Macquarie Technicals

AUSTRALIA Quantitative analysis
30 April 2009 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

Inside In doing this analysis, we also found that there was a significant difference in
What goes up must come down 2 behaviour between large cap and small cap stocks.
Market instability hurts fundamental Large cap stocks (ASX100) tend to be mean reverting which favours the
factors 3 mean reverting technical indicators over trend following strategies.
Using Technical Indicators as Quant Small cap stocks (ex-ASX100) tend to be trend following, this penalises the
factors 4 mean trend Technical indicators and favours momentum based strategies.
Performance of Technical Indicators in Improving the investment process with technicals
Australia 5
Incorporating a technical trading signal in an existing investment process can
Combining with other factors and use in add significant value as these signals tend to be lowly correlated to the standard
a factor model 10 value, momentum, quality, growth and analyst sentiment strategies. However,
careful consideration must be given to the higher turnover of these signals which
References 14
makes them hard to include in a process.
Appendices – Factor definitions 15
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
Analysts as a ‘knock out’ signal in a standard quant model.
Burke Lau
612 8232 0481 burke.lau@macquarie.com In general, we find that the ‘knock out’ method was a good way to incorporate
John Conomos
612 8232 5157 john.conomos@macquarie.com the predictive strength and timing ability of the technical indicators while
George Platt moderating the effects of higher turnover.
612 8232 6539 george.platt@macquarie.com
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
Let’s get technical

Short term technical Defining technical indicators
indicators are
typically enhanced With the recent volatility in the market as well as weakness in the performance of some of the
mean reverting standard quant factors like value, we turn our attention to faster burn short term strategies
signals. 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% Information
Macquarie Alpha
Coefficient Model
30%
12M Momentum
20%
3M Earnings
Revisions
10%
Earnings Yield
0% Price relative to
52 Week High
-10%
Force Index
-20%
Williams %R
-30%
Technical Indicators
Bollinger Band
-40% Commodity
Channel Index
Ju 97
Ju 04
Ja 98
Au 99
Ja 05
Au 06
Ju 94
Fe l 95
Ap 96
Nor 97
M 99
Ju 01
Fe l 02
Ap 03
Nor 04
M 06
08
De 01
De 08
Se 96
O 00
Se 03
O r 07
M t 00
M t 07
Relative Strength
v
v
n
n
n
n
c
c
p
g
ay
ay
b
b
ar
a
c
c
De
Indicator
Monthly IC 12M Average IC Money Flow Index
Stochastic
Oscillator
All Last 5 Last 3 Last 1 On Balance
History Years Years Year Volume Small Cap - ex ASX100
1M Mean Large Cap - ASX100
Average IC 5.6% 5.4% 4.8% 5.9% Reversion
MACD
Std Deviation IC 9.6% 7.9% 8.7% 12.3%
Success Rate 73.7% 73.3% 63.9% 50.0% -4% -2% 0% 2% 4% 6% 8% 10% 12%
Avg IC / Std Dev IC 0.59 0.68 0.55 0.48 Information Coefficient
t-stat 7.68 5.29 3.29 1.65
Source: Macquarie Research, April 2009
Other key technical indicators that also performed well were:

Force Index Bollinger Band
Commodity Channel Index Relative Strength Indicator
Technical Indicators Large caps mean revert; small caps trend follow
perform strongly in Mean reverting technical indicators work significantly better in large cap stocks. The standard
the large cap space. 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.
30 April 2009 2
Market instability hurts fundamental factors

Equity markets have proved extremely challenging for investors since the foundations of the
Higher time series
market began to shake in the second half of 2007. As soaring levels of uncertainty persisted,
and cross sectional
fundamental factors entrenched in security valuation have had highly volatile performance.
volatility has
This volatility has made life exceptionally difficult for investors as strong performance for a
increased returns to
month or two is often followed by severe underperformance.
mean reveting
strategies. 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
% 50% Information
25 Coefficient
40%
20 30%
20%
15
10%
10
0%
5 -10%
-20%
0
-30%
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
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
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)
30 April 2009 3
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
We test the standard When testing the technical indicators as quant factors, a key characteristic to keep in mind is
signals 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
We time series and rank one stock against another, eg on balance volume is a running total of (volume *
normalise technical direction of price move), which typically grows over time is very different for large cap vs.
indicators to detect small cap stocks.
changes in the
signal relative to its To take this into account we normalise a signal against its own history (using the last 12
own stock history. 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.
30 April 2009 4
Performance of technical indicators in

Australia
Fig 6 Technical factor performance in Australia – Dec 1994 – Mar 2009
Information Coefficient Q1-Q5 Long Short
Largest 1 Largest 1 Turnover
IC Avg/Sd IC Success Return Volatility Informatio Success
Average IC Month Month (2 Way
t-stat Rate (p.a) (p.a.) n Ratio Rate
Gain Loss 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 Reversal
1 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 / Low
Price 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 Period
Standard 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%
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%
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%
Accumulation / Distribution Index
Standard 0.2% 0.07 50.3% 1.9% 11.5% 0.16 51.5% 13.1% -11.7% 3076%
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%
Average True Range (Inverted) - 10 Period
Standard 1.7% 0.55 52.0% 3.9% 12.3% 0.31 59.1% 19.8% -13.5% 1234%
Chankin Money Flow Index - 21 Period
Standard 0.3% 0.11 52.6% 2.4% 11.1% 0.22 54.4% 8.5% -12.0% 2829%
On Balance Volume
12 Month Normalised 0.7% 0.31 51.5% -2.3% 9.5% -0.24 41.5% 11.4% -7.4% 3761%
Trading Intensity
1M / 12M Average Daily Volume 4.7% 1.93 70.2% 18.1% 11.7% 1.55 68.4% 9.2% -16.8% 2723%
Relative Volatility
1M / 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 5
A key characteristic of the Australian market is the stellar performance of the 12 month price
Momentum momentum factor. This is where stocks with high (low) 12-month price momentum continue
strategies are to outperform (underperform). The intuition behind the momentum phenomenon in Australia
strongly predictive resides in the recurrent cash injections from compulsory superannuation. Managers continue
in Australia 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.
Simple 1 month The key mean reverting technical indicators which added value in the Australian market are
mean reversion shown in Figure 7.
doesn’t work in
Australia
Fig 7 Information coefficient of mean reverting technical indicators
Williams %R
Force Index
Bollinger Band
Commodity
Channel Index
Relative Strength
Indicator
Money Flow Index
Stochastic
Oscillator
On Balance
Volume
1M Mean
Reversion
-2% -1% 0% 1% 2% 3% 4% 5% 6%
Information Coefficient
Source:Macquarie Research, April 2009
30 April 2009 6
Williams %R stands The Williams %R signal stands out from this list of technical indicators. It has a strong
out from the crowd 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% Information 50% Long Short

