than US

REITs

Pan-Asia Yes Yes Yes Yes

Pan-Asia ex. Japan Yes Yes Yes Yes

Developed Asia Yes Yes Yes Yes

Emerging Asia Yes Yes Yes Yes

US Yes No No NA

Australia No No No No

With regards to the impact of the GFC, the sub-period analysis revealed the severe affects of

the crisis on both the REIT and equity markets. During the GFC, market returns declined

sharply amid rapidly rising volatility. The cross-market correlations between global markets

have risen swiftly, resulting in the significant loss of diversification. However, the post-GFC

period saw a robust recovery of the REIT markets in Asia. Asian REITs, especially the

emerging Asia REITs, have been more robust than the corresponding stock markets post-

GFC. Figure 6.17 compares the performance in the post-GFC period against the pre-GFC

245

period. All REIT markets have enhanced returns post-GFC. Even though risk levels were still

high (except for the emerging Asia REITs), the risk-adjusted performance of all REIT

markets post-GFC were better than that pre-GFC. This is quite the opposite to the poorer risk-

adjusted performance of the share markets. Only two of the share markets, the emerging Asia

and US, have recovered from the GFC.

Figure 6.17: Summary Performance Post-GFC vs. Pre-GFC

Enhanced returns Reduced risk

Enhanced risk-adj.

returns

REITs Pan-Asia Yes No Yes

Pan-Asia ex. Japan Yes No Yes

Developed Asia Yes No Yes

Emerging Asia Yes Yes Yes

US Yes No Yes

Australia Yes No Yes

Shares

Pan-Asia Yes No No

Pan-Asia ex. Japan No No No

Developed Asia No No No

Emerging Asia Yes No Yes

US Yes No Yes

Australia Yes No No

The sub-period analysis further demonstrated the strong performance of Asian REITs over

the stock markets since the GFC (Figure 6.18). Most Asian REIT markets did not outperform

the stock markets prior to the GFC. However, during and after the GFC, Asian REITs have

consistently delivered better returns than stocks. In terms of risk, all Asian REIT markets

were less volatile than their corresponding stock markets in the periods before and after the

GFC. In terms of risk-adjusted performance, all REITs delivered superior returns to stocks

post-GFC. It is interesting to note that the pan-Asia ex. Japan REIT index was the only

market that consistently outperformed its corresponding stock index across all three sub-

periods. An investment strategy across Asian REITs without Japan appeared to be a good

strategy for international investors seeking exposure to the high growth potential of the Asian

property markets.

246

Figure 6.18: Summary Sub-period Performance of REITs vs. Stocks

Did REITs outperform

stocks? Did REITs have lower risk

than stocks?

Did REITs deliver

superior risk-adj.

returns to stocks?

REITs Pre-GFC GFC Post-GFC Pre-GFC GFC Post-GFC Pre-GFC Post-GFC

Pan-Asia No Yes Yes Yes No Yes No Yes

Pan-Asia ex. Japan Yes Yes Yes Yes Yes Yes Yes Yes

Developed Asia No Yes Yes Yes No Yes No Yes

Emerging Asia No Yes Yes Yes Yes Yes No Yes

US Yes No Yes No No No No Yes

Australia No No Yes Yes No Yes No Yes

As a result of their favourable risk-adjusted performance and low correlations with other

asset classes, Asian REITs played a significant role in the global investment portfolio. The

optimal portfolio over the full period would be composed of primarily Asian REITs (61%)

and US shares (31%). Asia shares only accounted for a minor portion of the optimal mix

(6%) due to their strong correlations with the US markets. Over the sub-periods, about a

quarter to a third of the portfolio should be allocated to Asian REITs. On the other hand, Asia

shares only played a significant role in the portfolio before the GFC, but were completely

excluded after the crisis.

247

7 CHAPTER 7

THE DYNAMICS OF RETURNS AND VOLATILITY IN

ASIAN REIT MARKETS

7.1 Introduction

The primary purpose of this study is to examine the dynamic transmission of REIT returns

and volatility between the seven Asian REIT markets: Japan, Singapore, Hong Kong,

Malaysia, Taiwan, Thailand and South Korea from June 15, 2006 to April 27, 2012.

The rationale for this study is motivated by a number of reasons. Firstly, since the early work

of Markowitz (1952) on portfolio theory, the issue of market integration has become an

important subject for both academics and investors. Previous studies on stock markets have

provided evidence that markets that are geographically and economically close tend to

influence one another and such integration or comovements is a result of closer political and

economic cooperation among countries.

Thus, this chapter seeks to understand the extent to which a certain movement in one REIT

market affects subsequent movements in other REIT markets in Asia. Secondly, most of the

previous studies mainly focused on the interactions between the U.S. and major Asian REIT

markets, namely Japan, Hong Kong and Singapore, while little attention has been paid to

emerging markets such as Malaysia, Taiwan, Thailand and South Korea. Recent years have

witnessed these four emerging markets enjoying rapid growth in market capitalization,

underpinned by strong economic growth and favourable regulatory changes. Therefore, the

linkages of these emerging markets with developed markets in the region deserve closer

attention. This study aims to answer the question on whether the developed REIT markets

influence the emerging REIT markets in Asia.

Thirdly, since a number of previous studies on stock markets found that international stock

market correlations tend to increase during a financial crisis, this study examines whether

REIT markets also experience the same phenomenon. In addition, the impact of the GFC on

Asian REIT returns and volatility are assessed. Therefore, this study makes the following

contributions to the existing literature on Asian listed property:

248

provision of empirical evidence on the nature of information transmission and market

linkages between Asian REIT markets at both the mean return and volatility levels;

extension of the current literature to four emerging REIT markets in Asia: Malaysia,

Taiwan, Thailand and South Korea; and

provision of updated evidence on the effects of the GFC on the performance of REIT

markets in Asia.

The remainder of the paper is organized as follows. Section 7.2 provides the preliminary

analysis of the data and empirical methodology. Section 7.3 reports the test results on mean

and volatility spillovers between the seven Asian REIT markets. Section 7.4 concludes the

chapter.

7.2 Preliminary Analysis

Table 7.1 illustrates the summary statistics for daily returns of the seven markets, as well as

statistics testing for normality and autocorrelation over the period from June 15, 2006 to

April 27, 2012. Taiwan (0.059%) delivered the highest average returns followed by Hong

Kong (0.056%) and Malaysia (0.056%), while the lowest figure was that of South Korea

(-0.005%). In terms of risk, as measured by the standard deviation, Japan (1.839%) and

Singapore (1.777%) are the most volatile markets, while Taiwan (0.761%) and Thailand

(0.702%) are the least risky markets during the period under investigation.

The plots of daily returns for each market index from June 15, 2006 to April 27, 2012 are

displayed in Table 7.2. Visual inspection of the volatility of all series suggests that there are

certain periods, particularly during the GFC period, which have higher volatility and are

riskier than others. Exceptionally in the case of Thailand, the volatility of returns was

significantly higher during 2006 when the Thai coup d’e´tat took place.

In addition, these volatile periods seem to be concentrated and followed by periods of lower

volatility. This reflects the ARCH or volatility clustering effects that commonly appear in

financial time series. However, additional tests are discussed later that verify this

phenomenon in the REIT index series.

249

Table 7.1: Descriptive Statistics of Asia REIT Index Returns: June 15, 2006-April 27, 2012



Korea

Mean 0.009% 0.044% 0.056% 0.056% 0.059% 0.042% -0.005%

Median 0.021% 0.072% 0.012% 0.042% 0.000% 0.042% 0.024%

Maximum 10.59% 20.42% 10.08% 4.20% 7.60% 7.24% 13.09%

Minimum -11.82% -17.74% -13.25% -5.22% -4.20% -8.50% -14.28%

Std. Dev. 1.839% 1.777% 1.275% 0.802% 0.761% 0.702% 1.755%

Skewness -0.205 0.306 -0.735 -0.326 1.320 -1.076 -0.355

Jarque-Bera 1,914 27,794 15,502 985 16,786 42,571 4,182

Probability 0.000 0.000 0.000 0.000 0.000 0.000 0.000

The stationarity of all return series are examined using the Augmented Dickey-Fuller (ADF),

Phillips-Person (PP) and Kwiatkowski-Phillips-Schmidt-Shin (KPSS) tests. The test results in

Table 7.3 support that all time series are stationary and contain no unit roots. This means that

the mean and variance of the Asian REIT return series are constant over time. Therefore, the

series are suitable for further analysis of time-varying return and volatility transmissions.

250

Table 7.2: Plots of REIT Index Returns: June 15, 2006-April 27, 2012

Notes: Shaded areas imply the period during the Global Financial Crisis.

-.12

-.08

-.04

.00

.04

.08

.12

II III IV I II III IV I II III IV I II III IV I II III IV I II III IV I II

2006 2007 2008 2009 2010 2011

Japan

-.2

-.1

.0

.1

.2

.3


2006 2007 2008 2009 2010 2011

Singapore

-.15

-.10

-.05

.00

.05

.10

.15


2006 2007 2008 2009 2010 2011

Hong Kong

-.06

-.04

-.02

.00

.02

.04

.06


2006 2007 2008 2009 2010 2011

Malaysia

-.06

-.04

-.02

.00

.02

.04

.06

.08


2006 2007 2008 2009 2010 2011

Taiwan

-.12

-.08

-.04

.00

.04

.08


2006 2007 2008 2009 2010 2011

Thailand

-.15

-.10

-.05

.00

.05

.10

.15


2006 2007 2008 2009 2010 2011

South Korea

251

Table 7.3: Unit Root Tests for Asia REIT Return Series: June 15, 2006-April 27, 2012


ADF C -38.79* -48.47* -47.13* -52.23* -22.46* -51.79* -23.28*

C & T -38.79* -48.48* -47.13* -52.26* -22.62* -51.90* -23.34*

PP C -50.76* -48.48* -47.19* -51.44* -32.68* -51.73* -33.97*

C & T -50.76* -48.49* -47.19* -51.55* -32.61* -51.90* -34.01*

KPSS C 0.16 0.28 0.12 0.13 0.61* 0.63* 0.40*

C & T 0.08 0.14 0.08 0.12 0.09 0.06 0.07

Notes: C and T denote use of constant and time trend in these tests.

*Significant at the 5% level.

7.3 Regression Analysis

7.3.1 Volatility Clustering Effects

The results of the Ljung-Box (LB) and heteroscedasticity tests for up to 24 orders are

reported in Table 7.4. The LB -statistics indicate significant autocorrelations in all seven

return series. This justifies the inclusion of the lag terms in the mean equation. In addition,

there is strong evidence of ARCH effects in the residuals, as suggested from the ARCH test

results. This means that there is significant volatility clustering effects in the REIT return

indices. Large changes in REIT returns tend to be followed by large changes, while small

changes tend to be followed by small changes. This phenomenon is quite similar to

observations in previous studies in financial time series. The presence of time-varying

volatility in the series further affirms the suitability of an EGARCH model in analysing the

dynamic volatility of Asian REIT markets.

