Engineering of features with economic value, also known as attributes, predictors or alpha factors, is the first step in the extraction of market alpha. The example in this article shows proper classification of DJIA stocks in the weekly timeframe has resulted in a strategy with MAR equal to 2.81 since 2013.
More details of how to develop and execute long/short strategies in the weekly timeframe with DLPAL LS software are offered in this article along with performance results.
Below is the output of DLPAL LS as of the close of Friday, December 1, 2017, which was also included in our premium report:
We use 5% profit target and stop-loss to calculate the weekly features. P-long is he long probability, P-short is the short probability, P-delta is the difference of the two and S is the Significance of the result (the higher the better.)
The DJIA index fell about 0.39% from of open of Monday to the close of Friday of the following week that ended Friday, December 8, 2017. On the other hand, our weekly strategy using the above features and based on a score equal to P-delta times significance being greater than 50 for long and less than -50 for short, generated a 0.40% gain with the selected long and short stocks, as shown below:
Below are charts of the performance of the strategy since 2013. First the equity and underwater equity charts. Click on images to enlarge.
CAGR is 10.16%, maximum drawdown is -3.62%. MAR is 2.81.
Below are charts of monthly returns and results of a Monte Carlo simulation for % drawdown:
Year-to-date return is about 13%. This accounts for commission of $0.01 per share. The Monte Carlo simulation results show that the probability of a drawdown greater than about 9% is less than 5%.
Click here for more details of how hedge funds can download a free fully-functional demo of DLPAL LS to test for a month.
We have a sales agent program for DLPAL LS and for those who would like to market the software and earn 50% commission. Click here for more details.
If you have any questions or comments, happy to connect on Twitter: @priceactionlab