are asean stock markets interdependent? econ 7710 che yuen shan guerra archimedes david
TRANSCRIPT
Are ASEAN Stock Markets Interdependent?
ECON 7710
CHE Yuen Shan
GUERRA Archimedes David
Evidence #1
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
00 01 02 03 04 05 06 07 08 09 10
JCI KLCI PSEISET STI
Evidence # 20
1,000
2,000
3,000
4,000
JCI
400
800
1,200
1,600
KLCI
0
1,000
2,000
3,000
4,000
5,000
PSEI
200
400
600
800
1,000
1,200
SET
1,000
2,000
3,000
4,000
01,000
2,000
3,000
4,000
JCI
STI
400 800 1,200 1,600
KLCI
0 2,000 4,000 6,000
PSEI
0 400 800 1,200
SET
1,000 2,000 3,000 4,000
STI
Evidence #3
Correlation JCI KLCI PSEI SET STI JCI 1.0000KLCI 0.9519 1.0000PSEI 0.9491 0.9639 1.0000SET 0.8015 0.8136 0.8008 1.0000STI 0.8631 0.9317 0.9487 0.7681 1.0000
Why is it important to know if stock markets are interdependent? Serve as the basis for global diversification
strategy Help guide regional financial policy
formulation Reflect regional economic couplings or
linkages
800
1,200
1,600
2,000
2,400
2,800
3,200
3,600
4,000
4,400
0 1,000 2,000 3,000 4,000
JCI
PSEI
Spurious Regression
800
1,200
1,600
2,000
2,400
2,800
3,200
3,600
4,000
4,400
0 1,000 2,000 3,000 4,000
JCI
PSEI
Spurious Regression
Cointegration tests aim to avoid “spurious regressions”, and show valid long term equilibrium relationships (Granger 1986)
The Data Adjusted weekly closing prices of five ASEAN
stock market indices: Indonesia – Jakarta Composite Index (JCI) Malaysia – Kuala Lumpur Composite Index (KLCI) Philippines – Philippine Stock Exchange Index
(PSEI) Singapore – Straits Times Index (STI) Thailand – Stock Exchange of Thailand Index (SET)
Advantages of weekly data Daily data contain “too much noise” (Bailey and
Stulz,1990) Monthly data highly seasonal (Roca et al., 1998)
The Data Period of study: January 3, 2000 to November
8, 2010 (567 weeks) After 1997 Asian financial crisis Covers widespread use of Internet and other
advances in communications technology Includes period from 2008 subprime crisis to today
The indices are in domestic currency to avoid problems from transforming the indices into one currency due to the cross-country exchange rate (Subramanian 2008)
The Data Descriptive Statistics:
Weekly Closing Prices
PSEi JCI KLCI STI SETMean 2,097 1,288 951 2,219 581 SD 821 862 247 624 192 Min 993 337 558 1,171 258 Median 1,917 1,074 902 2,087 654 Max 4,342 3,699 1,516 3,819 1,050 Skewness 0.6218 0.7162 0.5301 0.4780 -0.1714Kurtosis -0.6747 -0.6765 -0.8325 -0.7383 -1.2329
The Data Descriptive Statistics:
Weekly Closing Prices in Natural Log
PSEi JCI KLCI STI SETMean 7.5732 6.9227 6.8247 7.6658 6.3028SD 0.3877 0.7051 0.2548 0.2802 0.3673Min 6.901 5.8215 6.3237 7.0655 5.5533Median 7.5588 6.979 6.8043 7.6433 6.483Max 8.376 8.2159 7.324 8.2477 6.9563Skewness 0.113 0.0544 0.1693 0.0402 -0.5029Kurtosis -1.0991 -1.4155 -0.9684 -0.9108 -1.1903
Methodology1. Unit Root Test
To find out if the time series are nonstationary , and are integrated of the same order, I(d) An I(d) time series has to be differenced d times before
it becomes stationary
2. Cointegration Test To test the presence of long-run equilibrium
relationships among nonstationary time series, or if these share similar stochastic trends
3. Granger Causality Test To test possible short-term price linkages among
the ASEAN markets (Roca et al. ,1998)
Unit Root Test Augmented Dickey-Fuller (ADF) Test
Extends the standard Dickey-Fuller test to cover higher-order autoregressive processes, AR(p)
Tests the following hypotheses using the t-ratio H0 : = 0
H1 : < 0
where = - 1
tptptttt vyyyxyy ...' 22111
Unit Root Test Phillips-Perron (PP) Test
Modifies the t-ratio of the coefficient so that serial correlation does not affect the asymptotic distribution of the test statistic
sf
sefT
ftt 2/1
0
00
2/1
0
0
2
))ˆ()((~
Unit Root Test Results
All market index series are nonstationary All market index series are integrated of order
one, or I(1) The set of series are fit for cointegration
testing
Market Level First Difference* Level First Difference*Indonesia 1.8192 -25.1760 1.3378 -26.3160Malaysia 0.0094 -22.5287 -0.2363 -22.5905Philippines 0.4786 -24.1537 0.4427 -24.5323Singapore -0.7405 -23.2428 -1.0089 -24.9313Thailand -0.1224 -23.1362 -0.1868 -23.1875*Significant at 0.001 level for all markets
