macquarie technicals

Quantitative analysis AUSTRALIA

30 April 2009

Inside What goes up must come down 2 Market instability hurts fundamental factors 3 Using Technical Indicators as Quant factors 4 Performance of Technical Indicators in Australia 5 Combining with other factors and use in a factor model 10 References 14 Appendices – Factor definitions 15

Analysts Burke Lau 612 8232 0481 [email protected] John Conomos 612 8232 5157 [email protected] George Platt 612 8232 6539 [email protected]

Let’s get technical Technical indicators add value in Australia Technical indicators appear to have predictive power in Australian markets. These signals are typically a more sophisticated form of short term mean reversion signals. We think these results are encouraging as the standard one month mean reversion signal has typically not been predictive in Australia.

Volatile markets present opportunity to use technicals The large macro-driven swings in equity markets have resulted in deterioration of performance for a number of elementary quantitative factors. The factor volatility makes it necessary to utilise smarter techniques when searching for Alpha. A large number of technical indicators have had strong performance in the volatile period since August 2007.

Williams %R the most powerful signal The Williams %R signal stands out as the strongest performing technical indicator, with an IC 5.6%. The Williams %R is an enhanced mean reversion signal and measures the relative location of the closing price in relation to a recent high-low range. Other key indicators that worked well were: force index, commodity channel index and relative strength indicator.

Size matters – technicals work best in large cap stocks In doing this analysis, we also found that there was a significant difference in behaviour between large cap and small cap stocks.

Large cap stocks (ASX100) tend to be mean reverting which favours the mean reverting technical indicators over trend following strategies.

Small cap stocks (ex-ASX100) tend to be trend following, this penalises the mean trend Technical indicators and favours momentum based strategies.

Improving the investment process with technicals Incorporating a technical trading signal in an existing investment process can add significant value as these signals tend to be lowly correlated to the standard value, momentum, quality, growth and analyst sentiment strategies. However, careful consideration must be given to the higher turnover of these signals which makes them hard to include in a process.

We look at different ways of incorporating the technical indicators in an investment process. To do this we use the Williams %R:

as an extra factor in a standard four factor model; and

as a ‘knock out’ signal in a standard quant model.

In general, we find that the ‘knock out’ method was a good way to incorporate the predictive strength and timing ability of the technical indicators while moderating the effects of higher turnover.

Please refer to the important disclosures and analyst certification on inside back cover of this document, or on our website www.macquarie.com.au/research/disclosures.

Macquarie Research Equities - Report Quantitative analysis

30 April 2009 2

Let’s get technical Defining technical indicators With the recent volatility in the market as well as weakness in the performance of some of the standard quant factors like value, we turn our attention to faster burn short term strategies that may work better in volatile markets.

Technical indicators are generally short term mean reverting strategies. In this report, we examine a number of technical indicators in the Australian market and test their efficacy in forecasting short term stock performance.

Williams %R – the standout technical indicator Williams %R was the standout technical indicator from our analysis. It has had consistent performance through time and across the large cap and small cap stocks.

Fig 1 Williams %R information coefficient Fig 2 Technical performance in large vs. small caps

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

Dec 94Ju

l 95

Feb 9

6

Sep 96Apr

97Nov

97Ju

n 98Ja

n 99

Aug 99

Mar 00Oct

00

May 01

Dec 01Ju

l 02

Feb 0

3

Sep 03Apr

04Nov

04Ju

n 05Ja

n 06

Aug 06

Mar 07Oct

07

May 08

Dec 08

Information Coefficient

Monthly IC 12M Average IC

All History

Last 5 Years

Last 3 Years

Last 1 Year

Average IC 5.6% 5.4% 4.8% 5.9%Std Deviation IC 9.6% 7.9% 8.7% 12.3%Success Rate 73.7% 73.3% 63.9% 50.0%Avg IC / Std Dev IC 0.59 0.68 0.55 0.48t-stat 7.68 5.29 3.29 1.65

-4% -2% 0% 2% 4% 6% 8% 10% 12%

MACD

1M Mean Reversion

On Balance Volume

Stochastic Oscillator

Money Flow Index

Relative Strength Indicator

Commodity Channel Index

Bollinger Band

Williams %R

Force Index

Price relative to 52 Week High

Earnings Yield

3M Earnings Revisions

12M Momentum

Macquarie Alpha Model


Small Cap - ex ASX100Large Cap - ASX100

Tech

nica

lInd

icat

ors

Source: Macquarie Research, April 2009

Other key technical indicators that also performed well were:

Force Index


Bollinger Band


Large caps mean revert; small caps trend follow Mean reverting technical indicators work significantly better in large cap stocks. The standard one month mean reversion signal is negatively predictive in the Australian market (IC = -0.72%). But when split by size it outperforms in the large caps (IC=1.5%) while underperforming strongly in the small cap stocks (IC=-2.24%).

Adding value to a quant process Typically, mean reverting technical indicators are negatively correlated with the standard momentum, value and analyst revisions signals. This low correlation indicates that technical indicators can add significant value to an existing process.

A drawback of the typical technical indicators is that they have very high turnover levels. The Williams %R has an annual Q1-Q5 2-way turnover of 3835%. This is significantly higher than typical quant factors. The increased turnover from adding a technical signal to a standard quant process destroys most of the added alpha from the technical indicator.

An alternative is to use the technicals as a knock out overlay to a standard quant model. This technique can take advantage of the timing ability of the technical indicators without significantly increasing the turnover of the portfolio.

Short term technical indicators are

typically enhanced mean reverting

signals.

Technical Indicators perform strongly in

the large cap space.


30 April 2009 3

Market instability hurts fundamental factors Equity markets have proved extremely challenging for investors since the foundations of the market began to shake in the second half of 2007. As soaring levels of uncertainty persisted, fundamental factors entrenched in security valuation have had highly volatile performance. This volatility has made life exceptionally difficult for investors as strong performance for a month or two is often followed by severe underperformance.

Adding to the challenges for investors is the terrible performance of typical medium to longer term quant strategies such as value. Value factors are a vital input into most quantitative models and these factors have had arguably their worst performance in recent history (Figures 3 and 4).

Fig 3 ASX200 cross sectional volatility Fig 4 Performance of value factors

0

5

10

15

20

25

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008


Source:Macquarie Research, April 2009 Source: Macquarie Research, April 2009

Turning our attention to technicals While in the long run we believe volatility will reside and the fundamental factors will again return to the fore; in the short-term, the degree of macro instability indicates that the volatility will persist. With this in mind we turn our attention to short term focused technical indicators.

Our European team first looked at technical factors in its March Quantamentals report1. They tested (in European markets) whether technical factors can systematically add value on a standalone basis and if so, can they be incorporated into the investment process. We focus specifically on Australia to see how the results compare.

Technical analysis is a contentious topic, with many investors and members of the academic community expressing strong opinions on the validity of the signals. This often leads to heated debate with critics rejecting the use of these metrics. Technical analysis is often viewed as being too subjective as technical analysts will interpret the same chart in several different ways depending upon the metric employed and also the period being examined. Hence, investors typically do not include these metrics within their models.

1 Davies, C. “Quantamentals: Do technicals add value?”, Macquarie Research Equities, (2009)

Higher time series and cross sectional

volatility has increased returns to

mean reveting strategies.


30 April 2009 4

Using technical indicators as Quant factors The number of technical indicators in existence is significant, so we have mainly focused on the most commonly available indicators. These include the more frequently discussed indicators on websites and those most readily available on platforms such as Factset and Bloomberg. Figure 5 below shows the indicators that we tested. A brief description and definitions of these indicators are included in the Appendix.

Fig 5 Technical indicators tested Price Based Technical Indicators Price & volume Based Technical Indicators Volume Based Technical Indicators

Relative Strength Index Money Flow Index Trading Intensity Williams %R Chaikin Money Flow Index MACD On Balance Volume Stochastic Oscillator Force Index Commodity Channel Index Accumulation / Distribution Index Bollinger Band %b Moving Averages Price Relative to 52 Week High & Low Average Directional Movement Index (ADX) Relative Volatility Source: Macquarie Research, April 2009

When testing the technical indicators as quant factors, a key characteristic to keep in mind is that most of the indicators were designed to be used on a univariate basis, relative to its own history. Typically, these technical indicators attempt to detect when a stock (relative to its own history):

is over-bought

is over-sold

has broken out from a trading range

changes its price trends

Because of this, the scale of some of the technical indicators may not match up when compared cross-sectionally.

This is different from typical quant signals which tend to be used cross-sectionally to compare and rank one stock against another, eg on balance volume is a running total of (volume * direction of price move), which typically grows over time is very different for large cap vs. small cap stocks.

To take this into account we normalise a signal against its own history (using the last 12 months of data). This makes the signals:

comparable in magnitude,

picks up any changes relative to it’s own history and

filters out stock and market drift effects (momentum) in the signal.

Additionally, the majority of the technical indicators use a trailing window of data to calculate the underlying signal. We notice from our backtests that a window of ten trading days typically yields the best results. Given this result we will be using the 10-day-based metrics from this point forward rather than spend hours debating whether we should a 10-day, 14-day or 11-day window.

We test the standard signals

We time series normalise technical indicators to detect

changes in the signal relative to its

own stock history.