Table 7.4: Test for Serial Correlation and Heteroscedasticity: June 15, 2006-April 27, 2012


Q(4) 17.82* 8.81* 1.39* 3.68* 17.38* 11.00* 15.66*

Q(12) 39.71* 46.72* 15.63* 13.88* 30.44* 67.24* 46.14*

Q(24) 68.48* 86.17* 25.89* 42.58* 50.76* 97.69* 66.62*

Q2(4) 982.19* 440.02* 334.11* 185.17* 202.97* 51.48* 149.74*

Q2(12) 2,384.50* 583.48* 430.19* 205.11* 330.91* 123.13* 594.68*

Q2(24) 3,338.10* 695.68* 658.63* 267.37* 416.14* 164.78* 833.75*

ARCH(24) 39.62* 25.23* 17.44* 8.30* 10.82* 5.13* 23.37*

Notes: *Significant at the 5% level.

7.3.2 Cross-market Return and Volatility Spillovers

The estimated coefficients and standard errors for the conditional mean return equations are

presented in Panel A of Table 7.5. It can be seen that the returns of all Asian REIT indices,

except Malaysia, are highly correlated to their past values ( when ) and past

252

disturbance terms ( ), suggesting that historical innovations in these markets have strong

influence over expected returns. This explained the strong autocorrelation in all indices in the

preliminary analysis. These results indicate that Asian REIT markets are not fully efficient

and there are short-run momentums in REIT prices.

Results from the cross-market return spillovers indicate a strong influence of the developed

REIT markets over emerging markets. Singapore is the most efficient and influential market

in the region. As measured by the terms , changes in S-REIT prices significantly and

positively spread on to all regional REIT markets. The mean returns of Japan ( ) influence

the markets in Hong Kong and Taiwan substantially, while Hong Kong ( ) impacts Taiwan

and South Korea.

On the other hand, feedback effects from emerging markets to developed markets are limited.

In particular, Taiwan is the only emerging market that has return spillover effects to a

developed market, which is Hong Kong. The bidirectional interactions between Taiwan and

Hong Kong reflect the strong economic linkages between the two economies in the Greater

China Economic Area (GCEA). Among the emerging markets, there are no mean spillovers

between themselves, except for a unidirectional spillover from Malaysia to South Korea. In

general, it is more likely for mean returns to spill from developed REIT markets to emerging

markets than the other way around. This implies that investors could take advantage of price

innovations from leading REIT markets in the region to price REIT securities in other

markets in the region.

The parameter estimation for the second moment interdependencies can be seen in Panel B of

Table 7.5. Statistically significant values of the and parameters across all series confirm

the strong presence of ARCH and GARCH effects in the REIT indices. This implies that the

volatility of current REIT returns is related to the square of the previous innovations. Periods

of relative fluctuation in returns are followed by periods of relative calm. The coefficients

measuring asymmetry reveal that returns in most Asian REIT markets (i.e. Singapore,

Malaysia, Thailand and South Korea) experience asymmetry information effects. Negative

news has a much greater impact on REIT volatility than positive news.

253

In term of cross-market volatility spillovers, both Hong Kong and Singapore are the main

sources of volatility transmission to the other Asian REIT markets. The results indicate that

there is a high degree of volatility spillover from Hong Kong to Singapore, Malaysia,

Taiwan, Thailand and South Korea, while the volatility of Singapore transmits to Malaysia,

Taiwan and South Korea. On the other hand, volatility in Japan REITs is secluded and does

not influence any other markets. Interestingly, unlike return spillovers, which are often

unidirectional from the developed markets to the emerging markets, volatility spillovers tend

to be multidirectional. There are volatility feedback effects from smaller REIT markets to

more dominant REIT markets. The results show that Hong Kong received significant

volatility spillovers from Taiwan, Thailand and South Korea, despite the latter three being

emerging markets. In addition, shocks from South Korea REIT volatility transmitted to Japan

and Singapore. Among the emerging markets, unidirectional spillovers are detected from

Thailand to Taiwan and South Korea. Besides, there are significant bidirectional volatility

linkages between Taiwan and South Korea.

254

Table 7.5: EGARCH Model: Mean and Volatility Spillovers between the Asian REIT Markets: June 15, 2006-April 27, 2012


Panel A: Parameter estimation for conditional mean equations

α1 1.09 α2 3.88* α3 1.59 α4 4.82* α5 2.48* α6 8.32* α7 12.34*

β11 6.59* β21 1.71 β31 -2.13* β41 1.93 β51 -2.23* β61 0.51 β71 0.87

β12 7.67* β22 2.63* β32 4.28* β42 4.64* β52 3.95* β62 3.91* β72 4.56*

β13 -0.22 β23 -1.94 β33 -489.48* β43 -1.30 β53 1.99* β63 -1.07 β73 -2.64*

β14 1.26 β24 -0.19 β34 0.87 β44 -0.26 β54 0.53 β64 -1.10 β74 2.48*

β15 0.28 β25 0.56 β35 -2.96* β45 -0.92 β55 20.22* β65 -0.19 β75 -0.79

β16 1.49 β26 -1.73 β36 0.51 β46 1.57 β56 0.46 β66 332.05* β76 -1.70

β17 -0.62 β27 0.72 β37 1.42 β47 -0.84 β57 -0.27 β67 -0.43 β77 411.16*

γ1 -2.32* γ2 -3.35* γ3 -0.39 γ4 -3.12* γ5 -1.41 γ6 -5.17* γ7 -2.41*

φ1 -8.90* φ2 -2.20* φ3 4374898* φ4 0.14 φ5 -14.63* φ6 -345.37* φ7 -3842.78*

255



Panel B: Parameter estimation for conditional variance equations

δ1 -2.78* δ2 -5.87* δ3 -5.90* δ4 -5.84* δ5 -3.15* δ6 -4.02* δ7 -7.01*

ζ11 7.78* ζ21 9.31* ζ31 10.01* ζ41 8.21* ζ51 8.64* ζ61 13.73* ζ71 9.60*

ζ12 -2.81* ζ22 -2.83* ζ32 -0.55 ζ42 1.59 ζ52 -3.99* ζ62 -10.66* ζ72 3.88*

η1 -1.36 η2 -4.99* η3 0.07 η4 3.43* η5 1.94 η6 4.13* η7 -2.09*

θ11 4.12* θ21 4.36* θ31 3.93* θ41 2.87* θ51 9.92* θ61 12.77* θ71 4.02*

θ12 -0.47 θ22 4.33* θ32 10.17* θ42 9.25* θ52 -3.54* θ62 -3.31* θ72 38.00*

λ21 -0.67 λ31 1.48 λ41 -1.03 λ51 -1.75 λ61 1.23 λ71 1.31

λ12 -3.01*

λ32 0.63 λ42 -4.03* λ52 2.66* λ62 1.39 λ72 -2.90*

λ13 1.29 λ23 -4.12*

λ43 -3.47* λ53 -2.84* λ63 -3.25* λ73 4.52*

λ14 0.16 λ24 -0.71 λ34 -0.36

λ54 -1.01 λ64 -4.51* λ74 -1.53

λ15 -0.47 λ25 -0.77 λ35 5.44* λ45 -0.96

λ65 -0.93 λ75 -4.58*

λ16 -1.37 λ26 0.95 λ36 -2.17* λ46 0.83 λ56 2.20*

λ76 -2.15*

λ17 1.99* λ27 5.76* λ37 -3.25* λ47 1.21 λ57 -2.68* λ67 -1.19 π1 1.71 π2 3.27* π3 1.47 π4 -0.80 π5 2.73* π6 -2.99* π7 1.83

256



Panel C: Diagnostics on standardized residuals

Q(24) 22.87

29.35

18.43

28.12

25.78

39.71*

17.54

Q2(24) 24.67

21.29

167.91*

26.62

11.44

19.46

20.34

ARCH(24) 1.14

0.89

7.37*

1.12

0.51

0.75

0.78

Notes:

1 = Japan, 2 = Singapore, 3 = Hong Kong, 4 = Malaysia, 5 = Taiwan, 6 = Thailand, 7 = South Korea


257

7.3.3 Effects of the Global Financial Crisis

A dummy variable is incorporated to analyse the effects of the GFC on Asian REIT returns

and risk. The dummy variable takes the value one in the period from March 17, 2008 to March 6,

2009 and zero otherwise. While it is still unclear about the exact dates of the start and end of the

GFC, the March 17, 2008 is used as the beginning of the GFC in this study, since it was the date

Bear Stearns was taken over by J.P. Morgan. The March 6, 2009 is chosen as the end for the

GFC, because it was the date that the Asian REIT index rebounded strongly from its trough since

the start of the GFC.

The values of and measure the impact of the GFC on return (first moment) and volatility

(second moment) of every Asian REIT market respectively. Some interesting observations have

been observed. The results show that the impact of the GFC on Asian REIT returns and volatility

are quite different. While all markets experience significant reduction in returns, Asian REITs

have not experienced significant increases in volatility. Only two markets, Singapore and

Taiwan, had significantly higher volatility during the GFC. In contrast, Malaysia had lower risk

level during the GFC. Thailand also had lower volatility during the GFC, but the

statistic is not significant. These results highlight the robustness of Asian REIT markets during

the GFC.

7.3.4 Diagnostic Checks

Residual diagnostic tests summarized in Panel C of Table 7.5 show that the EGARCH models

satisfactorily capture the interactions between the seven Asian REIT markets. The Ljung-Box

statistics of up to 24 lags show no evidence of autocorrelation in the standardized residuals, with

all P-values greater than 0.05, with the exception of Thailand at . Besides, the

heteroscedasticity tests reject the presence of any ARCH effects in the residual series at the 0.05

significance levels. The diagnostic tests therefore support that the EGARCH models have

absorbed all autocorrelation, ARCH and GARCH effects present in the original return series.

7.4 Summary and Strategic Implications

The last decade has witnessed a rapid growth of the REIT sector across Asia. However, the body

of literature on REIT markets in Asia has been relatively modest as compared to that of the U.S.

258

and Australia. Within the Asian context, this study investigates the dynamic transmission of

REIT returns and volatility between seven countries in Asia: Japan, Singapore, Hong Kong,

Malaysia, Taiwan, Thailand and South Korea employing the EGARCH modification. In

addition, the effects of the GFC on REIT returns and volatility are also investigated using

dummy variables.

The data employed in this study are the daily closing prices for the total return REIT indices

from June 15, 2006 to April 27, 2012. The findings have important implications for international

investors in managing Asian REIT portfolios and developing trading and risk management

strategies.

The results from the mean spillover equations show that there is a strong tendency for REIT

returns to transmit from the developed markets to emerging markets. For example, lagged returns

from Singapore have significant spillover effects to other Asian REIT markets. The mean returns

of Japan influence the markets in Hong Kong and Taiwan substantially, while Hong Kong

impacts Taiwan and South Korea. This suggests that investors can take advantage of available

information from more dominant markets to predict movements of REIT returns in smaller

markets. In addition, investors should take into account past own innovations in developing

forecasting models of Asian REIT returns, as there are significant short-term price momentums

in Asian REIT markets.

