Thanks man! I've thought about it, yes. Specifically because I think the LOAD system is somewhat crude. You can just tell visually that when the entire slope is down, the mean reversion is going to be less effective. I think there is something to be said about the distance of the price to each signal and that could be used as some sort of signal as well.
That being said, I didn't try weekly or monthly rebalancing. But, because the window is so long (180), the turnover is very low, as shown by the transaction analysis plots.
I think that making an online portfolio selection strategy that keeps a ranking of the assets that are most mean-reverting / momentum based would improve the strategy as well.
Link I followed read "How I Fused Momentum and Mean-Reversion to Achieve 20% CAGR on ETFs Since 2000 [Paper to Profit]" , but 90% of the alpha of your writeup seems to come from individual stock selection, not ETFs - maybe link needs fixing?
Curious how you ensured a bias free stock selection process; pretty easy for a system to focus on picking surviving winning stocks without the right structure in place.
Really interesting fusion of momentum and mean reversion—two strategies that often seem at odds. How do you manage the transition points where momentum fades but the mean hasn’t yet asserted itself? Curious how you handle signal conflict in real time without overfitting or hesitation. Great read!
Momo is great. Avoid low beta stocks.
Appreciate the work. I love the idea of combing the momentum & reversal by different regime
Did you try weekly or even monthly rebalancing? I daily rebalancing may be too sensitive to cost/b-a spread...
I think it would be nice if there is any analysis on the correlation between the LOAD, pure momentum and reversal strategies.
Thanks man! I've thought about it, yes. Specifically because I think the LOAD system is somewhat crude. You can just tell visually that when the entire slope is down, the mean reversion is going to be less effective. I think there is something to be said about the distance of the price to each signal and that could be used as some sort of signal as well.
That being said, I didn't try weekly or monthly rebalancing. But, because the window is so long (180), the turnover is very low, as shown by the transaction analysis plots.
I think that making an online portfolio selection strategy that keeps a ranking of the assets that are most mean-reverting / momentum based would improve the strategy as well.
Link I followed read "How I Fused Momentum and Mean-Reversion to Achieve 20% CAGR on ETFs Since 2000 [Paper to Profit]" , but 90% of the alpha of your writeup seems to come from individual stock selection, not ETFs - maybe link needs fixing?
Curious how you ensured a bias free stock selection process; pretty easy for a system to focus on picking surviving winning stocks without the right structure in place.
Really interesting fusion of momentum and mean reversion—two strategies that often seem at odds. How do you manage the transition points where momentum fades but the mean hasn’t yet asserted itself? Curious how you handle signal conflict in real time without overfitting or hesitation. Great read!
Check out the reference paper to understand how exactly! It has to do with a simple regime filter and if the price is breaking recent highs or not.
Ah i see it now! Thank you for sharing! 💥
Interesting. I am looking for how to download the code.. I'm a paid subscriber but i don't see any links or emails with code?
Hi Bruce, I'm a new subscriber and had same problem, but just found the link to the Google Drive over here: https://substack.com/@papertoprofit/note/c-117024887 Click the Google link at the bottom.