Coefficient Returns
30% 40%
20%
30%
10%
0% 20%
-10% 10%
-20%
0%
-30%
-40% -10%
4
8
9
5
6
J 94
Ap 96
ar 9
J 01
Ap 03
a 6
08
7
4
5
2
Dey 01
c 8
Seb 96
O 00
p 3
O r 07
a 0
a 7
Juv 9
Juv 0
Jan 9
Aun 9
Ja 0
Aun 0
Juv 97
Juv 04
Ja 98
g 9
Ja 05
6
Mg9
Mg0
Nor 9
Nor 0
Dey 0
Ju 94
Ap 96
M 99
Ju 01
Ap 03
ar 6
08
Nor 97
Nor 04
Seb 0
5
Dey 01
b 2
Dey 08
Feul 9
Feul 0
Mct 0
Mct 0
Seb 96
O 00
Se 03
O 07
a 0
a 7
Aun 9
Aun 0
Mg0
Fe l 0
Fe l 9
M t0
M t0
n
c
c
p
n
De
c
c
p
p
ar
c
c
De
Monthly IC 12M Average IC LS Return 12M Rolling LS returns
Last 5 Last 3 Last 5 Last 3

All History Last 1 Year All History Last Year
Years Years Years Years
Average IC 5.6% 5.4% 4.8% 5.9% Quartile 1 Excess Return * 4.4% 3.8% 3.5% 3.7%
Std Deviation IC 9.6% 7.9% 8.7% 12.3% Quartile 5 Excess Return * -8.3% -10.4% -11.9% -23.3%
Success Rate 73.7% 73.3% 63.9% 50.0% Q1 - Q5 Return (p.a.) # 13.2% 15.1% 16.4% 32.1%
Avg IC / Std Dev IC 0.59 0.68 0.55 0.48 Q1 - Q5 Volatility (p.a.) # 10.2% 10.1% 12.2% 19.6%
Q1 - Q5 Information Ratio # 1.29 1.50 1.34 1.64
t-stat 7.68 5.29 3.29 1.65
Q1 - 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)
30 April 2009 7
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 Fig 11 Q1-Q3 L/S information ratio before costs in
large vs. small cap universes large vs small cap universe
Macquarie Alpha Macquarie Alpha
Model Model
12M Momentum 12M Momentum
3M Earnings 3M Earnings
Revisions Revisions
Earnings Yield Earnings Yield
Price relative to Price relative to
52 Week High 52 Week High
Force Index
Force Index
Williams %R
Williams %R
TechnicalIndicators
Bollinger Band
Technical Indicators
Bollinger Band
Commodity
Channel Index Commodity
Channel Index
Relative Strength
Indicator Relative Strength
Indicator
Money Flow Index
Money Flow Index
Stochastic
Oscillator Stochastic
On Balance Oscillator
Volume Small Cap - ex ASX100 On Balance
1M Mean Large Cap - ASX100 Volume Small Cap - ex ASX100
Reversion 1M Mean Large Cap - ASX100
Reversion
MACD
MACD
-4% -2% 0% 2% 4% 6% 8% 10% 12%
-1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00
Information Ratio
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
Technical indicators mean reverting signals, and perform significantly better in the large cap universe. While the
perform much more traditional momentum signals like 12M momentum and earnings revisions favour the
stronger in the large smaller cap universe.
cap markets
3
R. De Souza, “Macquarie Alpha Model – The alpha model goes under the knife”
30 April 2009 8
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 Fig 13 Q1-Q3 L/S after cost information ratio in ASX
in ASX 100 universe 100 universe with 15bp and 30bp transaction costs
Earnings Yield Earnings Yield
12M Momentum 12M Momentum
Macquarie Alpha Model Macquarie Alpha Model

3M Earnings 3M Earnings
Revisions Revisions
Price relative to Price relative to
52 Week High 52 Week High
Trading Intensity Trading Intensity
1M Mean 1M Mean
Reversion After Cost IR (30bps)
Reversion
MACD MACD After Cost IR (15bps)
On Balance On Balance
Before Cost IR
Volume Volume
Stochastic Stochastic
Oscillator Oscillator
Money Flow Index Money Flow Index
Force Index Force Index

Relative Strength Relative Strength
Indicator Indicator
Commodity Commodity
Channel Index Channel Index
Williams %R Williams %R
Bollinger Band Bollinger Band
0% 500% 1000% 1500% 2000% 2500% 3000% 3500% 4000% -1.00 -0.50 0.00 0.50 1.00 1.50
Average Information Ratio
Average Annual 2 way Turnover
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
Technical indicators
Figure 13. The high information ratio (even after costs) for a number of the other technical
have very high
factors (Williams %R, Bollinger Bands, Commodity Channel, RSI and Force Index) highlights
turnovers which can
the potential for technical indicators to add value in the investment process.
impact the Alpha
available from the Of the technical factors Williams %R stands out overall:
signals
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.
30 April 2009 9
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

Trend Following Mean Reverting
Distribution Index
Relative Volatility
Relative Strength
12M Momentum
Trading Intensity
Moving Average
Maquarie Alpha
6M Momentum
3M Momentum
Price relative to
Price relative to
Accumulation /
Bollinger Band
Channel Index
52 Week High
Earnings Yield
52 Week Low
3M Earnings
Money Flow
Williams %R
Commodity
Force Index
True Range
On Balance
Movement
Directional
Flow Index
Stochastic
Reversion
1M Mean
Oscillator
Revisions
Indicator
Average
Average
Volume
Chaikin
Money
Model
MACD
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
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%
52 Week Low
Price relative to
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%
52 Week High
Maquarie Alpha
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%
Model
3M Earnings
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%
Revisions
Trend Following
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 /
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%
Distribution Index
Chaikin
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%
Money Flow
Average
-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%
Directional
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%
Mean Reverting
Average
-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%
True Range
Commodity
-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%
Channel Index
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
-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%
Flow Index
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
-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%
Volume
Relative Strength
-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%
Indicator
Stochastic
-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%
Oscillator
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
-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%
Reversion
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 10
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
Technical indicators
trading intensity, accumulation / distribution index or price relative to 52 week highs).
are lowly correlated
to the standard A nice result is that the key predictive signals are not highly correlated to the standard quant
quant factors. 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
Technical indicators compared this to a new model which also includes the Williams %R factor.
in a multi factor
Figure 15 plots the two strategies. We tested this in the large cap universe of the MSCI
model struggle to
Australia, where the technical indicators would add the most value and be the least impacted
add value due to
by turnover constraints. Due to the smaller restricted universe we used a Q1-Q3 Long Short
higher turnover
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
600
Standard Model
160%
Standard Model With Williams
500
Standard Model (After Costs)
140%
Standard Model With Williams (After Costs)
120%
400
100%
300
80%
200
60%
100 40%
20%
0
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
0%
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.
30 April 2009 11
Overlay on a multifactor model

