Correlation and Trend Following
Ponder the statement:
“Statistics alone can never prove causality, but it can show you where to look.”
True trend following traders, if trading similar markets, will typically have very similar winning months and very similar losing months. A good historical example is the summer of 1998 (when Long Term Capital Management went bust). During August and September 1998 most trend followers had winning months. Interestingly, July 1998 was a losing month for most trend followers. Comparing monthly performance numbers of trend followers is best done through correlation analysis. Correlation, however, does not prove causality. It tells us where to begin the investigation. So when looking at correlations among trend followers, especially very large monthly gains or losses, it makes sense to look for the other side of the trade to better understand “why”.