On the other hand, the mechanism of volatility spillovers among Asian REIT markets appears to

be more multidirectional than that of return spillovers. Hong Kong and Singapore are the main

volatility emitters in the region. However, there are volatility feedback effects from the emerging

REIT markets to the developed REIT markets. The evidence of volatility linkages between Asian

REIT markets implies that investors can develop portfolio hedging strategies for Asian REITs.

The GFC analysis further indicated that while all markets experienced significant reductions in

returns, not all of them have experienced significant increases in volatility during the GFC. Some

markets even had lower risk during the GFC, such as Malaysia and Thailand. These results

highlight the robustness of the Asian REIT markets during the GFC.

259

8 CHAPTER 8

CONCLUSION

8.1 Introduction

Despite the growing significance of Asian REITs within both a domestic and global investment

context, empirical research on this topic has been thin and modest. This study has set out to

profile the risk and return characteristics of Asian REITs and their risk-adjusted performance in

comparison to the other asset classes of bonds, shares and property companies. The study has

also sought to examine whether Asian REITs provide diversification benefits to investors, as

well as the role of Asian REITs in a mixed-asset portfolio in both a local and international

context. Moreover, the impact of the GFC and the dynamic linkages between the Asian REIT

markets were also explored.

The analysis covered the seven Asian REIT markets; i.e. Japan, Singapore, Hong Kong,

Malaysia, Taiwan, Thailand and South Korea. The empirical results have been discussed and

summarised in the three main analysis chapters; Chapter 5: The Performance and Significance of

Asian REITs in Domestic Mixed-asset Portfolios, Chapter 6: The Significance of Asian REITs in

Regional and International Investment Portfolios and Chapter 7: The Dynamics of Returns and

Volatility in Asian REIT Markets. This chapter seeks to synthesise the major findings presented

in the body of the thesis in order to provide answers to the thesis research questions. In addition,

the theoretical and practical implications of the study will be highlighted. Lastly, the study

limitations and directions for future research will also be discussed.

8.2 Conclusion of Main Findings

8.2.1 Risk and Return Characteristics of Asian REITs

The empirical results from the individual country analyses indicated that Asian REITs delivered

better absolute returns than shares in most Asian markets. In particular, REITs outperformed

shares in Japan, Singapore, Hong Kong and Taiwan. The superior returns of Asian REITs over

shares reflect the high dividend yields of Asian REITs over this period. In terms of risk, Asian

REITs had lower volatility than shares in five (i.e. Hong Kong, Taiwan, Malaysia, Thailand and

260

South Korea) out of the seven markets. On average, Asian REIT returns are 18% less volatile

than share returns. This is attributed to the lower leverage levels of Asian REITs, which lead to

less sensitivity to interest rate fluctuations and financial risk.

When comparing Asian REITs to listed property companies for individual countries, the results

showed similar results to those between REITs and shares. REITs outperformed property

companies in most markets (i.e. Japan, Singapore, Hong Kong and Taiwan). Importantly, this

was achieved without taking on additional risk. The average volatility of Asian REITs was

equivalent to only 63% of the risk of property companies. One reason that may explain the high

volatility of property company returns is that most listed property companies in Asia are heavily

engaged in property development and non-property activities, making them more sensitive to

market risks. On the other hand, Asian REITs are more focused on income-producing property

with a stable income stream. The risk and return analysis also showed that Asian REITs share

more defensive characteristics with direct property than listed property companies. In contrast,

listed property companies are more like equities. This suggests that investors who seek exposure

to direct property through securitised vehicles should invest in Asian REITs rather than property

companies.

Within a pan-Asian context, the historical data suggested that the emerging Asian REIT markets

with the exception of South Korea tended to provide better returns, whilst having relatively

lower risk than the developed markets over the study period. These results were unexpected and

somehow contradicted the conventional perception that emerging markets should be riskier than

developed markets. The performance analysis of the UWS Emerging Asia REIT Index and the

UWS Developed Asia REIT Index also revealed similar observations. The emerging composite

Asia REIT index offered a superior return and risk profile than the developed Asia REIT index.

This can be explained by several reasons.

Firstly, the inferior returns of the developed Asian REIT markets over this period can be due to

the negative impact of the GFC and the ongoing sovereign debt issues in the US and Europe. On

the other hand, the emerging Asia REIT markets of Taiwan, Thailand and Malaysia are less

integrated into the global financial markets. Therefore, earnings of the emerging REITs are

261

largely driven by the local economic and political developments rather than international events.

One notable example was that of the Thai REIT market, which was much more volatile during

the 2006 Thai coup d'état than during the GFC period. Moreover, the emerging Asia REITs have

much lower leverage ratios than the developed Asia REITs, due to more restrictive regimes. The

average debt to net asset ratio of the emerging REITs was only 9% versus that of 35% for the

developed REITs in Asia as at May 2012 (APREA 2012). The low borrowing levels are

especially beneficial during the period of high uncertainly, in particular during the GFC. Another

reason for the low volatility of emerging Asia REIT indices was the low correlations between the

emerging REIT markets. Due to the low correlations, the overall performance of an emerging

REIT market portfolio can be relatively stable when investments were spread out over the

emerging markets.

8.2.2 Risk-adjusted Performance of Asian REITs

The risk-adjusted performance analysis for individual markets reported superior performance of

Asian REITs over shares in most of the Asian markets. In particular, stronger risk-adjusted

performance was seen for Japan, Singapore, Hong Kong and Taiwan, while Malaysia, Thailand

and South Korea REITs performed no better than their domestic stock markets. In comparison to

property companies, Asian REITs outperformed property companies across all markets on a risk-

adjusted basis, with an exception of South Korea REITs. This was due to Asian REITs delivering

better returns at lower volatility than property companies in most Asian markets.

Within a global context, Asian REITs also delivered strong risk-adjusted performance over the

study period. All of the composite Asia REIT indices; i.e. the pan-Asia REIT index, pan-Asian

excluding Japan REIT index, the developed Asia REIT index and the emerging REIT index,

outperformed their corresponding stock market indices. This was in contrast to the lacklustre

performance of the developed REIT markets of the US and Australia during this period. The

findings also revealed that all Asian REIT composite indices consistently outperformed both the

US and Australian REIT markets on a risk-adjusted basis. This highlights the sound investment

credentials of Asian REITs over this period. Low borrowing levels, steady earning streams and

favourable property market fundamentals have played important roles in negating the adverse

effects during periods of high instability in the global financial markets.

262

8.2.3 Diversification Benefits of Asian REITs

The empirical evidence from this study for individual markets showed that Asian REITs had low

or negative correlations with shares, property companies and bonds. The low correlations suggest

that Asian REIT returns do not move in tandem with the returns of other assets. As a result, the

inclusion of REITs will help smooth out the return fluctuations within a mixed-asset portfolio.

The average correlation between Asian REITs and shares was only 0.62 over the study period.

This is in contrast to the strong correlations observed between the returns of Asian property

companies and shares, averaging 0.82. Consequently, Asian REITs would provide more

diversification benefits for share investors than listed property companies. The low correlations

between Asian REIT and property companies (average r=0.62) also indicate that they are two

distinct asset classes and an inter-property diversification strategy is feasible. Asian REITs and

bonds are generally lowly or negatively correlated with each other, suggesting possible

diversification potential between them.

A detailed examination of the co-movements between the Asian markets discovered that the

Asian REIT markets were less integrated than the regional stock markets. The average

correlation coefficient between the Asian REIT markets was only 0.60, which was lower as

compared to that between the Asian equity markets of 0.71. The low correlations between the

Asian REIT markets suggest that investors can benefit from a pan-Asia REIT strategy, while

gaining access to the regional property markets. The lack of correlation between Asian REITs

has helped explain the lower volatility of the pan-Asia REIT index than the pan-Asia share index

over this period.

Extending the analysis to an international perspective, the study documented a low level of

correlation between the pan-Asia REIT indices with the international markets. The average

correlation between the pan-Asia indices and the S&P 500 index was only 0.64, with the lowest

figure being that between the emerging Asian REITs and US shares (r=0.49). This was in

contrast to the high correlations observed between pan-Asia shares and the US equity markets,

averaging 0.75 over the same period. Accordingly, US share investors would receive significant

portfolio enhancement by investing in Asian REITs. This was particularly more significant with

263

the emerging REIT markets of Malaysia, Taiwan, Thailand and South Korea, as they are less

integrated with the international markets.

8.2.4 The Role of Asian REITs in a Mixed-asset Portfolio

Asian REITs were found to offer attractive benefits to the investment portfolios in both the

domestic and international context. The asset allocation analysis for the local markets found that

including Asian REITs to a mixed-asset portfolio has increased the overall portfolio returns and

lowered portfolio risk. The results showed that when REITs were added to a mixed-asset

portfolio, the efficient frontier was raised upwards. On average, about one-third of the optimal

domestic mixed-asset portfolio should be allocated to REITs. Significant allocations to REITs

were targeted for Hong Kong (56.2%), Taiwan (54.3%), Singapore (53.0%) and Japan (50.1%),

reflecting the strong risk-adjusted performance of local REITs over shares and property

companies in these markets. Importantly, REITs were included in the efficient portfolios across

the risk spectrum meaning that all investors, from those who are seeking a defensive portfolio

strategy to those who are more growth-oriented and risk-tolerant, can find Asian REITs an

attractive asset class.

In regards to a pan-Asia REIT strategy, the study found that the optimal portfolio would consist

of mostly Thailand and Malaysia REITs at the low end of the risk spectrum, Taiwan and

Singapore REITs at the medium ranges and mostly Singapore REITs at the aggressive end of the

risk preference. For a global asset strategy, the analysis confirmed the significant role of pan-

Asia REITs in the asset mix. On average, the optimum portfolio would be composed of 61%

Asian REITs, 31% US shares and 8.2% Asian shares. US shares were more dominant at the

prudent risk level, then gradually reduced as risk levels escalated, allowing for the steadily

increased share of Asian REITs and Asian shares. At the high end of the risk spectrum, Asian

REITs accounted for the majority of the portfolio value.

8.2.5 Impact of the GFC on Asian REITs

The GFC during March 2008 and March 2009 has had a major impact on REIT markets across

Asia. The crisis has seen Asian REITs’ correlations with other asset classes increase

significantly, causing a significant loss of portfolio diversification benefits. Negative returns and

264

increased volatility were also evident across all markets. However, the study shown that the

impact of the GFC on Asian REITs was different between the developed and emerging REIT

markets. For the developed markets, REITs were affected as badly as shares and listed property

companies. However, for the emerging markets, the impact of the GFC on REITs was less severe

than on equities and property companies. This was due to the defensive characteristics of REITs

in the emerging markets with low leverage levels, steady income streams and a less aggressive

growth strategy.

The post-GFC period since April 2009 has seen Asian REITs gaining back some level of

diversification benefits, as the correlations with shares and property companies have been falling.

The strong recovery of Asian REIT performance has also been witnessed in both the return and

risk dimensions. Five out of seven Asian REIT markets have had better returns post-GFC than

the pre-GFC period. This was in comparison to only four of the Asian share markets and three of

the property security markets that have recovered from the GFC. In addition, most of Asian

REIT markets (4/7) have registered lower risk post-GFC versus the pre-GFC period. On the

other hand, the majority of share (6/7) and property company (5/7) markets still have higher risk

levels. This highlights the robustness of the Asian REIT market over this period of high

uncertainty.