Next, we look at a different way to combine the Williams %R into a multifactor model. Given
Technical indicators the high turnover of the signal we look to test the Williams %R as a ‘knockout filter’ for a
as a knock out signal model. We use the same four factor model of equally weighted value, revisions, quality and
add value without price momentum and again tested in the MSCI Australia universe. This time we divide the
significantly standard model and the Williams %R technical indicator into three equal weighted groupings.
increasing turnover. 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

Williams %R
Over Over
sold Middle bought Unconditional
High 7.7 3.8 -2.4 3.3
Standard
Model
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
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%.
30 April 2009 12
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
450
Standard Model (After Costs) 160%
400 Standard Model With Williams (After Costs)
140%
350 Standard Model With Williams Overlay (30bps Tcost)
120%
300
100%
250
200
80%
150 60%
100 40%
50 20%
0 0%
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 Standard Model With Standard Model With

Williams Williams Overlay
Source: FactSet, Macquarie Research, April 2009 Source: FactSet, Macquarie Research, April 2009
Fig 20 Annualised returns of model portfolios

50%
40%
30%
Annual Return
20%
10%
0%
-10%
Standard Model
-20% Standard Model With Williams
Standard Model With Williams Overlay
-30%
Source: FactSet, Macquarie Research, April 2009
30 April 2009 13
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 14
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) Æ 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)) Æ 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 Æ 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 15
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 16
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 17
Technical factor performance

Williams %R - Time Series Normalised
Information Coefficient 40%
Last 5 Last 3
All History Last 1 Year 30%
Years Years
20%
Average IC 5.6% 5.4% 4.8% 5.9%
Std Deviation IC 9.6% 7.9% 8.7% 12.3% 10%
Success Rate 73.7% 73.3% 63.9% 50.0% 0%
Avg IC / Std Dev IC 0.59 0.68 0.55 0.48 -10%
t-stat 7.68 5.29 3.29 1.65 -20%
-30%
-40%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
Monthly IC 12M Average IC
IC Horizon
Average IC
8.0%
1 month 5.6% 6.0%
3 month 1.4% 4.0%
6 month -0.3%
2.0%
12 month -1.8%
0.0%
24 month -1.5%
-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 24
IC Decay Accumulated IC
Long Short Q1-Q5
Last 5 Last 3
All History Last Year 10
Years Years
Excess Market Return
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% 1
0
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
Top Quintile Bottom Quintile
Long Short Q1-Q5 Rolling Excess Returns 50%

Long Short Returns
Last 5 Last 3 40%

All History Last 1 Year
Years Years 30%
Q1 - Q5 Return (p.a.) 13.1% 15.1% 16.4% 32.1%
20%
Q1 - Q5 Volatility (p.a.) 10.2% 10.1% 12.2% 19.6%
Q1 - Q5 Information Ratio 1.28 1.50 1.34 1.64 10%
Q1 - Q5 Turnover (p.a. 1 way) 1913% 1934% 1922% 1934%
0%
-10%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
LS Return 12M Rolling LS returns
Long Short Q1-Q5 Turnover 1-way

2500%
All History Last 5 Years Last 3 Years Last 1 Year 2000%

Turnover
1500%
Average Annual L/S Turnover 1913% 1934% 1922% 1934%
1000%
500%
0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Turnover 12M Rolling Turnover
Pure Factor Return

Last 5 Last 3 12%
Years Years
10%
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% 8%
Factor Return
Pure Factor Sharpe 1.81 1.78 1.81 1.86 6%

4%
2%
0%
-2%
-4%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Factor Return Annualised Factor Return
Source: IBES, MSCI, and Macquarie Research, April 2009
30 April 2009 18
Williams %R - Time Series Normalised - continued

Fractiles
Annual Excess Informatio
Volatility 100
Return Return n Ratio
Q1 14.3% 3.8% 17.0% 0.22
Cumulative Return
Q2 12.3% 1.7% 16.1% 0.11
Q3 12.2% 1.7% 16.2% 0.11 10
Q4 10.2% -0.3% 15.0% -0.02
Q5 4.7% -5.8% 15.8% -0.36
Market 10.5%
1
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
Q1 Q2 Q3 Q4 Q5 Market
Fractile Cross sectional Hit Rate 60.0%

Last 5 Last 3 58.0%
% Cross sectional hit rate

Years Years 56.0%
Q1 53.4% 52.5% 52.7% 52.0% 54.0%
Q2 50.2% 51.3% 51.0% 51.6% 52.0%
Q3 50.6% 51.5% 50.9% 53.1% 50.0%
Q4 48.6% 48.5% 48.8% 50.5% 48.0%
Q5 45.9% 45.7% 46.0% 42.2% 46.0%
44.0%
42.0%
40.0%
Q1 Q2 Q3 Q4 Q5
All History Last 5 Years Last 3 Years Last 1 Year
Size
8.0%
Last 5 Last 3 Latest No. of
Average IC All History Last 1 Year 7.0%

Years Years Stocks
6.0%
Large 6.15% 5.68% 6.40% 7.35% 94
Small 5.13% 5.30% 4.07% 5.37% 191 5.0%
All 5.63% 5.42% 4.76% 5.87% 285 4.0%
3.0%
Last 5 Last 3 2.0%
Std Dev IC All History Last 1 Year
Years Years
1.0%
Small 13.06% 12.85% 12.88% 14.90%
0.0%
Large 12.09% 8.99% 9.09% 12.91%
Large Small All
All 9.61% 7.93% 8.68% 12.31% All History Last 5 Years Last 3 Years Last 1 Year
Sector
Latest
Last 5 Last 3 18%

