Hey everyone,

 

I've been testing a strategy that I am developing where I Short stocks that showed a certian behavior at the first few houres of the day and close the position at the end of the day.

When backtesting the algorithms behavior over the last few years, I have noticed that:

Between 1.1.2015 - 1.7.2017 the strategy brings poor results - low succses rate (54%) and negative return

Between 1.8.2017 - 1.1.2020 the strategy brings good results - very high succses rate (78%) and a positive return

Since i'm picking all of the stocks by filltering them for a certian behavior (tradeing different stocks every day) I expected to see "uniform" statistical results from 2015 - 2020.

I can't figure out what had changed!

 

Can someone tell me what should I check for to understand the difference?

Thank You!