30 April 2009 5

Performance of technical indicators in Australia

Fig 6 Technical factor performance in Australia – Dec 1994 – Mar 2009 Information Coefficient Q1-Q5 Long Short

Average ICIC Avg/Sd

t-statIC Success

RateReturn

(p.a)Volatility

(p.a.)Informatio

n RatioSuccess

Rate

Largest 1 Month Gain

Largest 1 Month

Loss

Turnover (2 Way

p.a.) Price Momentum

3 Month Momentum 5.5% 1.47 67.8% 29.1% 17.7% 1.65 73.1% 19.2% -23.7% 2076%6 Month Momentum 5.7% 1.45 68.4% 27.6% 19.8% 1.40 73.1% 16.4% -31.1% 1476%12 Month Momentum 7.1% 1.70 71.3% 29.3% 19.9% 1.48 73.1% 16.2% -29.5% 1062%

Price Reversal1 Month Mean Reversion -0.7% -0.21 46.8% -10.5% 14.8% -0.71 43.3% 15.5% -16.1% 3655%

Relative Strength Indicator - 14 Day (Inverted)Standard 1.3% 0.42 51.5% -4.5% 12.1% -0.37 47.4% 12.7% -11.4% 3808%12M Normalised 3.6% 1.33 63.7% 5.9% 10.2% 0.58 58.5% 12.0% -9.3% 3888%

Williams %R (Inverted)Standard 4.2% 1.56 69.6% 7.5% 10.0% 0.75 63.7% 11.5% -9.7% 3776%12M Normalised 5.6% 2.22 73.7% 13.2% 10.2% 1.29 66.7% 12.3% -7.5% 3835%

52 Week High / LowPrice relative to 52 Week High 6.6% 1.57 73.1% 24.6% 18.3% 1.34 69.0% 20.7% -30.0% 1767%Price relative to 52 Week Low 2.7% 0.69 62.0% 15.9% 17.0% 0.94 66.7% 15.4% -28.2% 1363%

Bollinger Band %b (Inverted)%b 2.7% 0.89 60.8% 0.6% 10.5% 0.06 56.7% 11.7% -8.3% 3847%%b - 12 Month Normalised 4.4% 1.67 67.3% 7.7% 10.3% 0.76 64.9% 10.9% -7.3% 3886%

Force Index (inverted) - 10 PeriodStandard 3.8% 1.66 67.3% 6.8% 9.2% 0.73 60.8% 11.0% -10.9% 3332%12 Month Normalised 4.4% 1.94 69.0% 9.4% 8.8% 1.07 64.9% 9.9% -11.7% 3717%

Commodity Channel Index - 10 Period (Inverted)Standard 2.7% 1.00 59.1% -0.1% 10.1% -0.01 50.9% 10.4% -9.9% 3830%12 Month Normalised 4.1% 1.69 65.5% 8.0% 9.1% 0.88 63.2% 12.9% -6.8% 3865%

Stochastic Oscillator (Inverted)Standard -2.0% -0.62 44.4% -10.5% 12.9% -0.81 39.8% 17.1% -12.8% 3588%12M Normalised 1.0% 0.34 50.9% 0.9% 10.7% 0.08 52.6% 12.5% -10.5% 3654%

MACD (Inverted) 12 Month Normalised -0.8% -0.27 48.5% -7.4% 11.9% -0.63 44.4% 12.8% -11.2% 3395%

Money Flow Index (Inverted)Standard 0.7% 0.28 52.6% -4.4% 9.1% -0.48 38.0% 12.7% -8.1% 3758%12 Month Normalised 2.1% 0.92 62.6% 0.8% 8.5% 0.09 50.9% 10.9% -7.1% 3837%

Moving Average (20 Period / 60 Period)Standard 2.4% 0.56 52.0% -1.3% 19.7% -0.07 45.0% 30.1% -16.7% 2036%12 Month Normalised 1.2% 0.45 57.9% 1.9% 10.7% 0.17 55.6% 12.4% -12.5% 2872%

Accumulation / Distribution IndexStandard 0.2% 0.07 50.3% 1.9% 11.5% 0.16 51.5% 13.1% -11.7% 3076%12 Month Normalised -1.0% -0.41 50.3% -1.1% 9.8% -0.12 49.1% 9.5% -10.0% 3716%

Average Directional Movemebnt Index (ADX)Standard -0.2% -0.09 46.8% -1.0% 7.3% -0.13 49.1% 7.8% -6.3% 3877%12 Month Normalised -0.8% -0.43 47.4% -2.5% 7.1% -0.36 43.9% 6.5% -6.4% 3711%

Average True Range (Inverted) - 10 PeriodStandard 1.7% 0.55 52.0% 3.9% 12.3% 0.31 59.1% 19.8% -13.5% 1234%12 Month Normalised -0.7% -0.26 42.1% -5.6% 11.8% -0.48 41.5% 22.6% -11.3% 3302%

Chankin Money Flow Index - 21 PeriodStandard 0.3% 0.11 52.6% 2.4% 11.1% 0.22 54.4% 8.5% -12.0% 2829%12 Month Normalised -1.1% -0.42 48.0% -1.1% 9.8% -0.11 54.4% 7.5% -11.7% 3661%

On Balance Volume12 Month Normalised 0.7% 0.31 51.5% -2.3% 9.5% -0.24 41.5% 11.4% -7.4% 3761%

Trading Intensity1M / 12M Average Daily Volume 4.7% 1.93 70.2% 18.1% 11.7% 1.55 68.4% 9.2% -16.8% 2723%

Relative Volatility1M / 12M Daily Volatility 2.2% 0.95 60.8% 6.8% 9.7% 0.70 62.6% 8.6% -12.1% 3416%



30 April 2009 6

A key characteristic of the Australian market is the stellar performance of the 12 month price momentum factor. This is where stocks with high (low) 12-month price momentum continue to outperform (underperform). The intuition behind the momentum phenomenon in Australia resides in the recurrent cash injections from compulsory superannuation. Managers continue to invest the new funds into their preferred holdings, which in turn drives the performance of momentum.

On the flipside of 12 month momentum is the mean reversion effect. To measure this, we historically use 1 month mean reversion, which simply means stocks that have outperformed over the past month are likely to underperform in the current month. The underlying logic behind many technical indicators is some form of mean reversion. A number of the indicators are effectively ‘enhanced reversion signals’.

Figure 6 above shows the overall performance of a number of momentum and technical signals in the Australian market from December 1994 to April 2009. A key result is that the standard short term mean reversion signal of 1M mean reversion does not seem to work in Australia, with a negative IC of -0.8% and a long/short quintile return of -10.4% pa.

Surprisingly, a number of the short term technical indicators which are mean reverting strategies seemed to perform much better than the standard 1 month mean reversion signal and actually were strongly predictive.

The key mean reverting technical indicators which added value in the Australian market are shown in Figure 7.

Fig 7 Information coefficient of mean reverting technical indicators

-2% -1% 0% 1% 2% 3% 4% 5% 6%

1M Mean Reversion

On Balance Volume


Money Flow Index



Bollinger Band

Force Index

Williams %R

Information CoefficientSource:Macquarie Research, April 2009

Simple 1 month mean reversion doesn’t work in

Australia

Momentum strategies are

strongly predictive in Australia


30 April 2009 7

The Williams %R signal stands out from this list of technical indicators. It has a strong consistent IC as well as good quintile long/short performance. Also encouraging with this signal is its consistent performance in most developed markets2. The charts below highlight some performance statistics for the signal.

Fig 8 Williams %R information coefficient Fig 9 Q1-Q5 long / short performance

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

Dec 9

4Ju

l 95

Feb 9

6

Sep 9

6

Apr 9

7

Nov 9

7

Jun 9

8

Jan 9

9

Aug 9

9

Mar 00

Oct 00

May 01

Dec 0

1Ju

l 02

Feb 0

3

Sep 0

3

Apr 0

4

Nov 0

4

Jun 0

5

Jan 0

6

Aug 0

6

Mar 07

Oct 07

May 08

Dec 0

8



All History Last 5 Years

Last 3 Years Last 1 Year


-10%

0%

10%

20%

30%

40%

50%

Dec 9

4Ju

l 95

Feb 9

6

Sep 9

6

Apr 9

7

Nov 9

7

Jun 9

8

Jan 9

9

Aug 9

9

Mar 00

Oct 00

May 01

Dec 0

1Ju

l 02

Feb 0

3

Sep 0

3

Apr 0

4

Nov 0

4

Jun 0

5

Jan 0

6

Aug 0

6

Mar 07

Oct 07

May 08

Dec 0

8

Long Short Returns

LS Return 12M Rolling LS returns


Last 3 Years Last Year

Quartile 1 Excess Return * 4.4% 3.8% 3.5% 3.7%Quartile 5 Excess Return * -8.3% -10.4% -11.9% -23.3%Q1 - Q5 Return (p.a.) # 13.2% 15.1% 16.4% 32.1%Q1 - Q5 Volatility (p.a.) # 10.2% 10.1% 12.2% 19.6%Q1 - Q5 Information Ratio # 1.29 1.50 1.34 1.64Q1 - Q5 Turnover 1 way p.a.# 3841% 3868% 3844% 3867%* Monthly rebalanced annualised excess returns (long - market and market - short)# Monthly rebalanced Long Q1 - Short Q5


Of particular interested is the recent pick-up in the predictive strength of the Williams %R signal which is in line with a pick up in performance of a number of technical indicators we have noticed over the last year.

2 Davies, C. “Quantamentals: Do Technical’s add value?”, Macquarie Research Equities, (2009)

Williams %R stands out from the crowd


30 April 2009 8

Size performance In our report, dissecting the Alpha model in Australia3, we further examined factor performance across different universes. One of the key findings was the shifting performance of factors across the different universes definitions. To test this we divided the universe into two:

A large cap universe containing the top 100 stocks in the market and

A small cap universe consisting of the top 300 stocks ex the top 100 stocks.

Fig 10 Technical indicator information coefficient in large vs. small cap universes

Fig 11 Q1-Q3 L/S information ratio before costs in large vs small cap universe

-4% -2% 0% 2% 4% 6% 8% 10% 12%

MACD

1M Mean Reversion

On Balance Volume


Money Flow Index



Bollinger Band

Williams %R

Force Index


Earnings Yield


12M Momentum




Tech

nica

lInd

icat

ors

-1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00

MACD

1M Mean Reversion

On Balance Volume


Money Flow Index



Bollinger Band

Williams %R

Force Index


Earnings Yield


12M Momentum


Information Ratio


Tech

nica

lInd

icat

ors

Source: Macquarie Research, April 2009 Source: Macquarie Research, April 2009

The key result above is the significant difference in performance between the large cap and small cap universe for the technical indicators. The majority of the technical indicators are mean reverting signals, and perform significantly better in the large cap universe. While the more traditional momentum signals like 12M momentum and earnings revisions favour the smaller cap universe.

3 R. De Souza, “Macquarie Alpha Model – The alpha model goes under the knife”

Technical indicators perform much

stronger in the large cap markets


30 April 2009 9

Turnover A common criticism of technical indicators is the significant increase in turnover. Figure 12 highlights this by showing the average monthly two-way turnover on an annualised basis for a variety of factors. This is based upon a long short Q1-Q3 strategy over the ASX100 universe. For the more common Quant factors the turnover ranges from 600% to 1,200%. The turnover for the technical indicators is much higher with all the mean reverting technical factors measuring over 3,000%4.