All History Number of 16%
Years Years 14%
Stocks
12%
Energy 3.06% 4.64% 8.92% 32 10%
Materials 5.97% 1.36% -0.58% 57 8%
Industrials 5.93% 8.30% 7.85% 45 6%
Consumer Discretionary 5.44% 5.77% 7.43% 31 4%
2%
Consumer Staples 7.91% 3.27% 4.13% 16
0%
Health Care 4.32% 8.94% 8.85% 10 -2%
Financials 8.39% 6.21% 6.82% 73
Energy
Discretionary
Materials
Industrials
Financials
Utilities
Health Care
Consumer
Information
Telecomm.
Services
Consumer
Staples
Information Tech 7.49% 15.93% 15.69% 3

Tech
Telecomm. Services 13.53% 9.83% 2.50% 3

Utilities 9.99% 11.56% 14.37% 15
All History Last 5 Years Last 3 Years
Seasonality 12%
Information
IC Std Dev Samples

Coefficient 10%
January 10.9% 9.1% 15
February 1.2% 11.2% 15 8%
March 1.9% 9.2% 15
6%
April 7.4% 5.4% 14
May 6.9% 11.5% 14 4%
June 8.8% 9.3% 14
July 9.8% 11.1% 14 2%
August 2.6% 11.7% 14
September 3.7% 8.7% 14 0%
October 7.7% 7.8% 14
y
y
y
ay
ril
st
ch
ne
r
r
r
be
be
be
be
l
ar
ar
Ju
gu
Ap
M
ar
Ju
November 2.9% 7.0% 14

nu
ru
em
o
m
em
Au
M
ct
b
te
Ja
O
Fe
ov
ec
p
December 3.9% 8.3% 14

Se
Factor Coverage 400 100%

Last 5 Last 3 90%
% of Market Covered
All History Last 1 Year 350

Number of Stocks
Years Years 300

80%
Average number of stocks 228 291 322 319 70%
250 60%
Coverage of available stocks (%) 99% 99% 99% 100% 200 50%
150 40%
30%
100
20%
50 10%
0 0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Stocks with Factor % Coverage of Universe

% Coverage by Market Cap
30 April 2009 19
Force Index - Time Series Normalised

Last 5 Last 3
Years Years
20%
Average IC 4.4% 4.6% 3.7% 3.5%
Std Deviation IC 8.7% 7.3% 7.2% 9.1% 10%
Success Rate 69.0% 73.3% 69.4% 58.3% 0%
Avg IC / Std Dev IC 0.51 0.64 0.52 0.38 -10%
t-stat 6.74 4.93 3.11 1.32 -20%
-30%
-40%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
IC Horizon
Average IC
6.0%
1 month 4.4%
4.0%
3 month 2.2%
6 month 1.0% 2.0%
12 month -0.1% 0.0%
24 month -0.3% -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 24
Long Short Q1-Q5
Last 5 Last 3
Years Years

Q1 - Q5 Return (p.a.) # 9.4% 12.0% 11.4% 17.7%
Q1 - Q5 Volatility (p.a.) # 8.8% 7.1% 8.0% 11.7% 1
0
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

30%
Long Short Returns
Last 5 Last 3 25%

Years Years
Q1 - Q5 Return (p.a.) 9.4% 12.0% 11.4% 17.7% 15%
10%
Q1 - Q5 Volatility (p.a.) 8.8% 7.1% 8.0% 11.7% 5%
Q1 - Q5 Turnover (p.a. 1 way) 1855% 1899% 1894% 1858% -5%
-10%
-15%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

2500%

Turnover
1500%
1000%
500%
0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Pure Factor Return

Last 5 Last 3 12%
Years Years
10%
Factor Return
Pure Factor Sharpe 1.67 1.86 1.83 1.97 6%

4%
2%
0%
-2%
-4%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
30 April 2009 20
Force Index - Time Series Normalised - continued

Fractiles
Volatility 100
Q1 16.3% 5.8% 15.7% 0.37
Cumulative Return
Q2 12.2% 1.6% 16.9% 0.10
Q3 9.2% -1.3% 15.3% -0.09 10
Q4 7.5% -3.0% 16.6% -0.18
Q5 8.1% -2.4% 15.5% -0.16
Market 10.5%
1
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

Last 5 Last 3 58.0%

Years Years 56.0%
Q1 53.6% 52.8% 52.7% 54.1% 54.0%
Q2 50.0% 51.0% 50.8% 48.0% 52.0%
Q3 49.1% 48.9% 48.5% 50.4% 50.0%
Q4 47.9% 48.1% 48.2% 46.5% 48.0%
Q5 48.0% 48.6% 49.3% 50.6% 46.0%
44.0%
42.0%
40.0%
Q1 Q2 Q3 Q4 Q5
Size
6.0%

Years Years Stocks
Large 4.03% 2.86% 2.04% 0.93% 94 4.0%
Small 4.60% 5.23% 4.22% 4.60% 188
All 4.45% 4.62% 3.71% 3.46% 282 3.0%
2.0%
Last 5 Last 3
Years Years 1.0%
Small 12.16% 12.38% 13.76% 14.48%
0.0%
Large 10.73% 7.89% 6.71% 8.74%
Large Small All
Sector
Latest
Last 5 Last 3 25%

All History Number of
Years Years 20%
Stocks
Energy 3.40% 3.89% 4.41% 32 15%
Materials 5.25% 2.47% 0.50% 56 10%
Industrials 6.35% 3.63% 2.40% 45
5%
Consumer Discretionary 4.78% 5.84% 7.12% 31
Consumer Staples 5.36% -0.23% -1.89% 16 0%
Health Care 8.36% 10.66% 8.97% 10 -5%
Financials 6.01% 5.38% 5.65% 71
Energy
Discretionary
Materials
Industrials
Financials
Health Care
Utilities
Consumer
Information
Telecomm.
Services
Consumer
Staples

Tech

Utilities 1.00% 11.79% 13.60% 15
Seasonality 9%
Information
IC Std Dev Samples 8%