ADF PP
Cointegration Test Johansen Cointegration Test (1995)
Given a VAR of order p
where yt is a k-vector of nonstationary I(1) variables, xt is a d-vector of deterministic variables, and t is a vector of innovations. This VAR may be rewritten as
where
ttptptt BxyAyAy ...11
tt
p
iititt Bxyyy
1
11
p
ii IA
1
p
ijji A
1
Cointegration Test Assumptions
The level data yt have no deterministic trends and the cointegrating equations have intercepts:
Lag Interval: 1 1
)'( 011 ttt yBxy
Cointegration Test Results
Unrestricted Cointegration Rank Test (Trace)
Hypothesized Trace 0.05No. of CE(s) Eigenvalue Statistic Critical Value Prob.**None * 0.0675 77.7479 76.9728 0.0436At most 1 0.0310 38.2704 54.0790 0.5587At most 2 0.0158 20.4650 35.1928 0.6971At most 3 0.0123 11.4699 20.2618 0.4974At most 4 0.0079 4.4802 9.1645 0.3453
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values
Cointegration Test Results
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized Max-Eigen 0.05No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None * 0.0675 39.4775 34.8059 0.0129At most 1 0.0310 17.8054 28.5881 0.5923At most 2 0.0158 8.9951 22.2996 0.9043At most 3 0.0123 6.9897 15.8921 0.6708At most 4 0.007898 4.480234 9.164546 0.3453
Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values
Granger Causality Test The following error correction models are
tested: Unrestricted models:
Restricted models:
The null hypothesis is all the s of the unrestricted model equals zero. If the null hypothesis is rejected, then the series are Granger causes to each other.
ttit
p
iiit
p
iit weYXY
111
ttit
p
iiit
p
iit weXYX
111
ttit
p
iit weYY
11
ttit
p
iit weXX
11
Granger Causality Test Results
Index Explained PSEi JCI KLCI STI SET PSEi 3.60** 2.88* 114.18*** 12.08*** 4.29**
(0.0584) (0.0903) (0.0001) (0.0005) (0.0388)JCI 0.08 1.77 173.56*** 15.14 7.15***
(0.7774) (0.1842) (0.0001) (0.0001) (0.0077)KLCI 4.54** 2.21 0.17 16.46*** 7.55***
(0.0335) (0.1373) (0.6839) (0.0001) (0.0062)STI 2.67 10.63 111.99*** 4.68 0.02
(0.1029) (0.0012) (0.0001) (0.0309) (0.8817)SET 0.68 1.05 66.36*** 5.75 2
(0.4096) (0.3051) (0.0001) (0.0168) (0.1576)
Explanatory Index
Granger Causality Test Results The ASEAN market indices, except the PSEi,
are not significantly affected by their own previous prices
KCLI has a two-way linkage with all other markets. PSEi has a unidirectional causality with all the indices except KCLI
Both STI and SET can only be explained by the past price of KLCI significantly and are insignificantly explains be the other ASEAN market indices
References Abhyankar, A., “Linear and nonlinear Granger causality: Evidence from the
U.K. stock index futures market.” Journal of Futures Markets 18 (1998), 519-540.
Bailey, W. and Stulz, R., “Benefits of International Diversification: The Case
of the Pacific Basin Stock Markets.” Journal of Portfolio Management (Summer 1990), 57-61
EViews 6 User’s Guide II. Quantitative Micro Software, LLC (2007). Granger, C.W.J, “Causality, cointegration, and control.” Journal of Economic
Dynamics and Control 12 (1988), 551-559. Granger, C.W.J, “Developments in the Study of Co-integrated Economic
Variables.” Oxford Bulletin of Economics and Statistics 48 (1986), 226 Granger, C.W.J, “Investigating Causal Relationships by Econometric Models
and Cross Spectral Models.” Econometrica 37 (1969), 424-438.
References Helmut, L., Hans-Eggert, R., “Granger-causality in cointegrated VAR
processes: The case of the term structure.” Economics Letters 3 (1992), 263-268.
Johansen, S., Likelihood-based Inference in Cointegrated Vector Autoregressive Models. Oxford: Oxford University Press (1995).
Roca, E. D., Selvanathan, E. A., Shepherd, W. F., “Are the ASEAN Equity
Markets Interdependent?” ASEAN Economic Bulletin 15 (1998), 109-120. Subramanian, U., “Cointegration of Stock Markets in East Asia.” European
Journal of Economics, Finance and Administrative Sciences 14 (2008). http://support.sas.com/rnd/app/examples/ets/tourism/index.htm http://support.sas.com/documentation/cdl/en/etsug/60372/HTML/default/
etsug_varmax_sect049.htm http://support.sas.com/rnd/app/examples/ets/granger/index.htm