Fig 12 Technical indicator Q1-Q3 L/S annual turnover in ASX 100 universe

Fig 13 Q1-Q3 L/S after cost information ratio in ASX 100 universe with 15bp and 30bp transaction costs

0% 500% 1000% 1500% 2000% 2500% 3000% 3500% 4000%

Bollinger Band

Williams %R



Force Index

Money Flow Index


On Balance Volume

MACD

1M Mean Reversion

Trading Intensity




12M Momentum

Earnings Yield

Average Annual 2 way Turnover

-1.00 -0.50 0.00 0.50 1.00 1.50

Bollinger Band

Williams %R



Force Index

Money Flow Index


On Balance Volume

MACD

1M Mean Reversion

Trading Intensity




12M Momentum

Earnings Yield

Average Information Ratio

After Cost IR (30bps)

After Cost IR (15bps)

Before Cost IR


To further examine the real effects of turnover, we have calculated the information ratios for the same set of factors both pre and post transaction costs.5 The results are illustrated in Figure 13. The high information ratio (even after costs) for a number of the other technical factors (Williams %R, Bollinger Bands, Commodity Channel, RSI and Force Index) highlights the potential for technical indicators to add value in the investment process.

Of the technical factors Williams %R stands out overall:

it has a strong information ratio after costs,

performs consistently through time and

works well in both the large cap and small cap space.

4 Highest possible turnover for a long short quintile strategy is 4800% p.a. or 400% of the face value per month (200% long, 200% short). If the thirds are assumed to be randomly allocated then we would expect each third to only have 1/3 in common with the previous month, so the expected average turnover for a random long short portfolio is 3168% pa. So the mean reverting technical indicators get very close to a random portfolio in turnover terms, with near zero auto correlation in stocks held. 5 To calculate the post transaction cost information ratio, we take the simple long-short return for a particular month and then corresponding monthly two-way turnover multiplied by the assumed transaction cost of 30bp. This gives a monthly long-short return series adjusted for transaction costs from which we can calculate a post transaction cost turnover.

Technical indicators have very high

turnovers which can impact the Alpha

available from the signals


30 April 2009 10

Combining with other factors and use in a factor model Given the encouraging results of the univariate tests, we examine ways that investors may be able to implement the signals into their investment process.

Correlations with other factors A key consideration in deciding whether to use the technical indicators in an existing process is how correlated these signals are to the existing signals in the model. The more un-correlated the strategies the more value they will add to the process. Figure 14 below shows the IC correlation matrix for the different factors from December 1994 to March 2009. For comparison, we included in this analysis the Macquarie Alpha model as a representative quant model as well as some standard quant signals:

1M mean reversion,

12 month momentum,

Earnings revisions and

Earnings yield.

Fig 14 IC Correlation matrix for quant factors and technical indicators

12M

Mom

entu

m

6M M

omen

tum

3M M

omen

tum

Pric

e re

lativ

e to

52

Wee

k Lo

w

Pric

e re

lativ

e to

52

Wee

k H

igh

Maq

uarie

Alp

ha

Mod

el

3M E

arni

ngs

Revi

sion

s

Trad

ing

Inte

nsity

Rela

tive

Vola

tility

Acc

umul

atio

n /

Dis

trib

utio

n In

dex

Chai

kin

Mon

ey F

low

Aver

age

Dir

ectio

nal

Mov

emen

t

Earn

ings

Yie

ld

Mov

ing

Ave

rage

Forc

e In

dex

Aver

age

True

Ran

ge

Com

mod

ity

Chan

nel I

ndex

Will

iam

s %

R

Mon

ey

Flow

Inde

x

Bolli

nger

Ban

d

On

Bala

nce

Volu

me

Rela

tive

Stre

ngth

In

dica

tor

Stoc

hast

ic

Osc

illat

or

MAC

D

1M M

ean

Reve

rsio

n

12M Momentum 100% 85% 72% 74% 69% 72% 56% 52% 18% 16% 16% -8% -30% 4% -12% -54% -13% -23% -19% -20% -15% -19% -16% -46% -48%

6M Momentum 85% 100% 88% 68% 78% 64% 50% 59% 12% 31% 31% -4% -36% -13% -8% -62% -22% -31% -30% -32% -28% -32% -31% -55% -60%

3M Momentum 72% 88% 100% 59% 78% 51% 41% 56% 16% 39% 39% -3% -37% -27% -12% -57% -27% -36% -41% -42% -43% -47% -46% -62% -69%

Price relative to 52 Week Low

74% 68% 59% 100% 28% 42% 33% 35% 13% 16% 18% 2% -36% 5% -14% -41% -20% -25% -25% -28% -19% -27% -21% -49% -44%


69% 78% 78% 28% 100% 59% 51% 49% 15% 31% 30% -9% -27% -18% -12% -53% -23% -36% -33% -37% -36% -39% -40% -56% -62%

Maquarie Alpha Model

72% 64% 51% 42% 59% 100% 56% 41% 11% 6% 9% -6% 8% 10% 1% -41% -3% -10% 0% -5% 3% -4% -1% -26% -26%


56% 50% 41% 33% 51% 56% 100% 29% 4% 4% 4% -7% -12% -1% -5% -39% 2% -1% 0% -6% -2% -9% -2% -25% -23%

Trading Intensity 52% 59% 56% 35% 49% 41% 29% 100% -7% 16% 15% 5% -30% -25% -2% -46% -10% -19% -22% -19% -22% -16% -24% -37% -43%

Relative Volatility 18% 12% 16% 13% 15% 11% 4% -7% 100% 16% 17% -5% 1% 39% -6% 42% -1% -8% 2% -5% -10% -11% -8% -13% -12%

Accumulation / Distribution Index

16% 31% 39% 16% 31% 6% 4% 16% 16% 100% 98% -3% -12% -31% -3% -15% -21% -28% -39% -38% -53% -39% -51% -41% -44%

Chaikin Money Flow

16% 31% 39% 18% 30% 9% 4% 15% 17% 98% 100% -4% -10% -29% -2% -14% -22% -28% -36% -38% -53% -39% -50% -42% -44%

Average Directional

-8% -4% -3% 2% -9% -6% -7% 5% -5% -3% -4% 100% -5% -8% -12% -2% -18% -12% 6% -18% 9% -8% 0% -1% -4%

Earnings Yield -30% -36% -37% -36% -27% 8% -12% -30% 1% -12% -10% -5% 100% 13% 4% 39% 4% 10% 11% 13% 18% 12% 18% 30% 35%

Moving Average 4% -13% -27% 5% -18% 10% -1% -25% 39% -31% -29% -8% 13% 100% -8% 40% 3% 7% 20% 13% 22% 16% 40% 20% 26%

Force Index -12% -8% -12% -14% -12% 1% -5% -2% -6% -3% -2% -12% 4% -8% 100% -12% 45% 47% 24% 40% 15% 25% -2% 6% 14%

Average True Range

-54% -62% -57% -41% -53% -41% -39% -46% 42% -15% -14% -2% 39% 40% -12% 100% 3% 5% 13% 12% 10% 14% 24% 37% 40%


-13% -22% -27% -20% -23% -3% 2% -10% -1% -21% -22% -18% 4% 3% 45% 3% 100% 89% 58% 84% 49% 65% 47% 42% 46%

Williams %R -23% -31% -36% -25% -36% -10% -1% -19% -8% -28% -28% -12% 10% 7% 47% 5% 89% 100% 64% 86% 50% 70% 52% 50% 51%

Money Flow Index

-19% -30% -41% -25% -33% 0% 0% -22% 2% -39% -36% 6% 11% 20% 24% 13% 58% 64% 100% 72% 71% 75% 61% 59% 53%

Bollinger Band -20% -32% -42% -28% -37% -5% -6% -19% -5% -38% -38% -18% 13% 13% 40% 12% 84% 86% 72% 100% 67% 89% 72% 71% 68%

On Balance Volume

-15% -28% -43% -19% -36% 3% -2% -22% -10% -53% -53% 9% 18% 22% 15% 10% 49% 50% 71% 67% 100% 71% 73% 67% 72%


-19% -32% -47% -27% -39% -4% -9% -16% -11% -39% -39% -8% 12% 16% 25% 14% 65% 70% 75% 89% 71% 100% 77% 77% 72%


-16% -31% -46% -21% -40% -1% -2% -24% -8% -51% -50% 0% 18% 40% -2% 24% 47% 52% 61% 72% 73% 77% 100% 80% 79%

MACD -46% -55% -62% -49% -56% -26% -25% -37% -13% -41% -42% -1% 30% 20% 6% 37% 42% 50% 59% 71% 67% 77% 80% 100% 89%

1M Mean Reversion

-48% -60% -69% -44% -62% -26% -23% -43% -12% -44% -44% -4% 35% 26% 14% 40% 46% 51% 53% 68% 72% 72% 79% 89% 100%

Trend Following Mean Reverting

Tren

dFo

llow

ing

Mea

n Re

vert

ing

Source:Macquarie Research, April 2009

The factors in figure 14 are sorted from trend following (12 month momentum) to mean reverting (1 month mean reversion).


30 April 2009 11

Most technical factors tend to be mean reverting signals. These technical indicators are typically an enhanced version of the standard one month mean reversion signal – picking up when a stock has been oversold or overbought. A few of the technical indicators fall into the category of trend following signals - picking when a stock has broken out of a range, (eg trading intensity, accumulation / distribution index or price relative to 52 week highs).

A nice result is that the key predictive signals are not highly correlated to the standard quant factors of momentum, value and analyst sentiment. The key mean reverting technical indicators are also generally negatively correlated with the Macquarie Alpha model (Williams %R is -10% correlated).

Even though the technical indicators are highly correlated to 1 month mean reversion they tend to significantly outperform the standard 1M momentum measure and could be a good substitute for this signal.

The low correlations of the technical indicators with the standard quant factors as well as strong predictive strength indicate that it could add significant value to a quant screen or a multi-factor composite quant model.

Combining in a multifactor model To test how a technical indicator would perform in a multi-factor model we created a typical equal weighted four factor model of value, revisions, quality and price momentum6 and compared this to a new model which also includes the Williams %R factor.

Figure 15 plots the two strategies. We tested this in the large cap universe of the MSCI Australia, where the technical indicators would add the most value and be the least impacted by turnover constraints. Due to the smaller restricted universe we used a Q1-Q3 Long Short portfolio to measure the performance characteristics of these portfolios.

As can be seen, replacing the 1-month reversal term with Williams %R metric does enhance the overall strategy. From an information ratio point of view we see the information ratio jump from 0.65 to 0.92 for an equal weighted long-short basket assuming no transaction costs.

