Back testing results
The backtesting is a simulation of how our trading platform would have performed
if it were used in the past. We do this by running the trading algorithm ex-post
for the period 1997 until 2008. As in the real-time version, the trading
platform decided what positions to open and close on each day of the simulation
period, given all the data it would have had to rely on the particular day.
In the following table and chart we compare our strategies to the S&P 500 Index
which is a common indication of the market performance, and to the Russel 3000
Index which includes all the companies covered by our platform. Each strategy
was tested individually. We started each simulation with an initial investment
of $150k and allowed an additional borrowing of $50k (which was
paid back with interest at the end of the simulation - interest expenses were not included
in the calculation).
Backtesting Performance Summary For Period of 2009-1998
|
|
S&P 500
|
Russel 3000
|
Long Only*
|
Semi Bullish*
|
Market Neutral*
|
|
Total Yield
|
-15.75%
|
-12.66%
|
1300.27%
|
1649.70%
|
340.96%
|
|
Average Annual Yield
|
-1.55%
|
-1.22%
|
27.12%
|
29.72%
|
14.44%
|
|
Average Monthly Yield
|
-0.02%
|
0.01%
|
2.19%
|
2.42%
|
1.29%
|
|
Monthly STD
|
4.63%
|
4.73%
|
5.78%
|
6.88%
|
5.68%
|
|
Correlations With S&P 500
|
1.00
|
0.99
|
0.33
|
0.25
|
0.05
|
|
Correlations With Russel 3000
|
0.99
|
1.00
|
0.35
|
0.25
|
0.03
|
* The yields include trading commissions.
All three strategies outperformed the market significantly during the
simulation period, which includes the financial crisis that started in July
2007.