8.2.6 Dynamic Linkages between the Asian REIT Markets

The econometric analysis of Asian REIT markets’ linkages has yielded a number of interesting

findings. The results from the mean spillover equations showed that there is a strong tendency

for REIT returns to transmit from the developed markets to the emerging markets in Asia. In

particular, lagged returns from Singapore have significant spillover effects to the other Asian

REIT markets. The mean returns of Japan influence the markets in Hong Kong and Taiwan

substantially, while Hong Kong impacts Taiwan and South Korea. This suggests that investors

can take advantage of available information from more dominant markets to predict movements

of REIT returns in smaller markets. In addition, investors should take into account past own

innovations in developing forecasting models of Asian REIT returns, as there were significant

short-term price momentum in the Asian REIT markets.

265

On the other hand, the mechanisms of volatility spillovers among Asian REIT markets were

found to be more multidirectional. The developed markets of Hong Kong and Singapore were

still the main volatility emitters in the region. However, volatility feedback effects from the

emerging REIT markets to the developed REIT markets have been detected as well. This

indicates that the emerging markets would also contain useful information to forecast the

volatility in the developed markets. In addition, the evidence of volatility linkages between Asian

REIT markets implies that investors can develop portfolio hedging strategies for Asian REITs by

holding assets that have adverse price and volatility movements.

8.3 Contribution of the Study

This study has provided an empirical analysis of the investment characteristics and risk-adjusted

performance of Asian REITs, as well as identifying the important role of Asian REITs in a

mixed-asset portfolio for both local and international property investors. The following section

discusses the practical and theoretical contributions and their implications for future research.

8.3.1 Practical Contributions

This dissertation has made four principal practical contributions to the discipline of property

investment. Firstly, the study has contributed to the body of knowledge by extending the existing

literature on new Asian REIT markets. In particular, this study was the first study that has

provided an empirical analysis on the performance of the emerging REIT markets of Thailand

and South Korea. The REIT indices for Thailand and South Korea REITs have been built to

provide REIT market benchmarks for these markets. In addition, the pan-Asia REIT indices were

constructed to track the performance of listed Asian REITs across the seven markets. Such

indices have not been available publicly prior to this study and will prove to be useful ongoing

research tools for future Asian REIT analyses.

Secondly, this thesis offers important insights to the investment characteristics of Asian REITs

and their role in both the domestic and international portfolios. Importantly, research studies

focusing on the significance of Asian REITs in a global investment portfolio were limited,

despite strong interest from global property investors. Moreover, most of the previous studies

focused on Asian property companies rather than Asian REITs. The optimal portfolio weights of

266

Asian REITs in domestic and global portfolios have been examined. This, together with the

analysis of a pan-Asia REIT strategy, should be of considerable interest to both academics and

practitioners; in particular institutional investors.

The third contribution of this study is to expand the data time horizon to as recent as April 2012

to cover the GFC period in order to assess its impact on Asian REIT performance. The sample

period has also made the analysis more balanced, as it includes both the bull and bear market

cycles. Previous studies on Asian property markets have been mostly done for the periods prior

to the GFC, which might have provided biased results. Findings from this analysis will provide

important risk management implications for property portfolio managers and policy makers.

Last but not least, a further contribution of this thesis to property knowledge is the application of

advanced econometric techniques to examine Asian REIT returns. Although volatility clustering

analysis has been popular in other disciplines such as economics and finance, the importance of

modelling the time-varying dynamics of property time series have not been emphasised in the

past. This study employs an ARCH model in modelling REIT returns at both the first and

second-order moment. As most of the previous works on market linkages focused on interactions

at the first-order moment, this study offers new directions and understanding by expanding the

analysis on the interactions at the second-order moments of Asian REITs.

8.3.2 Theoretical Contributions

There are two important theoretical contributions of this dissertation. The first one is the

synthesising of theories from different academic disciplines, i.e. property, finance and

econometrics. This research contributes to the further development of property investment as a

multidisciplinary area. In the course of this research, the author has extended the work of

Markowitz (1952) on portfolio theory to a new asset class, which is Asian REITs. This study has

reaffirmed the applicability of the modern portfolio theory to international property investment.

Another theoretical contribution of this thesis is the drawing together of econometric theories on

the uncertainty of financial prices to applied property research. This study has provided evidence

on the time-varying variance of REIT prices using an ARCH model. Since the introduction of the

267

ARCH model by Engle (1982), numerous research papers applying this model strategy to

economic and financial time series have appeared. However, this study was among a rare number

of studies that have used such a technique for the analysis of REITs in Asia.

8.4 Limitations of Study

Although a great effort has been made in this research to reach its aims and objectives, there are

factors that are beyond the author’s control that constraint the scope of the study. The first

shortcoming of the analysis is the short time series of Asian REIT markets. This is a common

challenge for studies on Asian property, as these markets only have a short history of existence.

Due to the short time period available, the performance results may not represent the long-term

characteristics of Asian REITs.

The second limitation is on the asset choices for investors. The portfolio analysis for individual

markets limits the asset menu only to the core assets such as shares, bonds and property

companies in addition to REITs. Direct property in these Asian markets has not been included

due to the lack of benchmark series for Asian direct property. For international investors, the

portfolio optimisation is constraint in the US and Asian markets only. Alternative assets such as

mortgage-backed securities (MBS), derivatives and financial assets in other regions have not

been included.

Another limitation of this thesis lies in the assumptions under the Modern Portfolio Theory

(MPT). MPT relies on the assumptions that risk is measured by standard deviation. However,

there has been some criticism on whether the standard deviation is a good measure of risk, as

investors’ risk perceptions are homogenous. In addition, there were other deficiencies in the

MPT framework, such as the absence of transaction costs and tax considerations. In addition, the

asset allocation process relies on static variance and covariance and reflects backward-looking

information.

Lastly, the return and volatility dynamic analysis was restricted only to the interactions among

the seven Asian REIT markets. The impact of outside markets such as the US, Australia and the

European markets has not been considered. Moreover, the EGARCH model did not contain

268

macroeconomic variables such as inflation, interest rates and GDP growth. It would be

interesting to explore the effects of these additional variables on Asian REIT markets.

8.5 Recommendations for Future Research

The suggestions of further research envisage the solutions to overcome the limitations of this

dissertation, as well as the areas that deserve further investigation. First of all, a revisit of the

analysis in this study is recommended as longer time series data for Asian REITs becomes

available in the future, allowing for the examination of long-term performance over different

investment cycles. Secondly, the extension of the analysis to the prospective REIT markets of

China and India and the frontier markets of Pakistan and Philippines, when these markets have

matured, would be of great interest to both academics and practitioners.

Thirdly, the portfolio analysis can be extended to include alternative assets such as mortgage-

backed securities (MBS), property derivatives, as well as financial assets and REITs in other

regions such as Europe, Latin America, the Middle East and Africa. Fourthly, to overcome the

limitations of the traditional MPT model, alternative asset allocation frameworks such as down-

side risk, dynamic conditional asset allocation and regime switching can be considered. Finally,

the extension of the EGARCH model to include additional variables such as US and Australia

markets as well as macroeconomic variables such as inflation, interest rates and GDP growth

would be beneficial.

These recommendations for future research demonstrate that this study can be explored and

extended in multiple directions. However, this dissertation has achieved its ultimate goal of

providing a rigorous and insightful empirical analysis on Asian REITs and offering original

research contributions to the existing body of knowledge in the disciplinary area of property

investment.

8.6 Final Comments

Asian REITs have been an exciting development in recent years, allowing investors an additional

opportunity for listed property exposure in their respective stock markets. This thesis has

assessed a range of property investment issues in a rigorous manner regarding the integrity and

269

performance of Asian REITs. Many of the results have reinforced the quality and investment

stature of Asian REITs. The calibre of the research in this thesis has been further externally

validated by several publications by the author in leading property research journals in the US

and Australia. The following journal papers and conference presentations have been generated

during this research project:

Journal Articles:







Conference Presentations:

1. Pham, AK, Newell, G, Wing, CK & Wong, SK 2012, 'The significance of a pan-Asia

REIT strategy', presentation at the Asian Real Estate Society (AsRES) and American

Real Estate and Urban Economics Association (AREUEA) Joint International

Conference, 07-10 July 2012, Singapore.


Asian REIT markets', presentation at the Pacific Rim Real Estate Society 18th Annual

Conference, 15-18 January 2012, Adelaide, Australia.


asset portfolio', presentation at the Asian Real Estate Society (AsRES) and American

Real Estate and Urban Economics Association (AREUEA) joint International

Conference, 11-14 July 2011, Jeju Island, Korea.

270

REFERENCES

Addae-Dapaah, K & Kion, CB 1996, 'International diversification of property stock: a

Singaporean investor's viewpoint', Real Estate Finance Journal, Vol. 13, No. 3, pp. 54-66.

Addae-Dapaah, K & Yong, GL 1998, 'Currency risk and office investment in Asia Pacific', Real

Estate Finance, Vol. 5, No. 1, pp. 67-85.

Ambrose, BW, Lee, DW & Peak, J 2007, 'Comovement after joining an index: spillovers

of nonfundamental effects', Real Estate Economics, Vol. 35, No. 1, pp. 57-90.

APREA 2012, Regional REITs Monthly April 2012, Asia Pacific Real Estate Association,

Singapore.

ARES 2011, ARES J-REIT Report May 2011, The Association for Real Estate Securitization.

ARES 2012, ARES J-REIT Report May 2012, The Association for Real Estate Securitization.

Arnold, H & Kavanaugh, D 1988, 'Risks and rewards of off-shore investment: London, Paris and

Sydney', PREA Quarterly, Vol. 1, No. 12, pp. 7-10.

Asabere, PK, Kleiman, R & McGowan, C 1991, 'The risk-return attributes of international real

estate equities', Journal of Real Estate Research, Vol. 6, No. 2, pp. 143-52.

Bajtelsmit, V & Worzala, E 1995, 'Real estate allocation in pension fund portfolios', Journal of

Real Estate Portfolio Management, Vol. 1, No. 1, pp. 1-14.

Ball, M, Lizieri, C & MacGregor, BD 1998, The Economics of Commercial Property Markets,

Routledge, London.

Barkham, R & Geltner, D 1994, 'Unsmoothing British valuation-based returns without assuming

an efficient market', Journal of Property Research, Vol. 11, No. 2, pp. 81-95.

Barkham, R & Geltner, D 1995, 'Price discovery in American and British property markets', Real

Estate Economics, Vol. 23, No. 1, pp. 21-44.

271

Barry, CB, Rodriguez, M & Lipscomb, JB 1996, 'Diversification potential from real estate

companies in emerging capital markets', Journal of Real Estate Portfolio Management, Vol. 2,

No. 2, pp. 107-18.