Once transaction costs are taken into account (30bp) the information ratio increases only from 0.30 to 0.39. Most of the performance has been eroded by the significantly higher turnover. Figure 16 shows that the turnover has increased from 100% 2-way per month to 148% per month.

Fig 15 Standard and enhanced quant model performance Q1-Q3 performance

Fig 16 Monthly Q1-Q5 2-way turnover

0

100

200

300

400

500

600

Dec

-93

Dec

-94

Dec

-95

Dec

-96

Dec

-97

Dec

-98

Dec

-99

Dec

-00

Dec

-01

Dec

-02

Dec

-03

Dec

-04

Dec

-05

Dec

-06

Dec

-07

Dec

-08

Standard ModelStandard Model With WilliamsStandard Model (After Costs)Standard Model With Williams (After Costs)

0%

20%

40%

60%

80%

100%

120%

140%

160%

Standard Model Standard Model With Williams

Source: FactSet, Macquarie Research, March 2009 Source: FactSet, Macquarie Research, March 2009

We can see that using the Williams %R factor enhances returns both before costs and after costs. However, given the high turnover nature of the Williams %R metric it is hard to include it in the model at a significant weight.

6 For value, we use an equally weighted composite of our fundamental price to earnings and PEG ratio. For price momentum we use 12M momentum. For Revisions we set of revisions metrics (equally weight earnings, dividend, sales, cashflow and book). Finally, for our composite quality metric we equally weight Merton, quick ratio, dividend cover and Altman z-score.

Technical indicators are lowly correlated

to the standard quant factors.

Technical indicators in a multi factor

model struggle to add value due to higher turnover


30 April 2009 12

Overlay on a multifactor model Next, we look at a different way to combine the Williams %R into a multifactor model. Given the high turnover of the signal we look to test the Williams %R as a ‘knockout filter’ for a model. We use the same four factor model of equally weighted value, revisions, quality and price momentum and again tested in the MSCI Australia universe. This time we divide the standard model and the Williams %R technical indicator into three equal weighted groupings. From there we knock out the stocks that appear to be overbought or oversold according to the Williams %R indicator.

The annualised active mean returns of the standard model conditioned on the Williams %R are show in Figure 17. The key corners of interest are the top right and bottom left. The top right portfolio represents the stocks that score favourably on our standard model (long portfolio) but look overbought according to the Williams %R. We would expect these stocks to underperform. The bottom left portfolio represents stocks that are poor in our standard portfolio but have been recently oversold. We would expect these to outperform.

Fig 17 Annualised active returns of filtered portfolios

Over sold Middle

Over bought Unconditional

High 7.7 3.8 -2.4 3.3

Middle 5.8 1.5 -4.0 0.8

Low 3.4 -4.5 -11.2 -4.1 Unconditional 5.7 0.3 -6.0

Williams %R

Sta

ndar

d M

odel

Source: FactSet, Macquarie Research, April 2009

Further breaking down the above diagram (by corners):

For stocks with strong momentum that appear to be overbought (top right) we would expect to see underperformance. This is indeed the case with the annualised market relative return yielding -2.4% vs. 3.3% for the standard unconditional model (Figure 17).

Equally, we would expect strong performance from stocks with strong price momentum that are oversold on technical signals (top left). This combination of characteristics yields +7.7% (Figure 17).

Conversely, when looking at the portfolio that has poor momentum and appears overbought on technical signals (highlighted bottom right) we would expect to see worse performance. This is the case with this combination yielding -11.2% vs. -4.1% for the unconditional portfolio (Figure 17).

Finally, we would expect better performance from the combination of weak momentum and technically oversold stocks (bottom left). This again is true with the new combination yielding +3.4% compared to -4.1%.

Technical indicators as a knock out signal

add value without significantly

increasing turnover.


30 April 2009 13

Using the Williams %R signal as a ‘knockout filter’ for a standard quant model appears to be a practical way to add value to an investment process with technical indicators. The filter helps to avoid stocks in the top tertile that have been overbought and those in the bottom tertile that have been oversold. It is worth noting that the filtering methodology produces more or less the same turnover as the standard model.

Figure 18 shows the cumulative returns of our filtered portfolio versus the standard unfiltered portfolio. The overlay model has outperformed the standard model with Williams %R consistently since 1994 and has done so with a lower turnover level of 101% 2-way per month, vs. 148%. The overlay portfolio produced an increase in after cost information ratio from 0.39 to 0.66 from the standard model with Williams %R.

Fig 18 Cumulative returns for enhanced portfolio

Fig 19 Monthly Q1-Q5 two way turnover

0

50

100

150

200

250

300

350

400

450

Dec-

93

Dec-

94

Dec-

95

Dec-

96

Dec-

97

Dec-

98

Dec-

99

Dec-

00

Dec-

01

Dec-

02

Dec-

03

Dec-

04

Dec-

05

Dec-

06

Dec-

07

Dec-

08

Standard Model (After Costs)

Standard Model With Williams (After Costs)

Standard Model With Williams Overlay (30bps Tcost)

0%

20%

40%

60%

80%

100%

120%

140%

160%

Standard Model Standard Model With Williams

Standard Model With Williams Overlay

Source: FactSet, Macquarie Research, April 2009 Source: FactSet, Macquarie Research, April 2009

Fig 20 Annualised returns of model portfolios

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

Ann

ual R

etur

n

Standard ModelStandard Model With WilliamsStandard Model With Williams Overlay

Source: FactSet, Macquarie Research, April 2009


30 April 2009 14

References Andersen, T.G., Bollerslev, T., “Answering the sceptics: yes, standard volatility models do provide accurate Forecasts”, International Economic Review Vol 39, (1998)

Bessembinder, H., Chan, K., “Market efficiency and the returns to technical analysis. Financial Management” Vol 272 (1998).

Brock, W., Lakonishok, J., LeBaron, B., “Simple technical trading rules and the stochastic properties of stock returns” Journal of Finance vol 485 (1992)

Davies, C. “Quantamentals: Do Tehcnicals add value?”, Macquarie Research Equities, (2009)

De Souza, R., “Macquarie Alpha Model: The alpha model goes under the knife”, Macquarie Research Equities, (2009)

Fong, W., Yong, L. “Chasing trends: recursive moving average trading rules and Internet stocks” Journal of Empirical Finance Vol 12 (2005)

Kavajecz, K.A., Odders-White, E.R., “Technical analysis and liquidity provision” Review of Financial Studies Vol 17 (2005)

Kwon, K-Y., Kish, R.J., “A comparative study of technical trading strategies and return predictability: an extension of Brock, Lakonishok, and LeBaronk (1992) using NYSE and NASDAQ indices” Quarterly Review of Economics and Finance Vol 42 (2002)

Lee, D.D., Chan, H., Faff, R.W., Kalev, “Short-term contrarian investing — it is profitable — yes and no” Journal of Multinational Financial Management Vol 13 (2003)

Lo, A.W., Mamaysky, H., Wang, J., “Foundations of technical analysis: computational algorithms, statistical inference, and empirical implementation” Journal of Finance Vol 55 (2000)

Marshall, B.R., Cahan, R.H., Cahan, J.M. “Does intraday technical analysis in the U.S. equity market have value?” Journal of Empirical Finance Vol 15 (2008)

Ready, M.J., “Profits from technical trading rules” Financial Management Vol 313 (2002).

Sullivan, R., Timmermann, A., White, H., “Data-snooping, technical trading rule performance, and the bootstrap” Journal of Finance Vol 245 (1999)


30 April 2009 15

Appendices – factor definitions

Fig 21 Technical definitions

Relative Strength Index Developed by J. Welles Wilder in 1978 (New Concepts in Technical Trading Systems), it measures price relative to past prices. The indicator compares the magnitude of a stock’s recent gains to its recent losses and in effect measures the momentum behind recent price movements. That is it measures the internal price strength of a security. RSI = 100 – 100/(1+RS) RS = abs((Total gains over n periods/n) / (Total losses over n periods/n)) When the RSI falls below 30, it is deemed as a buy and when it rises above 70 it is deemed as sell. However, many practitioners will wait for the signal to turn. In terms of testing the signal given a high score is deemed as overbought we multiply the RSI by -1 so that overbought stocks score lowly. Additionally, given that a stock may continuously trade well within the bounds of the standard cut offs, we also look at the current level of the RSI of a stock relative to its own long term average and variance that is we normalise the RSI for each individual stock – we test on a rolling window of 3, 6 and 12 months. Williams %R Developed by Larry Williams, it measures the relative location of the closing price in relation to the high-low range over a set time period with the intention to identify over extended moves in a non-trending market. The indicator will typically peak and turn down a few days before the price peak of the underlying security and vice versa when forming a trough. It is calculated by first determining the highest high and lowest low over the preceding n periods, then taking the ratio of the highest high minus the current price divided by the highest high minus the lowest low. Values range between -100 and 0. W = -100 * ((Hn – Tc)/(Hn-Ln)) Hn = highest price in n periods; Ln = lowest price in n periods, Tc = today’s Close Values between 0 & -20 it is deemed as overbought; between -80 and -100 it is deemed as oversold. Like the RSI we once again multiply the score by -1 to make sure high scoring (overbought stocks close to 0) rank lowly. Additionally, we also test a normalised version of the signal on a 3-, 6- and 12-month basis. MACD - Moving Average Convergence/Divergence Developed by Gerald Appel in the 1960s, the MACD (Moving Average Convergence/Divergence) shows changes in relative movement of shorter and longer moving averages based upon a price series. The indicator has both trending and oscillator characteristics and tends to give less false signals than the faster RSI and Stochastic Oscillator. The MACD Line is the difference between the longer exponentially weighted moving average (typically 26 days) and the shorter exponentially weighted moving average (typically 12 days). A trigger line is then 9-day exponentially weighted moving average of the MACD. In 1986, Thomas Aspray added the MACD histogram which is simply a graphical representation of the difference between the MACD and trigger line. A buy signal is deemed to arise when the MACD rises through and above the trigger line, and the opposite is true for a sell signal. Given that we’re measuring this as a style factor, we take the view the more positive the difference (or the more positive the histogram) the more overbought a stock. Consequently, we multiply the MACD divergence by -1 so that a large divergence scores badly. We also test this factor on a normalised basis relative to a stocks own history over 3-, 6- and 12-month periods. Stochastic Oscillator Developed by George Lane in the 1950s, this metric is a momentum-based price oscillator which aims to catch tops and bottoms early. When prices close near to their highs, it is an indication of strength (bull) and vice versa when near lows (bear). In practice two oscillators are calculated. These are more commonly referred to the fast (%K) and slow (%D). The %K is the same as the Williams %R but ranges from 0 to 100 using the formula below: %K = ((Tc – Ln)/(Hn-Ln))*100 Hn = highest price in n periods; Ln = lowest price in n periods, Tc = today’s Close The %D is a simple moving average of the %K using an n day window For the purposes of our research we create a ratio of the %K relative to %D. When this ratio is positive then this is deemed as attractive and vice versa. Like the other metrics we also normalise the signal relative to a stocks own history and using a rolling 3-, 6- and 12-month window.