Coefficient
7%
January 0.9% 6.2% 15
February 0.7% 9.7% 15 6%
March 3.5% 7.4% 15 5%
April 4.3% 10.3% 14 4%
May 8.1% 8.4% 14 3%
June 6.9% 9.8% 14
2%
July 8.4% 9.5% 14
1%
August 5.6% 7.1% 14
September 1.0% 8.4% 14 0%
October 8.4% 9.7% 14
y
y
y
ay
ril
st
ch
ne
r
r
r
be
be
be
be
l
ar
ar
Ju
gu
Ap
M
ar
Ju
November 2.7% 8.1% 14

nu
ru
em
o
m
em
Au
M
ct
b
te
Ja
O
Fe
ov
ec
p
December 3.4% 5.6% 14

Se

Last 5 Last 3 90%
% of Market Covered

Number of Stocks
Years Years 300

80%
250 60%
150 40%
30%
100
20%
50 10%
0 0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D

30 April 2009 21
Bollinger Band - Time Series Normalised

Last 5 Last 3
Years Years
20%
Average IC 4.4% 4.6% 4.2% 4.6%
Std Deviation IC 10.1% 9.0% 9.7% 12.7% 10%
Success Rate 67.3% 71.7% 69.4% 58.3% 0%
Avg IC / Std Dev IC 0.44 0.51 0.43 0.36 -10%
t-stat 5.77 3.99 2.59 1.26 -20%
-30%
-40%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
IC Horizon
Average IC
6.0%
1 month 4.4%
4.0%
3 month 0.6%
6 month -0.5% 2.0%
12 month -2.3% 0.0%
24 month -1.8% -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 24
Long Short Q1-Q5
Last 5 Last 3
Years Years

Q1 - Q5 Return (p.a.) # 7.7% 8.6% 9.3% 18.3%
Q1 - Q5 Volatility (p.a.) # 10.3% 10.5% 12.5% 20.3% 1
0
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

30%
Long Short Returns
Last 5 Last 3 25%

Years Years 20%
Q1 - Q5 Return (p.a.) 7.7% 8.6% 9.3% 18.3% 15%
Q1 - Q5 Volatility (p.a.) 10.3% 10.5% 12.5% 20.3% 10%
Q1 - Q5 Turnover (p.a. 1 way) 1939% 1955% 1969% 1977% 0%
-5%
-10%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

2500%

Turnover
1500%
1000%
500%
0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Pure Factor Return

Last 5 Last 3 10%
Years Years
8%
Factor Return
Pure Factor Sharpe 1.21 1.15 1.35 1.32

4%
2%
0%
-2%
-4%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
30 April 2009 22
Bollinger Band - Time Series Normalised - continued

Fractiles
Volatility 100
Q1 13.2% 2.7% 16.7% 0.16
Cumulative Return
Q2 12.1% 1.5% 16.0% 0.09
Q3 11.1% 0.5% 16.3% 0.03 10
Q4 9.5% -1.1% 16.0% -0.07
Q5 7.0% -3.5% 14.9% -0.24
Market 10.6%
1
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

Last 5 Last 3 58.0%

Years Years 56.0%
Q1 53.3% 52.1% 51.4% 50.9% 54.0%
Q2 50.0% 51.6% 51.6% 51.8% 52.0%
Q3 49.5% 51.0% 51.6% 52.3% 50.0%
Q4 49.7% 48.3% 49.4% 50.7% 48.0%
Q5 46.5% 46.5% 45.7% 43.8% 46.0%
44.0%
42.0%
40.0%
Q1 Q2 Q3 Q4 Q5
Size
12.0%

Years Years Stocks
Large 5.81% 6.35% 6.96% 10.10% 94 8.0%
Small 3.40% 3.65% 2.97% 2.74% 191
All 4.44% 4.65% 4.17% 4.64% 285 6.0%
4.0%
Last 5 Last 3
Years Years 2.0%
Small 13.49% 14.36% 14.73% 15.24%
0.0%
Large 11.72% 9.70% 9.41% 12.89%
Large Small All
Sector
Latest
Last 5 Last 3 20%

Years Years 16%
Stocks
14%
Energy 2.89% 3.50% 7.10% 32 12%
Materials 5.07% 2.36% 2.23% 57 10%
Industrials 5.45% 6.53% 6.40% 45 8%
4%
2%
Health Care 2.49% 10.04% 9.36% 10 0%
Financials 8.13% 5.79% 6.63% 73
Energy
Discretionary
Materials
Industrials
Financials
Health Care
Utilities
Consumer
Information
Telecomm.
Services
Consumer
Staples

Tech

Utilities 5.55% 11.60% 13.20% 15
Seasonality 10%
Information
IC Std Dev Samples

Coefficient 8%
January 9.4% 11.9% 15
February -0.2% 12.1% 15 6%
March 1.6% 9.3% 15
4%
April 5.4% 6.9% 14
May 6.4% 10.9% 14 2%
June 7.1% 11.2% 14
July 9.0% 11.3% 14 0%
August 1.6% 10.2% 14
September 1.4% 8.7% 14 -2%
October 7.7% 8.3% 14
y
y
y
ay
ril
st
ch
ne
r
r
r
be
be
be
be
l
ar
ar
Ju
gu
Ap
M
ar
Ju
nu
ru
November 1.8% 6.7% 14

em
o
m
em
Au
M
ct
b
te
Ja
O
Fe
ov
ec
p
December 2.2% 8.5% 14

Se

Last 5 Last 3 90%
% of Market Covered

Number of Stocks
Years Years 300

80%
250 60%
150 40%
30%
100
20%
50 10%
0 0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D

30 April 2009 23
Commodity Channel Index - Time Series Normalised

Last 5 Last 3
Years Years
20%
Average IC 4.1% 4.0% 3.7% 3.9%
Std Deviation IC 9.1% 8.0% 8.5% 12.6% 10%
Success Rate 65.5% 66.7% 63.9% 50.0% 0%
Avg IC / Std Dev IC 0.45 0.50 0.44 0.31 -10%
t-stat 5.84 3.84 2.64 1.08 -20%
-30%
-40%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
IC Horizon
Average IC
5.0%
1 month 4.1% 4.0%
3 month 0.8% 3.0%
6 month -0.2% 2.0%
1.0%
12 month -1.5% 0.0%
24 month -1.5% -1.0%
-2.0%
-3.0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Long Short Q1-Q5
Last 5 Last 3
Years Years

Q1 - Q5 Return (p.a.) # 8.0% 9.1% 11.0% 24.7%
Q1 - Q5 Volatility (p.a.) # 9.1% 9.3% 11.3% 18.3% 1
0
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