Benjamin, JD, Sirmans, GS & Zietz, EN 2001, 'Returns and risk on real estate and other

investments: more evidences', Journal of Real Estate Portfolio Management, Vol. 7, No. 3, pp.

183-214.

Bera, AK & Higgins, ML 1993, 'ARCH Model: properties, estimation and testing', Journal of

Economic Surveys, Vol. 7, No. 4, pp. 305-66.

Bigman, T 2002, Investing in international listed property companies, PREA Quarterly.

Bley, J & Olson, DO 2005, 'An analysis of relative return behavior: REITs vs. stocks', Academy

of Accounting and Financial Studies Journal, Vol. 9, No. 2, pp. 71-88.

Blundell, GF & Ward, CWR 1987, 'Property portfolio allocation: a multi-factor model', Land

Development Studies, Vol. 4, No. 2, pp. 145-56.

Bollerslev, T, Chou, RY & Kroner, KF 1992, 'ARCH modeling in finance: a review of the theory

and empirical evidence', Journal of Econometrics, Vol. 52, No. 1-2, pp. 5-59.

Bond, SA, Karolyi, GA & Sanders, AB 2003, 'International real estate returns: a multifactor,

multicountry approach', Real Estate Economics, Vol. 31, No. 3, pp. 481-500.

Brinson, GP, Diermeier, JJ & Schlarbaum, GG 1986, 'A composite portfolio benchmark for

pension plans', Financial Analysts Journal, Vol. 42, No. 2, pp. 15-24.

Brown, GR 1991, Property Investment and the Capital Markets, E & FN Spon, London.

Brown, GR & Chau, KW 1997, 'Excess returns in the Hong Kong commercial real estate

market', The Journal of Real Estate Research, Vol. 14, No. 2, pp. 91-106.

Brown, GR & Schuck, EJ 1996, 'Optimal portfolio allocations to real estate', Journal of Real

Estate Portfolio Management, Vol. 2, No. 1, pp. 63-73.

Brown, P 1990a, 'Australian residential price expectations and inflation', The Valuer, Vol. 29, pp.

124-8.

272

Brown, PM 1990b, 'United Kingdom residential price expectations and inflation', Land

Development Studies, Vol. 7, No. 1, pp. 57-67.

Brown, RJ & Li, LH 2000, 'A note on intracity geographic diversification of Real Estate

portfolios: evidence from Hong Kong', Journal of Real Estate Portfolio Management, Vol. 6,

No. 2, pp. 130-40.

Brueggeman, WB & Fisher, JD 1997, Real Estate Finance and Investment, Teenth Edn,

McGraw-Hill Irwin, Chicago.

Cannon, SE & Vogt, SC 1995, 'REITs and their management: an analysis of organizational

structure, performance and management compensation', Journal of Real Estate Research, Vol.

10, No. 3, pp. 297-318.

Case, B, Yang, Y & Yildirim, Y 2010, 'Dynamic correlations among asset classes: REIT; and

stock returns', DOI 10.1007/s11146-010-9239-2.

CBRE 2011, REITs around Asia 1H2011, CB Richard Ellis, Hong Kong.

CBRE 2012, Asia REIT View Point February 2012, CB Richard Ellis, Hong Kong.

CFA Institute 2011, Asia-Pacific REITs: Building Trust through Better Governance, CFA

Institute.

Chan, KC, Hendershott, PH & Sanders, AB 1990, 'Risk and return on real estate: evidence from

equity REITs', AREUEA Journal, Vol. 18, No. 4, pp. 431-52.

Chandrashekaran, V 1999, 'Time-series properties and diversification benefits of REIT returns',

Journal of Real Estate Research, Vol. 17, No. 1, pp. 91-112.

Chatrath, A & Liang, Y 1998, 'REITs and inflation: a long-run perspective', Journal of Real

Estate Research, Vol. 16, No. 3, pp. 311-25.

Chau, KW 1997, 'Political uncertainty and the real estate risk premiums in Hong Kong', Journal

of Real Estate Research, Vol. 13, No. 3, pp. 297-315.

273

Chau, KW, Newell, G & Wong, SK 2003, 'Performance of property companies in Hong Kong: a

style analysis approach', Journal of Real Estate Portfolio Management, Vol. 9, No. 1, pp. 29-44.

Chen, SJ, Hsieh, CH & Jordan, BD 1997, 'Real estate and the arbitrage pricing theory:

macrovariables vs. derived factors', Real Estate Economics, Vol. 25, No. 3, pp. 505-32.

Cheng, P, Ziobrowski, AJ, Caines, RW & Ziobrowski, B 1999, 'Uncertainty and foreign real

estate', Journal of Real Estate Research, Vol. 18, No. 3, pp. 463-80.

Cheok, SMC, Sing, TF & Tsai, IC 2011, 'Diversification as a value-adding strategy for Asian

REITs: a myth or reality?', International Real Estate Review, Vol. 14, No. 2, pp. 184-207.

Cheung, YL, Tsang, SK & Mak, SC 1995, 'The causal relationships between residential property

prices and rentals in Hong Kong: 1982-91', Journal of Real Estate Finance and Economics, Vol.

10, No. 1, pp. 23-35.

Chiang, KCH 2010, 'On the comovement of REIT prices', Journal of Real Estate Research, Vol.

32, No. 2, pp. 187-200.

Chiang, KCH, Kozhevnikov, K, Lee, ML & Wisen, CH 2006, 'REIT mimicking portfolio

analysis', International Real Estate Review, Vol. 9, No. 1, pp. 95-111.

Chiang, YH & Ganesan, S 1996, 'Property investment in Hong Kong', Journal of Real Estate

Portfolio Management, Vol. 2, No. 2, pp. 141-58.

Chiang, YH, So, CKJ & Tang, BS 2008, 'Time-varying performance of four Asia-Pacific REITs',

Journal of Property Investment & Finance, Vol. 26, No. 3, pp. 210-31.

Cho, H, Kawaguchi, Y & Shilling, JD 2003, 'Unsmoothing commercial property returns: a

revision to Fisher-Geltner-Webb's unsmoothing methodology', Journal of Real Estate Finance

and Economics, Vol. 27, No. 3, pp. 393-405.

Chu, YQ & Sing, TF 2004, 'Inflation hedging characteristics of the Chinese real estate market',

Journal of Real Estate Portfolio Management, Vol. 10, No. 2, pp. 145-54.

Chua, A 1999, 'The role of international real estate in global mixed-asset investment portfolios',


274

Chui, ACW, Wei, KCJ & Titman, S 2003a, 'The crosssection of expected REIT returns', Real


Chui, ACW, Wei, KCJ & Titman, S 2003b, 'Intra-industry momentum: the case of REITs',

Journal of Financial Markets, Vol. 6, No. 3, pp. 363-87.

Clayton, J & MacKinnon, G 2001, 'The time-varying nature of the link between REIT, real estate

andfinancial asset returns', Journal of Real Estate Portfolio Management, Vol. 7, No. 1, pp. 43-5.

Clayton, J & MacKinnon, G 2003, 'The relative importance of stock, bond and real estate factors

in explaining REIT returns', Journal of Real Estate Finance and Economics, Vol. 27, No. 1, pp.

39-60.

Clifford Chance 2010, Introduction to REITs in China, Clifford Chance, Bejing.

Cole, RA 1988, 'A new look at commercial real estate returns', PhD thesis, The University of

North Carolina at Chapel Hill.

Coleman, M, Hudson-Wilson, S & Webb, JR 1994, 'Real estate in the multi-asset portfolio', in S

Hudson-Wilson & CH Wurtzebach (Eds), Managing real estate portfolios, Irwin, Burr Ridge,

pp. 98-123.

Conover, CM, Friday, HS & Sirmans, GS 2002, 'Diversification benefits from foreign real estate

investments', Journal of Real Estate Portfolio Management, Vol. 8, No. 1, pp. 17-25.

Cooperman, LG, Einhorn, SG & Melnikoff, M 1984, Property revisited: the role of property in

pension fund investments, Goldman Sachs, New York.

Corgel, JB, Smith, HC & Ling, DC 1998, Real estate perspectives :an introduction to real estate,

Fourth Edn, Irwin/McGraw-Hill, Boston.

Cotter, J & Stevenson, S 2006, 'Multivariate modeling of daily REIT volatility', Journal of Real

Estate Finance and Economics, Vol. 32, No. 3, pp. 302-25.

D'Arcy, É & Lee, S 1998, 'A real estate portfolio strategy for Europe: a review of the options',


275

Davidson, R & MacKinnon, JG 1993, Estimation and Inference in Econometrics, Oxford

University Press, New York.

Deloitte 2012, Japan Highlights 2012, Deloitte.

Dokko, Y, Edelstein, RH, Pomer, M & Urdang, ES 1991, 'Determinants of the rate of return for

nonresidential real estate: inflation expectations and market adjustment lags', Real Estate

Economics, Vol. 19, No. 1, pp. 52-69.

Edelstein, R & Quan, D 2006, 'How does appraisal smoothing bias real estate returns

measurement?', Journal of Real Estate Finance and Economics, Vol. 32, No. 1, pp. 41-60.

Eichholtz, PMA 1996, 'The stability of the covariance of international property share returns',


Eichholtz, PMA 1997, 'Real estate securities and common stocks: a first international look', Real


Eichholtz, PMA & Koedijk, KG 1996, 'The global real estate securities market', Real Estate

Finance, Vol. 13, No. 1, pp. 76-82.

Embrechts, P, McNeil, A & Straumann, D 2002, 'Correlation and dependency in risk

management: properties and pitfalls', in M Dempster (Ed.), Risk Management: Value at risk and

beyond, Cambridge University Press.

Engle, RF & Ng, VK 1993, 'Measuring and testing the impact of news on volatility', Journal of

Finance, Vol. 48, No. 5, pp. 1749-78.

Ennis, RM & Burik, P 1991, 'Pension fund real estate investment under a simple equilibrium

pricing model', Financial Analysts Journal, Vol. 47, No. 3, pp. 20-30.

EPRA 2011, Global REIT Survey 2011, European Public Real Estate Association.

Eves, C 2002, 'The role of rural land in mixed asset investment portfolios', in The 8th

Pacific Rim

Real Estate Society Conference, Christchurch, New Zealand.

Fama, EF & Schwert, GW 1977, 'Asset returns and inflation', Journal of Financial Economics,

Vol. 5, No. 2, pp. 115-46.

276

Fang, WS, Wang, KM & Nguyen, TBT 2008, 'Is real estate really an inflation hedge? Evidence

from Taiwan', Asian Economic Journal, Vol. 22, No. 2, pp. 209-24.

Fei, P, Ding, L & Deng, Y 2010, 'Correlation and volatility dynamics in REIT returns:

performance and portfolio considerations', Journal of Portfolio Management, Vol. 36, No. 2, pp.

113-25.

Feng, Z, Ghosh, C & Sirmans, CF 2006, 'Changes in REIT stock prices, trading volume and

institutional ownership resulting from S&P REIT index changes', Journal of Real Estate


Firstenberg, PM, Ross, SA & Zisler, RC 1988, 'Real estate: the whole story', in PL Bernstein &

FJ Fabozzi (Eds), Journal of Portfolio Management, pp. 22-34.