30 April 2009 16

Fig 21 Technical definitions (continued)

Commodity Channel Index Developed by Donald Lambert initially for use on commodities, it now is frequently used on equities and currencies. The CCI is calculated as the difference between the typical price and its simple moving average, divided by the mean of the typical price. The index is usually scaled by a factor of 1/0.015 to provide more readable numbers. CCI = (1/0.015) * (Typical Price – Simple Moving Average of Typical Price)/Mean Deviation Typical Price = average of high, low and close price Mean Deviation = Average(Abs(SMA Typical Price today – Typical Price for past n periods)) It is used to identify divergences and therefore acts as an overbought/oversold indicator. It typically oscillates around 0 ranging between +/- 100. When it rises above 100 it implies and overbought position and when it falls below -100 it represents an oversold position. Money Flow Index Similar to On Balance Volume, the Money Flow Index also incorporates the momentum characteristics based upon the relative location of the price. It is interpreted in a similar manner to the RSI in that its value ranges between 0 and 100 and a value < 20 is oversold while a value > 80 is overbought. It is calculated as follows Typical Price = (High Price + Low Price + Close Price)/3 Money Flow = Typical Price * Volume Money Ratio = Positive Money Flow/Negative Money Flow – where positive money flow is when today’s typical price > yesterday’s typical price. The Positive Money Flow is then the sum of that series over the past n periods. Negative money flow is when today’s typical price < yesterday’s typical price. The Money Flow Index = 100 – (100/(1+Money Ratio)) Like the RSI we multiply this signal by * -1 so that a high score (>80 overbought) actually ranks lowly. Additionally, we also test a normalised version of the signal over a 3-, 6- and 12-month basis. Chaikin Money Flow Developed by Marc Chaikin, this indicator is an enhancement of the Accumulation / Distribution Index. It simply takes the ADI and then scales this by the total volume traded over n periods (traders typically use 21 days). In terms of usage, the basic rule of thumb is that is the value is > 0 then this is deemed as bullish and < 0 is bearish. If the value is more that +/- 0.25 this is viewed as being a strong trend. A divergence can show up when the CMF makes a new high but the prices action creates a new flow. This implies that there is less selling pressure pushing the stock lower and therefore a bounce could occur. On Balance Volume Popularised by J Granville in his 1963 book Granville's New Key to Stock Market Profits, the metric was first investigated by Woods & Vignolia in 1940. The indicator relates volume to price change and indicates if volume is flowing into or out of a security. It adds and subtracts volume to a running total depending on whether price moves up or down. It is often used as an indicator of stocks under accumulation for a possible take over. The rationale for this is that potential bidders/acquirers of large stock positions tend to place bids just below the market and only take the ask price when a sizeable block is offered. Consequently, the indicator moves higher during a time when the stock price remains relatively flat. In terms of analysis investors should be wary of simply looking at the raw level and actually the direction tends to be more important as it indicated the flow. The metric is viewed as leading indicator of price trends and consequently a divergence between the indicator and price could signal that a reversal may about to occur. Having looked at the strategy it appears that like other metrics above it actually acts as a contrarian indicator in that when the OBV reaches new highs there are likely to be market participants taking profits. Consequently when we look at the signal and normalise it relative to a stocks own history we multiply the factor by -1.

Force Index Developed by Alexander Elder for intra day data this indicator measures the bullish (bearish) force during each upward (downward) movement - the force of each movement is defined by its trend, range and volume. While the index can be used on its own, practitioners typically prefer to use a moving average. A short moving average is used to identify when to open and close positions, while a longer moving average is used to show trends and their changes. We have adapted the calculation to use daily data as opposed to tick data. Force = (Price Today – Price Yesterday) * Volume Given the above, If the close of the current is > close of previous then force is positive and if the close of the current is < close of previous then force is negative. Additionally, the large the value can be either be driven by the greater difference in price or a greater the transaction volume. Like all oscillators, this indicator indicates a buy when it rises above the centre line and trend lines can be applied to determine the strength of the movement. Like the other oscillators, we multiply this factor by -1 so that a high score scores badly.



30 April 2009 17

Fig 21 Technical definitions (continued)

Accumulation / Distribution Index Developed by Marc Chaikin, this indicator is a cumulative total volume indicator which uses both prices and volume. The volume added each day is a function of the relative strength of price within the day. The nearer the closing price to its high of the day, the higher the added share of volume will be added and vice versa. This weighting coefficient is calculated as: ((Close – Low) – (High – Close)) / (High – Low) - This therefore ranges between +/-1. Consequently the incremental value added to the index on any given day is as follows: AD = (((Close – Low) – (High – Close)) / (High – Low)) * Volume When this indicator grows, it indicates that there is accumulation of a particular security as the majority of the share of the share of the volume is related to an upward trend in prices. Average Directional Movement Index (ADX) Developed by Wilder, the ADX attempts to evaluate the strength of a current trend, be it up or down. Although market direction is integral to the calculation of the ADX, it itself is not a directional indicator and therefore should be used in conjunction with other indicators. The ADX is derived from two other indicators the positive directional indicator (+DI) and negative directional indicator (-DI). When the +DI rises above the –DI, then a buy signal is deemed. The ADI is then the ratio of +DI/-DI on a moving average basis. The scale of the ADX ranges from 0 to 100 with low values (<20) indicating a weak trend and high (>40) a strong trend. When the signal moves up and through 20, it indicates that a trading range is potentially ending and a trend is forming. When declining from above 40, this can indicate that a trend could be slowing and a trading range may be developing. Average True Range Developed by J. Welles Wilder, this metric dynamically measures the typical range between prices (high, low, close) over a set period. The True range is the greater of the following three measures: Today’s High – Today’s Low Absolute( Today’s High – Yesterday’s Close) Absolute( Today’s Low – Yesterday’s Close) The Average True Range is the moving average of the true range values. The basic interpretation is the greater the value, the greater the possibility for a trend reversal; the smaller the value, the weaker the trend.

Bollinger Bands Developed by John Bollinger, these bands are used to visually judge the relationship between price and volatility changes. Bollinger bands consist of a centre line and an upper and lower band around that centre line. The centre line is a moving average, which could be simple or exponential, and we use a simple moving average of 20 days which is in line with providers such as FactSet and Bloomberg. The bands are then formed by calculating two standard deviations over the same period which are then added and subtracted from the centre line. They differ from simple bands around a mean in that in trending markets the bands will tend to expand, while in range-bound markets they will contract. The theory is that in range bound markets resistance should be found at the upper and lower levels. However, should the price break through, then this could indicate a breakout. Generally, Bollinger Bands can be useful to identify buy and sell signals but are not in themselves designed to determine the future direction of a security. They do, however, help identify periods of high and low volatility (bandwidth) and when prices are extreme and potentially unsustainable.



30 April 2009 18

Technical factor performance Williams %R - Time Series NormalisedInformation Coefficient




IC HorizonAverage IC

1 month 5.6%3 month 1.4%6 month -0.3%12 month -1.8%24 month -1.5%

Long Short Q1-Q5



Quintile 1 Excess Return * 4.4% 3.8% 3.5% 3.7%Quintile 5 Excess Return * -8.3% -10.4% -11.9% -23.3%Q1 - Q5 Return (p.a.) # 13.2% 15.1% 16.4% 32.1%Q1 - Q5 Volatility (p.a.) # 10.2% 10.1% 12.2% 19.6%Q1 - Q5 Information Ratio # 1.29 1.50 1.34 1.64Q1 - Q5 Turnover 1 way p.a.# 1913% 1934% 1922% 1934%* Monthly rebalanced annualised excess returns (long - market and market - short)# Monthly rebalanced Long Q1 - Short Q5

Long Short Q1-Q5 Rolling Excess Returns



Q1 - Q5 Return (p.a.) 13.1% 15.1% 16.4% 32.1%Q1 - Q5 Volatility (p.a.) 10.2% 10.1% 12.2% 19.6%Q1 - Q5 Information Ratio 1.28 1.50 1.34 1.64Q1 - Q5 Turnover (p.a. 1 way) 1913% 1934% 1922% 1934%

Long Short Q1-Q5 Turnover 1-way

All History Last 5 Years Last 3 Years Last 1 Year

Average Annual L/S Turnover 1913% 1934% 1922% 1934%

Pure Factor Return



Average pure factor return (p.a.) 5.11% 5.11% 6.23% 9.68%Pure Factor Volatility 2.82% 2.88% 3.44% 5.20%Pure Factor Sharpe 1.81 1.78 1.81 1.86

Source: IBES, MSCI, and Macquarie Research, April 2009

-40%-30%-20%-10%

0%10%20%30%40%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Info

rmat

ion

Coe

ffici

ent


0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Exce

ss M

arke

t Ret

urn

Top Quintile Bottom Quintile

-10%

0%

10%

20%

30%

40%

50%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Long

Sho

rt R

etur

ns


-4.0%-2.0%0.0%2.0%4.0%6.0%8.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Info

rmat

ion

Coe

ffici

ent

IC Decay Accumulated IC

-4%

-2%

0%

2%

4%

6%

8%

10%

12%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Fact

or R

etur

n

Factor Return Annualised Factor Return

0%

500%

1000%

1500%

2000%

2500%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Turn

over

Turnover 12M Rolling Turnover


30 April 2009 19

Williams %R - Time Series Normalised - continuedFractiles

Annual Return

Excess Return Volatility Informatio

n RatioQ1 14.3% 3.8% 17.0% 0.22Q2 12.3% 1.7% 16.1% 0.11Q3 12.2% 1.7% 16.2% 0.11Q4 10.2% -0.3% 15.0% -0.02Q5 4.7% -5.8% 15.8% -0.36Market 10.5%