25%
Long Short Returns
Last 5 Last 3 20%

Years Years 15%
Q1 - Q5 Return (p.a.) 7.9% 9.1% 11.0% 24.7% 10%
Q1 - Q5 Volatility (p.a.) 9.1% 9.3% 11.3% 18.3% 5%
-10%
-15%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

2500%

Turnover
1500%
1000%
500%
0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Pure Factor Return

Last 5 Last 3 10%
Years Years
8%
Factor Return

4%
2%
0%
-2%
-4%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
30 April 2009 24
Commodity Channel Index - Time Series Normalised - continued

Fractiles
Volatility 100
Q1 14.1% 3.6% 16.2% 0.22
Cumulative Return
Q2 11.0% 0.5% 16.5% 0.03
Q3 11.3% 0.8% 15.8% 0.05 10
Q4 8.8% -1.7% 16.0% -0.11
Q5 7.5% -3.0% 15.1% -0.20
Market 10.5%
1
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

Last 5 Last 3 58.0%

Years Years 56.0%
Q1 53.0% 52.2% 53.0% 54.2% 54.0%
Q2 50.4% 52.3% 51.5% 50.7% 52.0%
Q3 49.1% 49.7% 50.0% 50.9% 50.0%
Q4 49.2% 48.9% 48.6% 48.7% 48.0%
Q5 47.2% 46.6% 46.6% 45.2% 46.0%
44.0%
42.0%
40.0%
Q1 Q2 Q3 Q4 Q5
Size
7.0%

Years Years Stocks
Large 5.20% 5.03% 6.29% 6.59% 93 5.0%
Small 3.18% 3.66% 2.90% 3.32% 191 4.0%
All 4.06% 3.97% 3.75% 3.93% 284
3.0%
Last 5 Last 3 2.0%
Years Years 1.0%
Small 12.59% 13.18% 12.63% 13.04%
0.0%
Large 11.03% 8.75% 9.04% 13.93%
Large Small All
Sector
Latest
Last 5 Last 3 25%

Years Years
Stocks
15%
Energy 1.59% 2.29% 6.98% 32
10%
Materials 5.01% 0.90% -1.85% 56
Industrials 4.65% 9.19% 8.63% 45 5%
Consumer Staples 6.04% 1.43% 4.68% 16 -5%
Health Care 3.83% 11.72% 13.65% 10 -10%
Financials 5.97% 4.11% 4.97% 73
Energy
Discretionary
Materials
Industrials
Financials
Health Care
Utilities
Consumer
Information
Telecomm.
Services
Consumer
Staples

Tech
Telecomm. Services 2.23% -7.17% -3.61% 3

Utilities 3.85% 10.75% 12.69% 15
Seasonality 12%
Information

Coefficient
January 11.2% 8.2% 15 8%
February -0.6% 8.6% 15
March -0.8% 8.9% 15 6%
April 4.6% 5.6% 14 4%
May 7.2% 11.5% 14
June 6.7% 7.7% 14 2%
July 4.5% 10.0% 14 0%
August 2.1% 11.3% 14
September 2.3% 9.6% 14 -2%
October 6.5% 7.5% 14
y
y
y
ay
ril
st
ch
ne
r
r
r
be
be
be
be
l
ar
ar
Ju
gu
Ap
M
ar
Ju
nu
ru
November 2.9% 7.2% 14

em
o
m
em
Au
M
ct
b
te
Ja
O
Fe
ov
ec
p
December 2.4% 7.2% 14

Se

Last 5 Last 3 90%
% of Market Covered

Number of Stocks
Years Years 300

80%
250 60%
150 40%
30%
100
20%
50 10%
0 0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D

30 April 2009 25
Relative Strength Indicator - 14Days - Time Series Normalised

Last 5 Last 3
Years Years
20%
Average IC 3.6% 4.2% 3.2% 3.6%
Std Deviation IC 10.4% 9.2% 9.5% 12.3% 10%
Success Rate 63.7% 70.0% 66.7% 66.7% 0%
Avg IC / Std Dev IC 0.35 0.45 0.34 0.29 -10%
t-stat 4.59 3.52 2.03 1.01 -20%
-30%
-40%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D
IC Horizon
Average IC
4.0%
1 month 3.6%
3 month 0.7% 2.0%
6 month -0.3%
0.0%
12 month -1.9%
24 month -1.8% -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 24
Long Short Q1-Q5
Last 5 Last 3
Years Years

Q1 - Q5 Return (p.a.) # 5.9% 9.7% 7.9% 12.3%
Q1 - Q5 Volatility (p.a.) # 10.2% 10.9% 13.1% 20.1% 1
0
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

15%
Long Short Returns
Last 5 Last 3
Years Years
5%
Q1 - Q5 Return (p.a.) 5.9% 9.7% 7.9% 12.3%
0%
Q1 - Q5 Volatility (p.a.) 10.2% 10.9% 13.1% 20.1%
-5%
Q1 - Q5 Information Ratio 0.58 0.88 0.60 0.61
-15%
-20%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

2500%

Turnover
1500%
1000%
500%
0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
Pure Factor Return

Last 5 Last 3 8%
Years Years
Average pure factor return (p.a.) 2.84% 3.18% 3.68% 5.38% 6%
Pure Factor Volatility 2.74% 3.00% 3.63% 5.29%
Factor Return
4%
2%
0%
-2%
-4%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
30 April 2009 26
Relative Strength Indicator - 14Days - Time Series Normalised - continued

Fractiles
Volatility 10
Q1 11.7% 1.1% 16.9% 0.07
Cumulative Return
Q2 11.4% 0.8% 16.1% 0.05
Q3 12.2% 1.6% 16.1% 0.10
Q4 10.3% -0.2% 16.2% -0.01
Q5 7.1% -3.5% 14.7% -0.24 1
Market 10.6%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D
D

Last 5 Last 3 58.0%

Years Years 56.0%
Q1 52.4% 51.7% 51.5% 51.9% 54.0%
Q2 50.4% 51.9% 50.9% 52.0% 52.0%
Q3 49.8% 50.4% 50.8% 49.2% 50.0%
Q4 49.1% 49.5% 50.0% 49.6% 48.0%
Q5 47.3% 46.1% 46.4% 46.8% 46.0%
44.0%
42.0%
40.0%
Q1 Q2 Q3 Q4 Q5
Size
9.0%
Last 5 Last 3 Latest No. of 8.0%
Average IC All History Last 1 Year

