Last Updated on June 11, 2021 by Oddmund Groette
I guess most traders have heard that “the trend is your friend”. Personally, I close to never trade on trends, I go the opposite way: buy weakness and sell strength. However, this depends on the instrument you trade and of course on the time frame. In general, most currencies and commodities show a tendency to trend more than stocks and/or stock indices (at least according to my research). On the internet, you can find a lot about discussions which is the best strategy, mean reversion or trend following. For me this is a pointless discussion: There is nothing as best or worse: one has to do research and trade the result if it makes sense, be it mean reversion or trend.
I trade SPY/ES. I’m in the process of looking at some new strategies, but first I would like to check which time frame it’s best to look at (for mean reversion). So I checked how much mean reversion there is in this instrument. To do this I simply used a scatter plot to show the return for one period against the other. Comparing different time frames it seems like 2 days show the greatest mean reversion. Here is the scatter plot from 2005 up until July 2012:
The correlation is -0.12. As you can see from the plot there is an inverse relationship: if a 2 day period is positive, most likely the next 2 days will show opposite returns.
I have also tested this on some other markets:
EWA: Australia -0.07
RSX: Russia -0.08
EWZ: Brazil -0.11
I tested these results on a basket of ETFs: buy when RSI(2) drops below 10 and exit after two days. Both entry and exit are on the close:
As you can see the result is pretty good for such a simple strategy. However, what this does not show, is that the strategy shows weakness after the second half of 2010. 2011 would have been a loss as a portfolio, but 2012 holds up pretty well.
Doing the opposite, going short, shows a similar bar chart but with about half the average return per fill.
However, the mean reversion has somewhat abated over the last couple of years. Why? I have no idea. Perhaps it’s too many traders looking at the same thing? The markets always change, especially when there are clear patterns. Everyone trades them and they disappear.