Fractile Cross sectional Hit Rate



Q1 53.4% 52.5% 52.7% 52.0%Q2 50.2% 51.3% 51.0% 51.6%Q3 50.6% 51.5% 50.9% 53.1%Q4 48.6% 48.5% 48.8% 50.5%Q5 45.9% 45.7% 46.0% 42.2%

Size

Average IC All History Last 5 Years

Last 3 Years Last 1 Year Latest No. of

StocksLarge 6.15% 5.68% 6.40% 7.35% 94Small 5.13% 5.30% 4.07% 5.37% 191All 5.63% 5.42% 4.76% 5.87% 285

Std Dev IC All History Last 5 Years


Small 13.06% 12.85% 12.88% 14.90%Large 12.09% 8.99% 9.09% 12.91%All 9.61% 7.93% 8.68% 12.31%

Sector


Last 3 Years

Latest Number of

StocksEnergy 3.06% 4.64% 8.92% 32Materials 5.97% 1.36% -0.58% 57Industrials 5.93% 8.30% 7.85% 45Consumer Discretionary 5.44% 5.77% 7.43% 31Consumer Staples 7.91% 3.27% 4.13% 16Health Care 4.32% 8.94% 8.85% 10Financials 8.39% 6.21% 6.82% 73Information Tech 7.49% 15.93% 15.69% 3Telecomm. Services 13.53% 9.83% 2.50% 3Utilities 9.99% 11.56% 14.37% 15

SeasonalityInformation Coefficient IC Std Dev Samples

January 10.9% 9.1% 15February 1.2% 11.2% 15March 1.9% 9.2% 15April 7.4% 5.4% 14May 6.9% 11.5% 14June 8.8% 9.3% 14July 9.8% 11.1% 14August 2.6% 11.7% 14September 3.7% 8.7% 14October 7.7% 7.8% 14November 2.9% 7.0% 14December 3.9% 8.3% 14

Factor Coverage



Average number of stocks 228 291 322 319

Coverage of available stocks (%) 99% 99% 99% 100%


050

100150200250300350400

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Num

ber o

f Sto

cks

0%10%20%30%40%50%60%70%80%90%100%

% o

f Mar

ket C

over

ed

Stocks with Factor % Coverage of Universe% Coverage by Market Cap

0%

2%

4%

6%

8%

10%

12%

Janu

ary

Februa

ryMarc

hApri

lMay

June Ju

ly

Augus

t

Septem

ber

Octobe

r

Novem

ber

Decem

ber

Info

rmat

ion

Coe

ffici

ent

0

1

10

100

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Cum

ulat

ive

Ret

urn

Q1 Q2 Q3 Q4 Q5 Market

40.0%42.0%44.0%46.0%48.0%50.0%52.0%54.0%56.0%58.0%60.0%

Q1 Q2 Q3 Q4 Q5%

Cro

ss s

ectio

nal h

it ra

te


-2%0%2%4%6%8%

10%12%14%16%18%

Ener

gy

Mat

eria

ls

Indu

stria

ls

Con

sum

erD

iscr

etio

nary

Con

sum

erSt

aple

s

Hea

lth C

are

Fina

ncia

ls

Info

rmat

ion

Tech

Tele

com

m.

Serv

ices

Util

ities

Info

rmat

ion

Coe

ffici

ent

All History Last 5 Years Last 3 Years

0.0%1.0%2.0%3.0%4.0%5.0%6.0%7.0%8.0%

Large Small All

Info

rmat

ion

Coe

ffici

ent



30 April 2009 20

Force Index - Time Series Normalised






1 month 4.4%3 month 2.2%6 month 1.0%12 month -0.1%24 month -0.3%

Long Short Q1-Q5











Pure Factor Return





-40%-30%-20%-10%

0%10%20%30%40%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Info

rmat

ion

Coe

ffici

ent


0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Exce

ss M

arke

t Ret

urn


-15%-10%-5%0%5%

10%15%20%25%30%35%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Long

Sho

rt R

etur

ns


-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Info

rmat

ion

Coe

ffici

ent


-4%

-2%

0%

2%

4%

6%

8%

10%

12%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Fact

or R

etur

n


0%

500%

1000%

1500%

2000%

2500%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Turn

over



30 April 2009 21

Force Index - Time Series Normalised - continuedFractiles

Annual Return


n RatioQ1 16.3% 5.8% 15.7% 0.37Q2 12.2% 1.6% 16.9% 0.10Q3 9.2% -1.3% 15.3% -0.09Q4 7.5% -3.0% 16.6% -0.18Q5 8.1% -2.4% 15.5% -0.16Market 10.5%




Q1 53.6% 52.8% 52.7% 54.1%Q2 50.0% 51.0% 50.8% 48.0%Q3 49.1% 48.9% 48.5% 50.4%Q4 47.9% 48.1% 48.2% 46.5%Q5 48.0% 48.6% 49.3% 50.6%

Size






Small 12.16% 12.38% 13.76% 14.48%Large 10.73% 7.89% 6.71% 8.74%All 8.66% 7.25% 7.16% 9.09%

Sector


Last 3 Years

Latest Number of

StocksEnergy 3.40% 3.89% 4.41% 32Materials 5.25% 2.47% 0.50% 56Industrials 6.35% 3.63% 2.40% 45Consumer Discretionary 4.78% 5.84% 7.12% 31Consumer Staples 5.36% -0.23% -1.89% 16Health Care 8.36% 10.66% 8.97% 10Financials 6.01% 5.38% 5.65% 71Information Tech 8.56% 19.09% 10.40% 3Telecomm. Services 3.58% 2.00% 10.28% 3Utilities 1.00% 11.79% 13.60% 15



Factor Coverage






050

100150200250300350400

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Num

ber o

f Sto

cks

0%10%20%30%40%50%60%70%80%90%100%

% o

f Mar

ket C

over

ed


0%1%2%3%4%5%6%7%8%9%

Janu

ary

Februa

ryMarc

hApri

lMay

June Ju

ly

Augus

t

Septem

ber

Octobe

r

Novem

ber

Decem

ber

Info

rmat

ion

Coe

ffici

ent

0

1

10

100

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Cum

ulat

ive

Ret

urn


40.0%42.0%44.0%46.0%48.0%50.0%52.0%54.0%56.0%58.0%60.0%

Q1 Q2 Q3 Q4 Q5%

Cro

ss s

ectio

nal h

it ra

te


-5%

0%

5%

10%

15%

20%

25%

Ener

gy

Mat

eria

ls

Indu

stria

ls

Con

sum

erD

iscr

etio

nary

Con

sum

erSt

aple

s

Hea

lth C

are

Fina

ncia

ls

Info

rmat

ion

Tech

Tele

com

m.

Serv

ices

Util

ities

Info

rmat

ion

Coe

ffici

ent


0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

Large Small All

Info

rmat

ion

Coe

ffici

ent



30 April 2009 22

Bollinger Band - Time Series NormalisedInformation Coefficient






Long Short Q1-Q5











Pure Factor Return





-40%-30%-20%-10%

0%10%20%30%40%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Info

rmat

ion

Coe

ffici

ent


0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Exce

ss M

arke

t Ret

urn


-10%-5%0%5%

10%15%20%25%30%35%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Long

Sho

rt R

etur

ns


-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Info

rmat

ion

Coe

ffici

ent


-4%

-2%

0%

2%

4%

6%

8%

10%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Fact

or R

etur

n


0%

500%

1000%

1500%

2000%

2500%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Turn

over



30 April 2009 23

Bollinger Band - Time Series Normalised - continuedFractiles

Annual Return






Q1 53.3% 52.1% 51.4% 50.9%Q2 50.0% 51.6% 51.6% 51.8%Q3 49.5% 51.0% 51.6% 52.3%Q4 49.7% 48.3% 49.4% 50.7%Q5 46.5% 46.5% 45.7% 43.8%

Size






Small 13.49% 14.36% 14.73% 15.24%Large 11.72% 9.70% 9.41% 12.89%All 10.07% 9.04% 9.66% 12.72%

Sector


Last 3 Years

Latest Number of

StocksEnergy 2.89% 3.50% 7.10% 32Materials 5.07% 2.36% 2.23% 57Industrials 5.45% 6.53% 6.40% 45Consumer Discretionary 5.28% 3.76% 4.68% 31Consumer Staples 7.45% 2.96% 4.74% 16Health Care 2.49% 10.04% 9.36% 10Financials 8.13% 5.79% 6.63% 73Information Tech 2.73% 13.33% 18.67% 3Telecomm. Services 4.62% 2.67% 7.22% 3Utilities 5.55% 11.60% 13.20% 15


January 9.4% 11.9% 15February -0.2% 12.1% 15March 1.6% 9.3% 15April 5.4% 6.9% 14May 6.4% 10.9% 14June 7.1% 11.2% 14July 9.0% 11.3% 14August 1.6% 10.2% 14September 1.4% 8.7% 14October 7.7% 8.3% 14November 1.8% 6.7% 14December 2.2% 8.5% 14

Factor Coverage






050

100150200250300350400

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Num

ber o

f Sto

cks

0%10%20%30%40%50%60%70%80%90%100%

% o

f Mar

ket C

over

ed


-2%

0%

2%

4%

6%

8%

10%

Janu

ary

Februa

ryMarc

hApri

lMay

June Ju

ly

Augus

t

Septem

ber

Octobe

r

Novem

ber

Decem

ber

Info

rmat

ion

Coe

ffici

ent

0

1

10

100

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Cum

ulat

ive

Ret

urn


40.0%42.0%44.0%46.0%48.0%50.0%52.0%54.0%56.0%58.0%60.0%

Q1 Q2 Q3 Q4 Q5%

Cro

ss s

ectio

nal h

it ra

te


0%2%4%6%8%

10%12%14%16%18%20%

Ener

gy

Mat

eria

ls

Indu

stria

ls

Con

sum

erD

iscr

etio

nary

Con

sum

erSt

aple

s

Hea

lth C

are

Fina

ncia

ls

Info

rmat

ion

Tech

Tele

com

m.

