3. Strategy Overview
The strategy is designed to exploit the mispricing of the
paired futures.
Theoretical mathematic and economic model
foundation
The residual in the risk factor model is considered as the
spread.
Signals in entering ,adding and existing position
Backtesting results show over 22.75% annual return and
1.86% monthly return.
Appropriate risk control management
3
4. Model Construction
New approach to risk factor model
Spread stochastic model by Elliott, Hoek
and Malcolm (2005)
Use the model to find the equilibrium of
the residual spread.
Our model overcomes weaknesses in
other models in this field.
4
5. Trading Signals
(2012-6-1 to 2012-6-30)
5
31May 05Jun 10Jun 15Jun 20Jun 25Jun 30Jun
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
Spread, Trading Band and Trading Signal
Spread
Observed Spread
Estimated Spread
Sell band
Upper stop loss limit
Buy band
Lower stop loss limit
Buy
Sell
7. Return Performance
(2012-6-1 to 2012-6-30)
7
31May 05Jun 10Jun 15Jun 20Jun 25Jun 30Jun
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
Daily Return of the Strategy
Return
8. Return Performance
(2012-6-1 to 2012-6-30)
8
31May 05Jun 10Jun 15Jun 20Jun 25Jun 30Jun
-0.04
-0.02
0
0.02
0.04
0.06
0.08
Cumulative Return of the Strategy
Return
9. Backtesting
2003-01-01 to 2013-02-28 daily data
Testing Portfolio :
WTI Crude Oil Contract CL1
Natural Gas Contract NG1
Leverage constraints that capital
9
10. Superior Backtested Returns
Unique trading strategy, designed to exploit market
information and inefficiencies, yields excess returns over
the past 10 years
Strong and extensively-tested quantitative foundation
Investment process emphasizes robustness and risk
management
--- 22.75% Annual Returns
--- 4.22% Annualized Standard Deviation
--- 5.39 Sharpe Ratio
10
12. Backtesting
Return Analysis
Total Trade Numbers 491
Annualized
Standard Deviation
4.22%
Average Monthly
Trade Numbers
4
Average Monthly
Standard Deviation
1.02%
Annual Return 22.75% Best Monthly Return 40.25%
Average Monthly Return 1.86% Worst Monthly Return -17.31%
%Positive Months 70% Max Drawdown -17.51%
%Positive
12-Months Rolling
71% Sharpe Ratio 5.39
12
13. Backtesting
Sharpe Ratio Analysis
Trading Strategy
Crude Oil Price
Index
NYSE AMEX
Natural Gas Index
Annual Return 22.75% 13.82% 16.10%
Annualized Standard
Deviation
4.22% 4.97% 5.09%
Sharpe Ratio 5.39 2.78 3.16
When compared to returns of the benchmark, it offers:
• Higher returns
• Lower standard deviation
13
16. Backtesting
Monthly Return VS Benchmark
16
Feb2004 Jul2005 Nov2006 Apr2008 Aug2009 Dec2010 May2012
-1
0
1
2
3
4
5
6
7
8
Return
Cumulative Monthly Return of Trading Strategy VS Benchmark
Pairs Trading Strategy
Crude Oil Price Index
NYSE AMEX Natural Gas Index
17. Backtesting
Annual Return VS Benchmark
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Crude Oil Price Index
NYSE AMEX Natural Gas Index
Trading Strategy
17
18. Backtesting
Monthly Return VS Benchmark
Crude Oil Price Index
NYSE AMEX Natural
Gas Index
Alpha 0.018586 0.019103
Beta -0.0001158 -0.034482
Correlation -1.4535e-04 -0.0348
• Positive Alpha
• Negative Beta
• Low correlation with the market
18
19. Backtesting
Performance in 2008 Crisis
19
Feb2008 Apr2008 May2008 Jul2008 Aug2008 Oct2008 Dec2008
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Return
Cumulative Monthly Return in 2008 Crisis
Spread Strategy
Crude Oil Price Index
NYSE AMEX Natural Gas Index
20. Risk Control
Leverage control
Leverage: less than 20, and even smaller
during crisis time.
Position limit
Maximum position: cannot add more than
5 times of initial positions.
20
21. Risk Control
Stop loss limit control
Set an appropriate stop loss limit according
to the investor’s requirement and the analysis
from the market environment and adjust it
according to market dynamics.
21
22. Risk Control
Regular review of model
Review the market environment and data
obtained regularly to decide how frequently
to update our model parameters.
During crisis time, reset parameters more
frequently.
22