Fisher, JD, Geltner, DM & Webb, RB 1994, 'Value indices of commercial real estate: a

comparison of index construction methods', Journal of Real Estate Finance and Economics, Vol.

9, No. 2, pp. 137-64.

Fogler, HR 1984, 'Twenty percent in real estate: can theory justify it?', Journal of Portfolio

Management, Vol. 10, No. 2, pp. 6-13.

Fraser, W 1985, 'The risk of property to the institutional investor', Journal of Property Valuation

and Investment, Vol. 4, No. 1, pp. 45-59.

Fu, YL, W KLo, Wai Chung 1993, The dynamics of residential property markets and the stock

market in Hong Kong, City Polytechnic of Hong Kong, Kowloon, Hong Kong.

Ganesan, S & Chiang, YH 1998, 'The inflation-hedging characteristics of real and financial

assets in Hong Kong', Journal of Real Estate Portfolio Management, Vol. 4, No. 1, pp. 55-67.

Garvey, R, Santry, G & Stevenson, S 2001, 'The linkages between real estate securities in the

Asia Pacific', Pacific Rim Property Research Journal, Vol. 7, No. 4, pp. 240-58.

Gatzlaff, D & Geltner, D 1998, 'A transaction-based index of commercial property and its

comparison to the NCREIF index', Real Estate Finance, Vol. 15, No. 1, pp. 7-22.

277

Geltner, D 1989a, 'Bias in appraisal-based returns', Real Estate Economics, Vol. 17, No. 3, pp.

338-52.

Geltner, D 1989b, 'Estimating real estate's systematic risk from aggregate level appraisal-based

returns', Real Estate Economics, Vol. 17, No. 4, pp. 463-81.

Geltner, D 1991, 'Smoothing in appraisal-based returns', Journal of Real Estate Finance and


Geltner, D 1993, 'Temporal aggregation in real estate return indices', Journal of the American

Real Estate and Urban Economic Association, Vol. 21, No. 2, pp. 141-66.

Geltner, D & Goetzmann, W 2000, 'Two decades of commercial property returns: a repeated-

measures regression-based version of the NCREIF index', Journal of Real Estate Finance and


Geltner, DM, Miller, NG, Clayton, J & Eichholtz, P 2001, Commercial real estate analysis and

investments, Second Edn, Cengage Learning, Mason, OH, USA.

Ghosh, C, Miles, M & Sirmans, CF 1996, 'Are REITs stocks?', Real Estate Finance, Vol. 13, No.

3, pp. 46-53.

Giliberto, SM 1988, 'A note on the use of appraisal data in indexes of performance

measurement', Real Estate Economics, Vol. 16, No. 1, pp. 77-83.

Giliberto, SM 1989, Real estate vs. financial assets: an updated comparison of returns in the

United States and the United Kingdom, Salomon Brothers, Inc, New York.

Giliberto, SM 1990, 'Equity real estate investment trusts and real estate returns', Journal of Real


Giliberto, SM 1992, Real estate risk and return: 1991 survey results, Salomon Brothers, New

York.

Glascock, JL, Lu, CL & So, RW 2000, 'Further evidence on the integration of REIT, bond and

stock returns', Journal of Real Estate Finance and Economics, Vol. 20, No. 2, pp. 177-94.

278

Glascock, JL, Lu, CL & So, RW 2002, 'REIT returns and inflation: perverse or reverse causality

effects?', Journal of Real Estate Finance and Economics, Vol. 24, No. 3, pp. 301-17.

Goebel, PH & Kim, KS 1989, 'Performance evaluation of finite-life real estate investment trusts',


Gold, RB 1985, Real estate: can insitutional portfolios be diversified without it?, JMB

Institutional Realty Corporation, Chicago.

Gordon, JN 1991, 'The diversification potential of international property investments', Real


Gordon, JN & Canter, TA 1999, 'International real estate securities: a test for capital markets

intergration', Journal of Real Estate Portfolio Management, Vol. 5, No. 2, pp. 161-70.

Gordon, JN, Canter, TA & Webb, JR 1998, 'The effects of international real estate securities on

portfolio diversification', Journal of Real Estate Portfolio Management, Vol. 4, No. 2, pp. 83-92.

Gyourko, J & Keim, DB 1992, 'What does the stock market tell us about real estate returns?',

Real Estate Economics, Vol. 20, No. 3, pp. 457-85.

Gyourko, J & Linneman, P 1988, 'Owner-occupied homes, income-producing properties, and

REITs as inflation hedges: empirical findings', Journal of Real Estate Finance and Economics,

Vol. 1, No. 4, pp. 347-72.

Hamelink, F & Hoesli, M 1996, 'Swiss real estate as a hedge against inflation: new evidence

using hedonic and autoregressive models', Journal of Property Finance, Vol. 7, No. 1, pp. 33-49.

Hamilton, JD 1994, Time Series Analysis, Princeton University Press, Princeton.

Han, J & Liang, Y 1995, 'The historical performance of real estate investment trusts', Journal of

Real Estate Research, Vol. 10, No. 3, pp. 235-62.

Hartzell, D 1986, Real estate in the portfolio, Salomon Brothers, New York.

Hartzell, D, Hekman, J & Miles, M 1986, 'Diversification categories in investment real estate',


279

Hartzell, D, Hekman, JS & Miles, ME 1987, 'Real estate returns and inflation', AREUEA

Journal, Vol. 15, No. 1, pp. 617-37.

Hasting, A & Nordby, H 2007, 'Benefits of global diversification on a real estate portfolio',

Journal of Portfolio Management, Vol. 33, pp. 53-62.

Hayashi, F 2000, Econometrics, Princeton University Press, Princeton.

Hoag, JW 1980, 'Toward indices of real estate value and return', Journal of Finance, Vol. 35,

No. 2, pp. 569-80.

Hoesli, M 1994, 'Real estate as a hedge against inflation: lessons from the Swiss case', Journal of

Property Valuation and Investment, Vol. 12, No. 3, pp. 51-9.

Hoesli, M & MacGregor, BD 2000, Property Investment Principles and Practice of Portfolio

Management, Longman.

Hoesli, M, MacGregor, BD, Matysiak, G & Nanthakumaran, N 1997, 'The short-term inflation-

hedging characteristics of U.K. real estate', Journal of Real Estate Finance and Economics, Vol.

15, No. 1, pp. 27-57.

Hoesli, MH, Lekander, J & Witkiewicz, W 2002, 'International evidence on real estate as a

portfolio diversifier', Journal of Real Estate Research, Vol. 26, No. 2, p. 161.

Huang, JZ & Zhong, ZDK 2006, 'Time variation in diversification benefits of commodity,

REITs, and TIPS', paper presented to 2006 China International Conference in Finance, Xian,

China.

Ibbotson, RG & Siegel, LB 1984, 'Real estate returns: a comparison with other investments',


IMF 2012, World Economic Outlook Databases, International Monetary Fund, viewed 30 June

2012, <http://www.imf.org/external/ns/cs.aspx?id=28>.

Irwin, SH & Landa, D 1987, 'Real estate, futures and gold as portfolio assets', Journal of


http://www.imf.org/external/ns/cs.aspx?id=28%3e

280

JLL 2011, Asia Pacific Property Digest - Q2 2011, Jones Lang LaSalle.

JLL 2012, Real Estate Transparency: Back on Track, Jones Lang LaSalle.

Jones Lang Wootton 1995, Direct and indirect property: close relation?, Jones Lang Wootton

Research, Sydney.

Kallberg, JG, Liu, CH & Greig, DW 1996, 'The role of real estate in the portfolio allocation

process', Real Estate Economics, Vol. 24, No. 3, pp. 359-77.

Khoo, T, Hartzell, D & Hoesli, M 1993, 'An investigation of the change in real estate investment

trusts betas', Journal of the American Real Estate and Urban Economic Association, Vol. 21, No.

2, pp. 107-30.

Kim, BS 2009, 'The impact of the global financial crisis on Asia-Pacific real estate markets:

evidence from Korea, Japan, Australia and US REITs', Pacific Rim Property Research Journal,

Vol. 15, No. 4, pp. 398-416.

Kleiman, R & Farragher, E 1992, 'Investment characteristics of international real estate equities',

Real Estate Review, Vol. 22, No. 3, pp. 33-9.

Kundus, SS & Sing, TF 2011, 'Interest alignment and insider shareholdings in the emerging

Asian REIT markets', Journal of Real Estate Portfolio Management, Vol. 17, No. 2, pp. 127-37.

Kutsuna, K, Dimovski, W & Brooks, R 2008, 'The pricing and underwriting cost of Japanese

REIT IPOs', Journal of Property Research, Vol. 25, No. 3, pp. 221-39.

Larsen, AB & McQueen, GR 1995, 'REITs, real estate, and inflation: lessons from the gold

market', Journal of Real Estate Finance and Economics, Vol. 10, No. 3, pp. 285-97.

Le Moigne, C & Viveiros, É 2008, 'The inflation-hedging ability of Canadian direct real estate

return and its components, income, and appreciation', Journal of Real Estate Portfolio


Lee, CL & Ting, KH 2009, 'The role of Malaysian securitised real estate in a mixed-asset

portfolio', Journal of Financial Management of Property and Construction, Vol. 14, No. 3, pp.

208-30.

281

Lee, CL & Ting, KH 2011, 'Linkages between Malaysian housing prices, property companies

and stocks', Pacific Rim Property Research Journal, Vol. 17, No. 2, pp. 287-312.

Lee, ML & Chiang, K 2010, 'Long-run behaviour of equity REITs: become more like common

stocks after the early 1990s?', Journal of Property Investment & Finance, Vol. 28, No. 6, pp.

454-65.

Lee, ML, Kuo, SH, Lee, MT & Lin, CW 2011a, 'Market signals associated with Taiwan REIT

IPOs: reactions of non-REIT real estate stocks', Journal of Real Estate Literature, Vol. 19, No. 1,

pp. 93-110.

Lee, MT & Lee, ML 2011, 'Long-run inflation-hedging properties of United State equity real

estate investment trusts (REITs): before and after the structural break in the 1990s', African

Journal of Business Management, Vol. 6, No. 6, pp. 2162-8.

Lee, MT, Lee, ML, Lai, FT & Yang, TH 2011b, 'Do real estate stocks hedge inflation in the long

run? Evidence from three East Asian emerging markets', Journal of Real Estate Literature, Vol.

19, No. 2, pp. 347-72.

Lee, S, Byrne, P & French, N 1996, 'Assessing the future of property in the multi-asset portfolio:

the case of UK pension funds', Journal of Property Research, Vol. 13, No. 3, pp. 197-209.

Li, DD & Yung, K 2007, 'REIT returns: between the Pacific and the Atlantic', Review of

Accounting and Finance, Vol. 6, No. 4, pp. 353-69.

Liang, Y, Chatrath, A & McIntosh, W 1996, 'Apartment REITs and apartment real estate', The


Lim, LC, McGreal, S & Webb, JR 2008, 'Institutional benchmarks for international real estate

investment', Journal of Real Estate Portfolio Management, Vol. 14, No. 2, pp. 93-104.

