I have been playing with my strategy for a while. I am wanting to use like $5000 capital base for my algorithm. I have created a fundamental algorithm that buys 10 stocks monthly loosely based on Benjamin Graham fundamentals and indicators. It has a %10 trailing stop loss. I do well on the 2008 and 2010 backtest. But when I start in 2014 or 2015 the results are poor.
I am looking for suggestions on why my algorithm is performing well starting on certain dates and not on others. Am I missing a statistical model to find more correlations to generate alpha? Any help would be greatly appreciated.