Serv

ices

Util

ities

Info

rmat

ion

Coe

ffici

ent


0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

Large Small All

Info

rmat

ion

Coe

ffici

ent



30 April 2009 24

Commodity Channel Index - Time Series NormalisedInformation Coefficient






Long Short Q1-Q5











Pure Factor Return





-40%-30%-20%-10%

0%10%20%30%40%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Info

rmat

ion

Coe

ffici

ent


0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Exce

ss M

arke

t Ret

urn


-15%-10%

-5%0%5%

10%15%20%25%30%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Long

Sho

rt R

etur

ns


-3.0%-2.0%-1.0%0.0%1.0%2.0%3.0%4.0%5.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Info

rmat

ion

Coe

ffici

ent


-4%

-2%

0%

2%

4%

6%

8%

10%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Fact

or R

etur

n


0%

500%

1000%

1500%

2000%

2500%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Turn

over



30 April 2009 25

Commodity Channel Index - Time Series Normalised - continuedFractiles

Annual Return






Q1 53.0% 52.2% 53.0% 54.2%Q2 50.4% 52.3% 51.5% 50.7%Q3 49.1% 49.7% 50.0% 50.9%Q4 49.2% 48.9% 48.6% 48.7%Q5 47.2% 46.6% 46.6% 45.2%

Size






Small 12.59% 13.18% 12.63% 13.04%Large 11.03% 8.75% 9.04% 13.93%All 9.12% 8.01% 8.51% 12.59%

Sector


Last 3 Years

Latest Number of

StocksEnergy 1.59% 2.29% 6.98% 32Materials 5.01% 0.90% -1.85% 56Industrials 4.65% 9.19% 8.63% 45Consumer Discretionary 6.06% 4.32% 5.67% 31Consumer Staples 6.04% 1.43% 4.68% 16Health Care 3.83% 11.72% 13.65% 10Financials 5.97% 4.11% 4.97% 73Information Tech 0.85% 10.80% 20.74% 3Telecomm. Services 2.23% -7.17% -3.61% 3Utilities 3.85% 10.75% 12.69% 15


January 11.2% 8.2% 15February -0.6% 8.6% 15March -0.8% 8.9% 15April 4.6% 5.6% 14May 7.2% 11.5% 14June 6.7% 7.7% 14July 4.5% 10.0% 14August 2.1% 11.3% 14September 2.3% 9.6% 14October 6.5% 7.5% 14November 2.9% 7.2% 14December 2.4% 7.2% 14

Factor Coverage






050

100150200250300350400

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Num

ber o

f Sto

cks

0%10%20%30%40%50%60%70%80%90%100%

% o

f Mar

ket C

over

ed


-2%

0%

2%

4%

6%

8%

10%

12%

Janu

ary

Februa

ryMarc

hApri

lMay

June Ju

ly

Augus

t

Septem

ber

Octobe

r

Novem

ber

Decem

ber

Info

rmat

ion

Coe

ffici

ent

0

1

10

100

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Cum

ulat

ive

Ret

urn


40.0%42.0%44.0%46.0%48.0%50.0%52.0%54.0%56.0%58.0%60.0%

Q1 Q2 Q3 Q4 Q5%

Cro

ss s

ectio

nal h

it ra

te


-10%-5%0%5%

10%15%20%25%

Ener

gy

Mat

eria

ls

Indu

stria

ls

Con

sum

erD

iscr

etio

nary

Con

sum

erSt

aple

s

Hea

lth C

are

Fina

ncia

ls

Info

rmat

ion

Tech

Tele

com

m.

Serv

ices

Util

ities

Info

rmat

ion

Coe

ffici

ent


0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

Large Small All

Info

rmat

ion

Coe

ffici

ent



30 April 2009 26

Relative Strength Indicator - 14Days - Time Series Normalised







Long Short Q1-Q5











Pure Factor Return





-40%-30%-20%-10%

0%10%20%30%40%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Info

rmat

ion

Coe

ffici

ent


0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Exce

ss M

arke

t Ret

urn


-20%-15%-10%

-5%0%5%

10%15%20%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Long

Sho

rt R

etur

ns


-4.0%

-2.0%

0.0%

2.0%

4.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Info

rmat

ion

Coe

ffici

ent


-4%

-2%

0%

2%

4%

6%

8%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Fact

or R

etur

n


0%

500%

1000%

1500%

2000%

2500%

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Turn

over



30 April 2009 27

Relative Strength Indicator - 14Days - Time Series Normalised - continuedFractiles

Annual Return






Q1 52.4% 51.7% 51.5% 51.9%Q2 50.4% 51.9% 50.9% 52.0%Q3 49.8% 50.4% 50.8% 49.2%Q4 49.1% 49.5% 50.0% 49.6%Q5 47.3% 46.1% 46.4% 46.8%

Size






Small 13.88% 15.00% 16.53% 19.91%Large 11.56% 10.12% 9.09% 11.97%All 10.35% 9.20% 9.50% 12.33%

Sector


Last 3 Years

Latest Number of

StocksEnergy -0.61% 1.73% 5.09% 32Materials 3.21% 0.88% 0.38% 56Industrials 4.85% 6.06% 4.87% 45Consumer Discretionary 5.46% 2.80% 2.42% 31Consumer Staples 5.48% 4.16% 4.73% 16Health Care 1.24% 9.52% 7.96% 10Financials 7.10% 4.94% 6.29% 73Information Tech 0.42% 7.20% 16.04% 3Telecomm. Services 1.44% 12.50% 5.56% 3Utilities 4.51% 8.01% 7.41% 15



Factor Coverage






050

100150200250300350400

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Num

ber o

f Sto

cks

0%10%20%30%40%50%60%70%80%90%100%

% o

f Mar

ket C

over

ed


0%1%2%3%4%5%6%7%8%9%

10%

Janu

ary

Februa

ryMarc

hApri

lMay

June Ju

ly

Augus

t

Septem

ber

Octobe

r

Novem

ber

Decem

ber

Info

rmat

ion

Coe

ffici

ent

0

1

10

Dec 94

Dec 95

Dec 96

Dec 97

Dec 98

Dec 99

Dec 00

Dec 01

Dec 02

Dec 03

Dec 04

Dec 05

Dec 06

Dec 07

Dec 08

Cum

ulat

ive

Ret

urn


40.0%42.0%44.0%46.0%48.0%50.0%52.0%54.0%56.0%58.0%60.0%

Q1 Q2 Q3 Q4 Q5%

Cro

ss s

ectio

nal h

it ra

te


-2%0%2%4%6%8%

10%12%14%16%18%

Ener

gy

Mat

eria

ls

Indu

stria

ls

Con

sum

erD

iscr

etio

nary

Con

sum

erSt

aple

s

Hea

lth C

are

Fina

ncia

ls

Info

rmat

ion

Tech

Tele

com

m.

Serv

ices

Util

ities

Info

rmat

ion

Coe

ffici

ent


0.0%1.0%2.0%3.0%4.0%5.0%6.0%7.0%8.0%9.0%

Large Small All

Info

rmat

ion

Coe

ffici

ent



30 April 2009 28

Important disclosures: Recommendation definitions Macquarie - Australia/New Zealand Outperform – return >5% in excess of benchmark return Neutral – return within 5% of benchmark return Underperform – return >5% below benchmark return Macquarie – Asia/Europe Outperform – expected return >+10% Neutral – expected return from -10% to +10% Underperform – expected return <-10% Macquarie First South - South Africa Outperform – expected return >+10% Neutral – expected return from -10% to +10% Underperform – expected return <-10% Macquarie - Canada Outperform – return >5% in excess of benchmark return Neutral – return within 5% of benchmark return Underperform – return >5% below benchmark return Macquarie - USA Outperform (Buy) – return >5% in excess of benchmark return (Russell 3000) Neutral (Hold) – return within 5% of benchmark return (Russell 3000) Underperform (Sell)– return >5% below benchmark return (Russell 3000) Recommendations – 12 months Note: Quant recommendations may differ from Fundamental Analyst recommendations

Volatility index definition* This is calculated from the volatility of historical price movements. Very high–highest risk – Stock should be expected to move up or down 60–100% in a year – investors should be aware this stock is highly speculative. High – stock should be expected to move up or down at least 40–60% in a year – investors should be aware this stock could be speculative. Medium – stock should be expected to move up or down at least 30–40% in a year. Low–medium – stock should be expected to move up or down at least 25–30% in a year. Low – stock should be expected to move up or down at least 15–25% in a year. * Applicable to Australian/NZ/Canada stocks only

Financial definitions All "Adjusted" data items have had the following adjustments made: Added back: goodwill amortisation, provision for catastrophe reserves, IFRS derivatives & hedging, IFRS impairments & IFRS interest expense Excluded: non recurring items, asset revals, property revals, appraisal value uplift, preference dividends & minority interests EPS = adjusted net profit / efpowa* ROA = adjusted ebit / average total assets ROA Banks/Insurance = adjusted net profit /average total assets ROE = adjusted net profit / average shareholders funds Gross cashflow = adjusted net profit + depreciation *equivalent fully paid ordinary weighted average number of shares All Reported numbers for Australian/NZ listed stocks are modelled under IFRS (International Financial Reporting Standards).

Recommendation proportions – For quarter ending 31 March 2009 AU/NZ Asia RSA USA CA EUR Outperform 40.44% 49.55% 44.83% 38.49% 67.19% 43.84% Neutral 38.60% 15.57% 39.66% 46.43% 28.12% 39.04% Underperform 20.96% 34.88% 15.52% 15.08% 4.69% 17.12%