Limmack, RJ & Ward, CWR 1988, 'Property returns and inflation', Land Development Studies,

Vol. 5, No. 1, pp. 47-55.

Ling, DC & Naranjo, A 1997, 'Economic risk factors and commercial real estate returns', The

Journal of Real Estate Finance and Economics, Vol. 14, No. 3, pp. 283-307.

282

Ling, DC & Naranjo, A 1999, 'The integration of commercial real estate markets and stock

markets', Real Estate Economics, Vol. 27, No. 3, pp. 483-516.

Ling, DC, Naranjo, A & Ryngaert, MD 2000, 'The predictability of equity REIT returns: time

variation and economic significance', Journal of Real Estate Finance and Economics, Vol. 20,

No. 2, pp. 117-36.

Liow, KH 1997, 'The historical performance of Singapore property stocks', Journal of Property


Liow, KH 1998, 'Relationship between Singapore stock and real estate returns', Journal of Real

Estate and Construction, Vol. 8, pp. 1-16.

Liow, KH 2000, 'The dynamics of the Singapore commercial property market', Journal of

Property Research, Vol. 17, No. 4, pp. 279-92.

Liow, KH 2001, 'The abnormal return performance of Singapore property companies', Pacific


Liow, KH 2007, 'The dynamics of return volatility and systematic risk in international real estate

security markets', Journal of Property Research, Vol. 24, No. 1, pp. 1-29.

Liow, KH 2011, 'Co-movements and correlations across Asian securitized real estate and stock

markets', Real Estate Economics, Vol. 40, No. 1, pp. 97-129.

Liow, KH & Adair, A 2009, 'Do Asian real estate companies add value to investment portfolio?',


Liow, KH, Ho, KHD, Ibrahim, MF & Chen, Z 2009, 'Correlation and volatility dynamics in

international real estate securities markets', Journal of Real Estate Finance and Economics, Vol.

39, No. 2, pp. 202-23.

Liow, KH & Sim, MC 2006, 'The risk and return profile of Asian real estate stocks', Pacific Rim

Property Research Journal, Vol. 12, No. 3, pp. 283-310.

Liu, CH & Mei, J 1998, 'The predictability of international real estate markets, exchange risks

and diversification consequences', Real Estate Economics, Vol. 26, No. 1, pp. 3-40.

283

MacGregor, BD & Nanthakumaran, N 1992, 'The allocation to property in the multi-asset

portfolio: the evidence and theory reconsidered', Journal of Property Research, Vol. 9, No. 1, pp.

5-32.

MacKinnon, GH & Al Zaman, A 2009, 'Real estate for the long term: the effect of return

predictability on long-horizon allocations', Real Estate Economics, Vol. 37, No. 1, pp. 117-53.

Macquarie 2011, Global Property Insight, Macquarie Equities Research.

Markowitz, H 1952, 'Portfolio Selection', The Journal of Finance, Vol. 7, No. 1, pp. 77-91.

Marks, E 1986, 'Real rates of returns among foreign investors in US real estate', Real Estate

Finance Journal, Vol. 1, No. 3, pp. 56-61.

Maroney, N & Naka, A 2006, 'Diversification benefits of Japanese real estate over the last four

decades', Journal of Real Estate Finance and Economics, Vol. 33, No. 3, pp. 259-74.

Maurer, R & Reiner, F 2002, 'International asset allocation with real estate securities in a

shortfall risk framework: the view point of German and US investors', Journal of Real Estate


Miles, D 1996, 'Property and inflation', Journal of Property Finance, Vol. 7, No. 1, pp. 21-32.

Mueller, AG & Mueller, GR 2003, 'Public and private real estate in a mixed-asset portfolio',


Mueller, GR & Laposa, SP 1996, 'REIT returns: a property-type perspective', Real Estate


Mueller, GR, Pauley, KR & Morrill, WKJ 1994, 'Should REITs be included in a mixed-asset

portfolio?', Real Estate Finance, Vol. 11, No. 1, pp. 23-8.

Mull, SR & Soenen, LA 1997, 'U.S. REITs as an asset class in international investment

portfolios', Financial Analysts Journal, Vol. 53, No. 2, pp. 55-61.

Murphy, JA & Kleinman, RT 1989, 'The inflation-hedging characteristics of equity REITs: an

empirical study', Quarterly Review of Economics and Business, Vol. 29, pp. 95-101.

284

Myer, FCN & Webb, JR 1994a, 'Statistical properties of returns: financial assets versus

commercial real estate', Journal of Real Estate Finance and Economics, Vol. 8, No. 3, pp. 267-

82.

Myer, NFC & Webb, JR 1993, 'Return properties of equity REITs, common stocks, and

commercial real estate: a comparison', Journal of Real Estate Research, Vol. 8, No. 1, pp. 87-

106.

Myer, NFC & Webb, JR 1994b, 'Retail stocks, retail REITs and retail real estate', Journal of Real


NAREIT 2012, FTSE NAREIT US Real Estate Index Historical Values & Returns, National

Association of Real Estate Investment Trusts, viewed 04 July 2012,

<http://www.reit.com/DataAndResearch/IndexData/FNUS-Historical-Data.aspx>.

Newell, G 1995a, 'Inflation-hedging attributes of Australian commercial property', Australian

Land Economics Review, Vol. 1, pp. 31-7.

Newell, G 1995b, 'Is Canadian real estate a hedge against inflation?', The Canadian Appraiser,

Vol. 39, pp. 25-7.

Newell, G 1996, 'The inflation-hedging characteristics of Australian commercial property: 1984-

1995', Journal of Property Finance, Vol. 7, No. 1, pp. 6-20.

Newell, G 2008, 'PPA lecture notes', University of Western Sydney.

Newell, G 2012, The investment characteristics and benefits of Asian REITs for retail investors,

Asia Pacific Real Estate Association (APREA).

Newell, G & Boyd, T 1995, 'Inflation-hedging attributes of New Zealand commercial property',

New Zealand Valuers’ Journal, No. December, pp. 50-3.

Newell, G, Chan, KW & Wong, SK 2004, 'The level of direct property in Hong Kong property

company performance', Journal of Property Investment & Finance, Vol. 22, No. 6, pp. 512-32.

Newell, G & Chau, KW 1996, 'Linkages between direct and indirect property performance in

Hong Kong', Journal of Property Finance, Vol. 7, No. 4, pp. 9-29.

http://www.reit.com/DataAndResearch/IndexData/FNUS-Historical-Data.aspx%3e

285

Newell, G, Chau, KW & Pretorius, F 1996, 'Adjusting the volatility of the Hong Kong property

market', Journal of Real Estate and Construction, Vol. 6, No. 1, pp. 1-16.

Newell, G, Chau, KW & Wong, SK 2009a, 'The significance of Chinese commercial property in

an Asian property portfolio', Journal of Property Investment & Finance, Vol. 27, No. 2, pp. 102-

19.

Newell, G, Chau, KW, Wong, SK & McKinnell, K 2007, 'Factors influencing the performance of

Hong Kong real estate companies', Journal of Real Estate Portfolio Management, Vol. 13, No. 1,

pp. 75-86.

Newell, G & Kamineni, R 2007, 'The significance and performance of real estate markets in

India', Journal of Real Estate Portfolio Management, Vol. 13, No. 2, pp. 161-72.

Newell, G & MacFarlane, J 1995a, 'Improved risk estimation using appraisal smoothed real

estate returns', Journal of Real Estate Portfolio Management, Vol. 1, No. 1, pp. 51-7.

Newell, G & MacFarlane, J 1995b, 'Linkages between property trust performance and property

market returns', in 1st International Real Estate Society Conference, Stockholm.

Newell, G & MacFarlane, J 1996, 'Risk estimation and appraisal-smoothing in the UK property

returns', Journal of Property Research, Vol. 13, No. 1, pp. 1-12.

Newell, G & Osmadi, A 2009, 'The development and preliminary performance analysis of

Islamic REITs in Malaysia', Journal of Property Research, Vol. 26, No. 4, pp. 329-47.

Newell, G & Osmadi, A 2010, 'Assessing the importance of factors influencing the future

development of REITs in Malaysia', Pacific Rim Property Research Journal, Vol. 16, No. 3, pp.

358-74.

Newell, G & Peng, HW 2012, 'The significance and performance of Japan REITs in a mixed-

asset portfolios', Pacfific Rim Property Research Journal, Vol. 18, No. 1, pp. 21-34.

Newell, G & Webb, JR 1996, 'Assessing risk for international real estate investments', Journal of

Real Estate Research, Vol. 11, No. 2, pp. 103-16.

286

Newell, G & Webb, JR 1998, 'Real estate performance benchmarks in New Zealand and South

Africa', Journal of Real Estate Literature, Vol. 6, No. 2, pp. 137-43.

Newell, G, Wing, CK, Kei, WS & Hiang, LK 2009b, 'The significance and performance of

property securities markets in the Asian IFCs', Journal of Property Research, Vol. 26, No. 2, pp.

125–48.

Newell, G & Worzala, E 1995, 'The role of international property in investment portfolio',

Journal of Property Finance, Vol. 6, No. 1, pp. 55-63.

Newell, G, Wu, Y, Chau, KW & Wong, SK 2010, 'The development and performance of REITs

in Hong Kong', Pacific Rim Property Research Journal, Vol. 16, No. 2, pp. 191-206.

Nguyen, TK 2010, 'The significance and performance of listed property companies in Vietnam',

Pacfific Rim Property Research Journal, Vol. 16, No. 2, pp. 221-45.

Nguyen, TK 2011, 'The significance and performance of listed property companies in The

Phillipines', Pacific Rim Property Research Journal, Vol. 7, No. 2, pp. 260-86.

Nguyen, TK 2012, 'The volatility spillovers in Asian listed property companies in developed and

emerging markets', Pacific Rim Property Research Journal, Vol. 18, No. 1, pp. 49-65.

Norman, EJ, Sirmans, SG & Benjamin, JD 1995, 'The historical environment of real estate

returns', Journal of Real Estate Portfolio Management, Vol. 1, No. 1, pp. 1-24.

Oikarinen, E, Hoesli, M & Serrano, C 2011, 'The long-run dynamics between direct and

securitized real estate', Journal of Real Estate Research, Vol. 33, No. 1, pp. 73-103.

Ong, SE 1994, 'Structural and vector autoregressive approaches to modelling real estate and

property stock prices in Singapore', Journal of Property Finance, Vol. 5, No. 4, pp. 4-18.

Ong, SE 1995, 'Singapore real estate and property stocks - a co-integration test', Journal of

Property Research, Vol. 12, No. 1, pp. 29-39.

Ong, SE, Ooi, JTL & Kawaguichi, Y 2011, 'Seasoned equity issuance by Japan and Singapore

REITs', Journal of Real Estate Finance and Economics, Vol. 43, No. 1, pp. 205-20.

287

Ooi, JTL, Newell, G & Sing, TF 2006, 'The growth of REIT markets in Asia', Journal of Real

Estate Literature, Vol. 14, No. 2, pp. 203-22.