Years Years Stocks 7.0%
Large 5.34% 6.28% 6.22% 8.34% 94 6.0%
Small 2.52% 3.07% 1.78% 1.84% 190
5.0%
All 3.63% 4.18% 3.22% 3.58% 284
4.0%
3.0%
Last 5 Last 3
Std Dev IC All History Last 1 Year 2.0%
Years Years
1.0%
Small 13.88% 15.00% 16.53% 19.91%
0.0%
Large 11.56% 10.12% 9.09% 11.97%
Large Small All
Sector
Latest
Last 5 Last 3 18%

Years Years 14%
Stocks
12%
Energy -0.61% 1.73% 5.09% 32 10%
Materials 3.21% 0.88% 0.38% 56 8%
Industrials 4.85% 6.06% 4.87% 45 6%
2%
0%
Health Care 1.24% 9.52% 7.96% 10 -2%
Financials 7.10% 4.94% 6.29% 73
Energy
Discretionary
Health Care
Materials
Industrials
Financials
Utilities
Consumer
Information
Telecomm.
Services
Consumer
Staples

Tech

Utilities 4.51% 8.01% 7.41% 15
Seasonality 10%
Information

Coefficient 8%
January 7.5% 11.5% 15 7%
February 0.6% 11.3% 15 6%
March 3.3% 9.6% 15
5%
April 4.1% 8.9% 14
4%
May 3.2% 10.6% 14
3%
June 6.0% 12.2% 14
July 9.1% 11.2% 14 2%
August 0.7% 8.4% 14 1%
September 2.4% 9.4% 14 0%
October 4.4% 10.5% 14
ry
ay
ly
r il
st
ch
er
r
be
be
be
r
Ju
gu
Ap
a
ua
ob
M
ar
Ju
nu
November 0.2% 9.3% 14

m
em
em
Au
br
ct
te
Ja
O
Fe
ov
ec
p
December 1.9% 10.4% 14

Se

Last 5 Last 3 90%
% of Market Covered

Number of Stocks
Years Years 300

80%
250 60%
150 40%
30%
100
20%
50 10%
0 0%
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
ec
D

30 April 2009 27
Important disclosures:
Recommendation definitions Volatility index definition* Financial definitions
Macquarie - Australia/New Zealand This is calculated from the volatility of historical All "Adjusted" data items have had the following
Outperform – return >5% in excess of benchmark return price movements. adjustments made:
Neutral – return within 5% of benchmark return Added back: goodwill amortisation, provision for
Underperform – return >5% below benchmark return Very high–highest risk – Stock should be catastrophe reserves, IFRS derivatives & hedging,
Macquarie – Asia/Europe expected to move up or down 60–100% in a year – IFRS impairments & IFRS interest expense
Outperform – expected return >+10% investors should be aware this stock is highly Excluded: non recurring items, asset revals, property
Neutral – expected return from -10% to +10% speculative. revals, appraisal value uplift, preference dividends &
Underperform – expected return <-10% minority interests
High – stock should be expected to move up or
Macquarie First South - South Africa down at least 40–60% in a year – investors should EPS = adjusted net profit / efpowa*
Outperform – expected return >+10% be aware this stock could be speculative. ROA = adjusted ebit / average total assets
Neutral – expected return from -10% to +10% ROA Banks/Insurance = adjusted net profit /average
Underperform – expected return <-10% Medium – stock should be expected to move up or total assets
Macquarie - Canada down at least 30–40% in a year. ROE = adjusted net profit / average shareholders funds
Outperform – return >5% in excess of benchmark return Gross cashflow = adjusted net profit + depreciation
Neutral – return within 5% of benchmark return Low–medium – stock should be expected to move *equivalent fully paid ordinary weighted average
Underperform – return >5% below benchmark return up or down at least 25–30% in a year. number of shares
Macquarie - USA
Low – stock should be expected to move up or All Reported numbers for Australian/NZ listed stocks
Outperform (Buy) – return >5% in excess of benchmark
down at least 15–25% in a year. are modelled under IFRS (International Financial
return (Russell 3000)
* Applicable to Australian/NZ/Canada stocks only Reporting Standards).
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
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 Bangkok Calgary Hong Kong Jakarta Johannesburg Kuala Lumpur
Tel: (649) 377 6433 Tel: (662) 694 7999 Tel: (1 403) 218 6650 Tel: (852) 2823 3588 Tel: (62 21) 515 1818 Tel: (2711) 583 2000 Tel: (60 3) 2059 8833
London Manila Melbourne Montreal Mumbai Perth Seoul
Tel: (44 20) 3037 4400 Tel: (63 2) 857 0888 Tel: (613) 9635 8139 Tel: (1 514) 925 2850 Tel: (91 22) 6653 3000 Tel: (618) 9224 0888 Tel: (82 2) 3705 8500
Shanghai Singapore Sydney Taipei Tokyo Toronto New York
Tel: (86 21) 6841 3355 Tel: (65) 6231 1111 Tel: (612) 8232 9555 Tel: (886 2) 2734 7500 Tel: (81 3) 3512 7900 Tel: (1 416) 848 3500 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 28
Notes
30 April 2009 29
Research
Heads of Equity Research Industrials Emerging Leaders
David Rickards (Global Co – Head) (852) 2823 3538 Transportation /Commercial Services Adam Simpson (612) 8232 4439
John O’Connell (Global Co – Head) (612) 8232 7544 Warren Doak (New Zealand) (649) 363 1416 Benn Skender (612) 8232 6846
Ian Myles (612) 8232 4157 Andrew Wackett (618) 9224 0867
Consumer Staples
Russell Shaw (612) 8232 7124 Quantitative
Food & Beverages Transportation - Infrastructure
Greg Dring (612) 8232 3104 Martin Emery (Hong Kong) (852) 2823 3582
Warren Doak (New Zealand) (649) 363 1416
Scott Hamilton (612) 8232 3544
Consumer Discretionary Ian Myles (612) 8232 4157
George Platt (612) 8232 6539
Media and Tourism & Leisure Materials Data Services (Australia & New Zealand)
Alex Pollak (612) 8232 3172 Chemicals/Containers, Packaging/Paper & Sheridan Duffy (612) 8232 9786
Retailing Forest Products, Construction Materials
Warren Doak (New Zealand) (649) 363 1416 Stephen Hudson (New Zealand) (649) 363 1414 Economics and Strategy
Greg Dring (612) 8232 3104 John Purtell (612) 8232 8633 Tanya Branwhite (Strategy) (612) 8232 7628
Energy Global Metals & Mining Richard Gibbs (Head of Economics) (612) 8232 3935
Jim Copland (612) 8232 0397 Neale Goldston-Morris (Strategy) (612) 8232 7562
Adrian Wood (612) 8232 8531 Brian Redican (Aus Economics) (612) 8232 7016
Len Eldridge (618) 9224 0838
Financials Brendan Harris (612) 8232 3575 Mark Tierney (Int’l Economics) (612) 8232 3121
Sophie Spartalis (612) 8232 5159
Banks
Ben Zucker (612) 8232 6089 Real Estate Find our research at
Tom Quarmby (612) 8232 8668 Macquarie: www.macquarie.com.au/research
Property Trusts & Developers Thomson: www.thomson.com/financial
Diversified Financials Callum Bramah (612) 8232 7647 Reuters: www.knowledge.reuters.com
Deana Mitchell (612) 8232 4576 Paul Checchin (612) 8232 4197 Bloomberg: MAC GO
Insurance Factset: http://www.factset.com/home.aspx
Tony Jackson (612) 8232 4442
Telecommunications
Contact Gareth Warfield for access (612) 8232 3207
Deana Mitchell (612) 8232 4576 Andrew Levy (612) 8232 5165
Healthcare Utilities See and hear our analysts at