Analyst Certification: The views expressed in this research accurately reflect the personal views of the analyst(s) about the subject securities or issuers and no part of the compensation of the analyst(s) was, is, or will be directly or indirectly related to the inclusion of specific recommendations or views in this research. The analyst principally responsible for the preparation of this research receives compensation based on overall revenues of Macquarie Group Ltd ABN 94 122 169 279 (AFSL No. 318062 )(MGL) and its related entities (the Macquarie Group) and has taken reasonable care to achieve and maintain independence and objectivity in making any recommendations. Disclaimers: Macquarie Securities (Australia) Ltd; Macquarie Capital (Europe) Ltd; Macquarie Capital Markets Canada Ltd; Macquarie Capital Markets North America Ltd; Macquarie Capital (USA) Inc; Macquarie Capital Securities Ltd; Macquarie Capital Securities (Singapore) Pte Ltd; Macquarie Securities (NZ) Ltd; and Macquarie First South Securities (Pty) Limited are not authorised deposit-taking institutions for the purposes of the Banking Act 1959 (Commonwealth of Australia), and their obligations do not represent deposits or other liabilities of Macquarie Bank Limited ABN 46 008 583 542 (MBL) or MGL. MBL does not guarantee or otherwise provide assurance in respect of the obligations of any of the above mentioned entities. MGL provides a guarantee to the Monetary Authority of Singapore in respect of the obligations and liabilities of Macquarie Capital Securities (Singapore) Pte Ltd for up to SGD 35 million. This research has been prepared for the general use of the wholesale clients of the Macquarie Group and must not be copied, either in whole or in part, or distributed to any other person. If you are not the intended recipient you must not use or disclose the information in this research in any way. Nothing in this research shall be construed as a solicitation to buy or sell any security or product, or to engage in or refrain from engaging in any transaction. In preparing this research, we did not take into account the investment objectives, financial situation and particular needs of the reader. Before making an investment decision on the basis of this research, the reader needs to consider, with or without the assistance of an adviser, whether the advice is appropriate in light of their particular investment needs, objectives and financial circumstances. There are risks involved in securities trading. The price of securities can and does fluctuate, and an individual security may even become valueless. International investors are reminded of the additional risks inherent in international investments, such as currency fluctuations and international stock market or economic conditions, which may adversely affect the value of the investment. This research is based on information obtained from sources believed to be reliable but we do not make any representation or warranty that it is accurate, complete or up to date. We accept no obligation to correct or update the information or opinions in it. Opinions expressed are subject to change without notice. No member of the Macquarie Group accepts any liability whatsoever for any direct, indirect, consequential or other loss arising from any use of this research and/or further communication in relation to this research. Other Disclaimers: In Canada, securities research is prepared, approved and distributed by Macquarie Capital Markets Canada Ltd, a participating organisation of the Toronto Stock Exchange, TSX Venture Exchange & Montréal Exchange. Macquarie Capital Markets North America Ltd., which is a registered broker-dealer and member of FINRA, accepts responsibility for the contents of reports issued by Macquarie Capital Markets Canada Ltd in the United States and to US persons and any person wishing to effect transactions in the securities described in the reports issued by Macquarie Capital Markets Canada Ltd should do so with Macquarie Capital Markets North America Ltd. Securities research is issued and distributed by Macquarie Securities (Australia) Ltd (AFSL No. 238947) in Australia, a participating organisation of the Australian Securities Exchange; Macquarie Securities (NZ) Ltd in New Zealand, a licensed sharebroker and New Zealand Exchange Firm; Macquarie Capital (Europe) Ltd in the United Kingdom, which is authorised and regulated by the Financial Services Authority (No. 193905); Macquarie Capital Securities Ltd in Hong Kong, which is licensed and regulated by the Securities and Futures Commission; Macquarie Capital Securities (Japan) Limited in Japan, a member of the Tokyo Stock Exchange, Inc., Osaka Securities Exchange Co. Ltd, and Jasdaq Securities Exchange, Inc. (Financial Instruments Firm, Kanto Financial Bureau(kin-sho) No. 231, a member of Japan securities Dealers Association and Financial Futures Association of Japan); Macquarie First South Securities (Pty) Limited in South Africa, a member of the JSE Limited and in Singapore, Macquarie Capital Securities (Singapore) Pte Ltd (Company Registration Number: 198702912C), a Capital Markets Services licence holder under the Securities and Futures Act to deal in securities and provide custodial services in Singapore. Pursuant to the Financial Advisers (Amendment) Regulations 2005, Macquarie Capital Securities (Singapore) Pte Ltd is exempt from complying with sections 25, 27 and 36 of the Financial Advisers Act. Clients should contact analysts at, and execute transactions through, a Macquarie Group entity in their home jurisdiction unless governing law permits otherwise. Macquarie Capital (USA) Inc., which is a registered broker-dealer and member of FINRA, accepts responsibility for the content of each research report prepared by one of its non-US affiliates when the research report is distributed in the United States by Macquarie Capital (USA) Inc. Macquarie Capital (USA) Inc. affiliate research reports and affiliate employees are not subject to the disclosure requirements of FINRA rules. Any persons receiving this report directly from Macquarie Capital (USA) Inc. and wishing to effect a transaction in any security described herein should do so with Macquarie Capital (USA) Inc. The information contained in this document is confidential. If you are not the intended recipient, you must not disclose or use the information in this document in any way. If you received it in error, please tell us immediately by return e-mail and delete the document. We do not guarantee the integrity of any e-mails or attached files and are not responsible for any changes made to them by any other person. MGL has established and implemented a conflicts policy at group level (which may be revised and updated from time to time) (the "Conflicts Policy") pursuant to regulatory requirements (including the FSA Rules) which sets out how we must seek to identify and manage all material conflicts of interest. Disclosures with respect to the issuers, if any, mentioned in this research are available at www.macquarie.com/research/disclosures. © Macquarie Group Auckland Tel: (649) 377 6433

London Tel: (44 20) 3037 4400

Shanghai Tel: (86 21) 6841 3355

Bangkok Tel: (662) 694 7999

Manila Tel: (63 2) 857 0888

Singapore Tel: (65) 6231 1111

Calgary Tel: (1 403) 218 6650

Melbourne Tel: (613) 9635 8139

Sydney Tel: (612) 8232 9555

Hong Kong Tel: (852) 2823 3588

Montreal Tel: (1 514) 925 2850

Taipei Tel: (886 2) 2734 7500

Jakarta Tel: (62 21) 515 1818

Mumbai Tel: (91 22) 6653 3000

Tokyo Tel: (81 3) 3512 7900

Johannesburg Tel: (2711) 583 2000

Perth Tel: (618) 9224 0888

Toronto Tel: (1 416) 848 3500

Kuala Lumpur Tel: (60 3) 2059 8833

Seoul Tel: (82 2) 3705 8500

New York Tel: (1 212) 231 2500

Available to clients on the world wide web at www.macquarie.com/research and through Thomson Financial, FactSet, Reuters and Bloomberg.


30 April 2009 29

Notes

April 09

Research Heads of Equity Research David Rickards (Global Co – Head) (852) 2823 3538 John O’Connell (Global Co – Head) (612) 8232 7544

Consumer Staples Food & Beverages Greg Dring (612) 8232 3104

Consumer Discretionary Media and Tourism & Leisure Alex Pollak (612) 8232 3172 Retailing Warren Doak (New Zealand) (649) 363 1416 Greg Dring (612) 8232 3104

Energy Adrian Wood (612) 8232 8531

Financials Banks Ben Zucker (612) 8232 6089 Tom Quarmby (612) 8232 8668 Diversified Financials Deana Mitchell (612) 8232 4576 Insurance Tony Jackson (612) 8232 4442 Deana Mitchell (612) 8232 4576

Healthcare Steve Wheen (612) 8232 4130

Industrials Capital Goods Warren Doak (New Zealand) (649) 363 1416 Greg Dring (612) 8232 3104 John Purtell (612) 8232 8633

Industrials Transportation /Commercial Services Warren Doak (New Zealand) (649) 363 1416 Ian Myles (612) 8232 4157 Russell Shaw (612) 8232 7124 Transportation - Infrastructure Warren Doak (New Zealand) (649) 363 1416 Ian Myles (612) 8232 4157

Materials Chemicals/Containers, Packaging/Paper & Forest Products, Construction Materials Stephen Hudson (New Zealand) (649) 363 1414 John Purtell (612) 8232 8633 Global Metals & Mining Jim Copland (612) 8232 0397 Len Eldridge (618) 9224 0838 Brendan Harris (612) 8232 3575 Sophie Spartalis (612) 8232 5159

Real Estate Property Trusts & Developers Callum Bramah (612) 8232 7647 Paul Checchin (612) 8232 4197

Telecommunications Andrew Levy (612) 8232 5165

Utilities Stephen Hudson (New Zealand) (649) 363 1414 Gavin Maher (612) 8232 4151

Commodities & Precious Metals Jim Lennon (London) (44 20) 3037 4271 Adam Rowley (London) (44 20) 3037 4272

Emerging Leaders Adam Simpson (612) 8232 4439 Benn Skender (612) 8232 6846 Andrew Wackett (618) 9224 0867

Quantitative Martin Emery (Hong Kong) (852) 2823 3582 Scott Hamilton (612) 8232 3544 George Platt (612) 8232 6539

Data Services (Australia & New Zealand) Sheridan Duffy (612) 8232 9786

Economics and Strategy Tanya Branwhite (Strategy) (612) 8232 7628 Richard Gibbs (Head of Economics) (612) 8232 3935 Neale Goldston-Morris (Strategy) (612) 8232 7562 Brian Redican (Aus Economics) (612) 8232 7016 Mark Tierney (Int’l Economics) (612) 8232 3121

Find our research at Macquarie: www.macquarie.com.au/research Thomson: www.thomson.com/financial Reuters: www.knowledge.reuters.com Bloomberg: MAC GO Factset: http://www.factset.com/home.aspx Contact Gareth Warfield for access (612) 8232 3207 See and hear our analysts at www.macquarie.com.au/macquariedigital

Toll free from overseas Canada 1800 989 8159 Hong Kong 800 96 2049 Japan 0053 161 6437 New York 1888 622 7862 Singapore 800 616 1037

Email addresses [email protected] eg. [email protected]

Sales Equities Martin Dacron (Australia) (612) 8232 7421 Scott Dolling (Asia) (852) 2823 3705 Rob Fabbro (Continental Europe) (44 20) 7065 2031 Charles Nelson (UK) (44 20) 7065 2032 Duane O’Donnell (Melbourne) (613) 9635 9183 Dave Roberton (New Zealand) (649) 363 1498 Luke Sullivan (New York) (1 212) 231 2507 Stevan Vrcelj (Head of Global Sales) (612) 8232 5999

Specialist Sales Matthew Clegg (Index) (612) 8232 5653 Andrew Mouat (Property Trusts) (612) 8232 3151 George Platt (Quantitative) (612) 8232 6539 Phil Zammit (Emerging Leaders) (612) 8232 3122

Alternative Strategies Anthony Panaretto (Sales) (612) 8232 4500 Shannon Donohoe (Stock borrow & loan) (612) 8232 6997 Mark Donnelly (Equity finance) (612) 8232 7664 Cameron Duncan (Converts) (612) 8232 7405 Anthony Hourigan (Derivatives) (612) 8232 9884

Treasury & Commodities Gavin Bradley (Metals & Mining) (612) 8232 4248 Matthew Forgham (Metals & Mining) (44 20) 3037 4615 Emma Winspear (Futures) (613) 9635 8275 James Mactier (Metals & Mining) (618) 9224 0612 Ian Miller (Futures) (612) 8232 3555 Will Richardson (Foreign Exch) (612) 8232 4777

Syndication Peter Curry (612) 8232 4039 Paul Staines (612) 8232 7781

macquarie technicals

Documents