Pagan, A 1996, 'The econometrics of financial markets', Journal of Empirical Finance, Vol. 3,

No. 1, pp. 15-102.

Park, JY, Mullineaux, DJ & Chew, I-K 1990, 'Are REITs inflation hedges?', The Journal of Real

Estate Finance and Economics, Vol. 3, No. 1, pp. 91-103.

Peng, L & Schulz, R 2012, 'Does the diversification potential of securitized real estate vary over

time and should investors care?', Journal of Real Estate Finance and Economics, DOI

10.1007/s11146-011-9357-5.

Pham, AK 2011a, 'The performance of Thai-REITs in a mixed-asset portfolio', Pacific Rim

Property Research Journal, Vol. 17, No. 2, pp. 197-214.

Pham, AK 2011b, 'The significance and performance of South Korean REITs in a mixed-asset

portfolio', Journal of Real Estate Literature, Vol. 19, No. 2, pp. 373-90.

Pham, AK 2012, 'The dynamics of return and volatility in the emerging and developed Asian

REIT markets', Journal of Real Estate Literature, Vol. 20, No. 1, pp. 79-96.

Phillips, PCB & Perron, P 1988, 'Testing for a unit root in time series regression', Biometrika,

Vol. 75, No. 2, pp. 335–46.

Pierzak, E 2001, Exploring international property securities for us investors, Henderson Global

Investors Property Economics & Research.

Pramerica 2012, A Bird's Eye View of Global Real Estate Markets: 2012 Update, Pramerica.

Quek, MCH & Ong, SE 2008, 'Securitising China real estate: a tale of two China-centric REITs',


Reid, I 1989, 'Creating a global real estate investment strategy', Money Management Forum, pp.

101-4.

Robichek, AA, Cohn, RA & Pringle, JJ 1972, 'Returns on alternative investment media and

implications for portfolio construction', The Journal of Business, Vol. 45, No. 3, pp. 427-43.

288

Ross, SA & Webb, JR 1985, 'Diversification and risk in international real property investments:

an empirical study', International Property Investment Journal, Vol. 2, No. 3, pp. 359-77.

Ross, SA & Zisler, RC 1987, A close look at equity real estate risk, Goldman Sach.

Ross, SA & Zisler, RC 1991, 'Risk and return in real estate', Journal of Real Estate Finance and


Roulac, SR 1976, 'Can real estate outperform common stocks?', Journal of Portfolio


Rubens, JH, Bond, MT & Webb, JR 1989, 'The inflation-hedging effectiveness of real estate',


S&P 2012, Global Property and REIT Quantitative Analysis: 1st Quarter 2012, Standard and

Poors.

Seiler, MJ, Webb, JR & Myer, FCN 1999, 'Diversification issues in real estate investment',

Journal of Real Estate Literature, Vol. 7, No. 2, pp. 163-79.

Simpson, MW, Ramchander, S & Webb, JR 2009, 'The asymmetric response of equity REIT

returns to inflation', Journal of Real Estate Finance and Economics, Vol. 34, No. 4, pp. 513-29.

Sing, TF, Ho, KHD & Mak, MF 2002, 'Real estate market reaction to public listings and

acquisition news of Malaysian REITs', Journal of Real Estate Portfolio Management, Vol. 8,

No. 2, pp. 209-27.

Sing, TF & Ling, SC 2003, 'The role of Singapore REITs in a downside risk asset allocation

framework', Journal of Real Estate Portfolio Management, Vol. 9, No. 3, pp. 219-36.

Sing, TF & Low, SHY 2001, 'A comparative analysis of inflation-hedging characteristics of

property and non-property assets in Singapore', Journal of Real Estate Portfolio Management,

Vol. 6, No. 4, pp. 373-86.

Sing, TF & Sng, SBS 2003, 'Conditional variance tests of integration between direct and indirect

real estate markets', Journal of Property Investment & Finance, Vol. 21, No. 4, pp. 366-82.

289

Sirmans, CF & Worzala, E 2003a, 'International direct real estate investment: a review of the

literature', Urban Studies, Vol. 40, No. 5-6, pp. 1081-114.

Sirmans, CF & Worzala, E 2003b, 'Investing in international real estate stocks: a review of the

literature', Urban Studies, Vol. 40, No. 5-6, pp. 1115-49.

Sirmans, GS & Sirmans, CF 1987, 'The historical perspective of real estate returns', The Journal

of Portfolio Management, Vol. 13, No. 3, pp. 22-31.

Smith, KV 1980, 'Historical returns of real estate equity portfolios', in SN Levine (Ed.),

Investment Managers Handbook, Down Jones-Richard D Irwin, Homewood, pp. 426-42.

Smith, KV & Shulman, D 1976, 'The performance record of the equity REITs', Finance Analyst

Journal, Vol. 32, No. 5, pp. 61-6.

Stevenson, S 2000, 'International real estate diversification: empirical tests using hedged indices',

Journal of Real Estate Research, Vol. 19, No. 1-2, pp. 105-31.

Stevenson, S 2002, 'An examination of volatility spillovers in REIT returns', Journal of Real


Stevenson, S & Murray, L 1999, 'An examination of the inflation hedging ability of Irish real

estate', Journal of Real Estate Portfolio Management, Vol. 5, No. 1, pp. 59-69.

Sweeney, F 1988, 'International real estate diversification: a viable investment strategy', Property


Sweeney, F 1989, 'Investment strategy: a property market without frontiers', Estate Gazette, Vol.

89, No. 35, pp. 23-30.

Thomson Reuters 2011, Thomson Reuters Indices Country & Region Index Methodology,

Thomson Reuters.

TI 2012, Corruption Perception Index, Transparency International.

Ting, KH & Choi, KW 2011, 'Effects of property portfolio characteristics on M-REIT risks',

Pacific Rim Property Research Journal, Vol. 17, No. 2, pp. 230-47.

290

Trust 2011, Asia-Pacific REIT Survey 2011, The Trust Company and Baker & McKenzie.

Tsai, IC & Lee, CF 2012, 'The convergent behavior in REIT markets', Journal of Property

Investment and Finance, Vol. 30, No. 1, pp. 42-57.

Tse, RYC 2001, 'Impact of property prices on stock price in Hong Kong', Review of Pacific

Basin Financial Markets and Policies, Vol. 4, No. 1, pp. 1-15.

Tse, RYC, Chiang, YH & Raftery, J 1999, 'Office property returns in Shanghai, Guangzhou, and

Shenzhen', Journal of Real Estate Literature, Vol. 7, No. 2, pp. 197-208.

Tse, RYC & Webb, JR 2000, 'Public versus private real estate in Hong Kong', Journal of Real


Tse, RYC & Webb, JR 2001, 'Public vs. private estate in Hong Kong using adaptive

expectations', Journal of Real Estate Portfolio Management, Vol. 7, No. 2, pp. 143-9.

UN 2010, World Population Prospects, the 2010 Revision, United Nations, Department of

Economic an Social Affairs.

US SEC 2012, Real Estate Investment Trusts (REITs), U.S. Securities and Exchange

Commission, viewed 30 April 2012, <http://www.sec.gov/answers/reits.htm>.

Webb, JR, Chau, KW & Li, LH 1997, 'Past and future sources of real estate returns in Hong

Kong', The Journal of Real Estate Research, Vol. 13, No. 3, pp. 251-71.

Webb, JR, Curcio, RJ & Rubens, JH 1988, 'Diversification gains from including real estate in

mixed-asset portfolio', Decision Sciences, Vol. 19, No. 2, pp. 434-52.

Webb, JR, Miles, M & Guilkey, D 1992, 'Transactions-driven commercial real estate returns: the

panacea to asset allocation models?', Journal of the American Real Estate and Urban Economic

Association, Vol. 20, No. 2, pp. 325-57.

Webb, JR & Rubens, JH 1986, 'Portfolio considerations in the valuation of real estate', Real


http://www.sec.gov/answers/reits.htm%3e

291

Webb, JR & Rubens, JH 1987, 'How much in real estate? a surprising answer', The Journal of


Webb, JR & Rubens, JH 1989, 'Diversification gains from including foreign real estate in a

mixed asset portfolio', paper presented to American Real Estate Society Meetings, San

Francisco.

Webb, JR & Sirmans, CF 1980, 'Yields and risk measures for real estate, 1966-77', Journal of


Webb, JR & Tse, RYC 2000, 'Regional comparison of office prices and rentals in China:

Evidence from Shanghai, Guangzhou and Shenzhen', Journal of Real Estate Portfolio


WEF 2011, The Global Competitiveness Report 2010-2011, World Economic Forum.

WEF 2012, The Global Competitiveness Report 2011–2012, World Economic Forum.

Wendt, PF & Wong, SN 1965, 'Investment performance: common stocks v. apartment houses',

Journal of Finance, Vol. 20, No. 4, pp. 633-46.

Westerheide, P 2006, 'Cointegration of real estate stocks and REITs with common stocks, bonds

and consumer price inflation: an international comparison', ZEW - Centre for European

Economic Research, Discussion Paper No. 06-057.

Wilson, P & Okunev, J 1999, 'Special analysis of real estate and financial assets markets',


Wurtzebach, C 1991, 'Looking at non-U.S. real estate', Real Estate Review, Vol. 21, No. 1, pp.

48-54.

Wurtzebach, CH, Mueller, GR & Machi, D 1991, 'The impact of inflation and vacancy on real

estate returns', Journal of Real Estate Research, Vol. 6, No. 2, pp. 153-68.

Yobaccio, E, Rubens, JH & Ketcham, DC 1994, 'The inflation-hedging properties of risk assets:

the case of REITs', Journal of Real Estate Research, Vol. 10, No. 3, pp. 279-97.

292

Zhou, X & Clements, S 2010, 'The inflation hedging ability of real estate in China', Journal of

Real Estate Portfolio Management, Vol. 16, No. 3, pp. 267-77.

Ziering, B, Winograd, B & McIntosh, W 1997, 'The evolution of public and private market

investing in the new real estate capital markets', Real Estate Finance, Vol. 14, No. 2, pp. 21-8.

Ziobrowski, BJ & Ziobrowski, AJ 1997, 'Higher real estate risk and mixed-asset portfolio

performance', Journal of Real Estate Portfolio Management, Vol. 3, No. 2, pp. 107-15.

293

APPENDICES

List of Publications







List of Conference Presentations

1. Pham, AK, Newell, G, Wing, CK & Wong, SK 2012, 'The significance of a pan-Asia

REIT strategy', presentation at the Asian Real Estate Society (AsRES) and American

Real Estate and Urban Economics Association (AREUEA) Joint International

Conference, 07-10 July 2012, Singapore.


Asian REIT markets', presentation at the Pacific Rim Real Estate Society 18th

Annual

Conference, 15-18 January 2012, Adelaide, Australia.


asset portfolio', presentation at the Asian Real Estate Society (AsRES) and American

Real Estate and Urban Economics Association (AREUEA) joint International

Conference, 11-14 July 2011, Jeju Island, Korea.

an empirical analysis of real estate investment …17… · real estate investment trusts in asia:...

Documents