www.macquarie.com.au/macquariedigital
Steve Wheen (612) 8232 4130 Stephen Hudson (New Zealand) (649) 363 1414
Gavin Maher (612) 8232 4151
Industrials Toll free from overseas
Commodities & Precious Metals Canada 1800 989 8159
Capital Goods
Warren Doak (New Zealand) (649) 363 1416 Jim Lennon (London) (44 20) 3037 4271 Hong Kong 800 96 2049
Greg Dring (612) 8232 3104 Adam Rowley (London) (44 20) 3037 4272 Japan 0053 161 6437
John Purtell (612) 8232 8633 New York 1888 622 7862
Singapore 800 616 1037
Email addresses
FirstName.Surname@macquarie.com
eg. David.Rickards@macquarie.com
Sales
Equities Specialist Sales Treasury & Commodities
Martin Dacron (Australia) (612) 8232 7421 Matthew Clegg (Index) (612) 8232 5653 Gavin Bradley (Metals & Mining) (612) 8232 4248
Scott Dolling (Asia) (852) 2823 3705 Andrew Mouat (Property Trusts) (612) 8232 3151 Matthew Forgham
Rob Fabbro (Continental Europe) (44 20) 7065 2031 George Platt (Quantitative) (612) 8232 6539 (Metals & Mining) (44 20) 3037 4615
Charles Nelson (UK) (44 20) 7065 2032 Phil Zammit (Emerging Leaders) (612) 8232 3122 Emma Winspear (Futures) (613) 9635 8275
Duane O’Donnell (Melbourne) (613) 9635 9183 James Mactier (Metals & Mining) (618) 9224 0612
Alternative Strategies
Dave Roberton (New Zealand) (649) 363 1498 Ian Miller (Futures) (612) 8232 3555
Anthony Panaretto (Sales) (612) 8232 4500 Will Richardson (Foreign Exch) (612) 8232 4777
Luke Sullivan (New York) (1 212) 231 2507
Shannon Donohoe
Stevan Vrcelj (Head of Global Sales) (612) 8232 5999 (Stock borrow & loan) (612) 8232 6997 Syndication
Mark Donnelly (Equity finance) (612) 8232 7664 Peter Curry (612) 8232 4039
Cameron Duncan (Converts) (612) 8232 7405 Paul Staines (612) 8232 7781
Anthony Hourigan (Derivatives) (612) 8232 9884
April 09

Macquarie Technicals

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Macquarie Technicals

Uploaded by

Copyright:

Available Formats

AUSTRALIA Quantitative analysis

30 April 2009 Let’s get technical

Volatile markets present opportunity to use technicals

Williams %R the most powerful signal

Size matters – technicals work best in large cap stocks

Let’s get technical

Williams %R – the standout technical indicator

Other key technical indicators that also performed well were:

Adding value to a quant process

Market instability hurts fundamental factors

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

Turning our attention to technicals

Using technical indicators as Quant factors

Fig 5 Technical indicators tested

 has broken out from a trading range

 changes its price trends

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

Performance of technical indicators in

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

40% Information 50% Long Short

Monthly IC 12M Average IC LS Return 12M Rolling LS returns

Last 5 Last 3 Last 5 Last 3

12M Momentum 12M Momentum

Macquarie Alpha Model Macquarie Alpha Model

Force Index Force Index

Bollinger Band Bollinger Band

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

 performs consistently through time and

Combining with other factors and use in a

Correlations with other factors

 Earnings revisions and

Fig 14 IC Correlation matrix for quant factors and technical indicators

Source:Macquarie Research, April 2009

Combining in a multifactor model

Fig 15 Standard and enhanced quant model

Overlay on a multifactor model

Fig 17 Annualised active returns of filtered portfolios

Middle 5.8 1.5 -4.0 0.8

Source: FactSet, Macquarie Research, April 2009

Further breaking down the above diagram (by corners):

Standard Model Standard Model With Standard Model With

Fig 20 Annualised returns of model portfolios

Source: FactSet, Macquarie Research, April 2009

Appendices – factor definitions

MACD - Moving Average Convergence/Divergence

The %D is a simple moving average of the %K using an n day window

Fig 21 Technical definitions (continued)

Typical Price = average of high, low and close price

Money Flow Index

Typical Price = (High Price + Low Price + Close Price)/3

The Money Flow Index = 100 – (100/(1+Money Ratio))

Chaikin Money Flow

Force = (Price Today – Price Yesterday) * Volume

Fig 21 Technical definitions (continued)

AD = (((Close – Low) – (High – Close)) / (High – Low)) * Volume

Average Directional Movement Index (ADX)

Average True Range

Today’s High – Today’s Low

Source: Macquarie Research, April 2009

Technical factor performance

Quintile 1 Excess Return * 4.4% 3.8% 3.5% 3.7%

Long Short Q1-Q5 Rolling Excess Returns 50%

has broken out from a trading range

changes its price trends

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

performs consistently through time and

Earnings revisions and