Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Why mean reversion has such bad performance

I am still very sceptical that algorithms can beat market index on a long run. I am yet to find an algorithm that consistently beats the market (in bull and bear periods). This particular algorithm which uses mean reversion lost more than 60% since January 2015. Any clues how parameters in this algorithm can be tweeked to better the performance.

2 responses

its actually really easy to beat the market with different value/growth strategies. mix that with the mantra 'switch to tlt' and you can double the market or more over time
however, i totally agree if you are trying low beta/uncorrelated/high frequency/zscore stuff

Hi Tyler, many thanks for your quick reply. Is there any algorithm on quantopian that does what you mentioned? I can clone that and test. Also I noticed that I used an initial capital of 10k. I get +60% if I use a 100k capital and +50k if I use 50k capital. I guess it has to do with the high price stocks that were part of the selection basket for the algo.