A recent article on the Pensions and Investments website (Crisis Alpha Everywhere) written by Katherine Kaminsky, the Director of Investment Strategies at Campbell & Company, looked at “crises” across various assets, noting that one can usually find a crisis somewhere within the investment landscape at any given time. “Crisis” was defined by Kaminsky as the 10 worst quarters (independently viewed) for an asset between 1990 and 2016. The conclusion was that trend-following tends to do better as more assets are in “crisis”.
Adding to Kaminsky’s piece, RCM extended the analysis (see “What is Crisis Alpha?”) by looking at individual years going back to 2007, showing that as the number of assets in crisis increases, returns to CTAs also tend to increase, though the correlation isn’t perfect. Given our recent posts looking at both trend-following and counter-trend performance through time, it seemed logical to see how well a simple equity counter-trend system1 captures “crisis alpha” as measured by RCM. Additionally, I added the SocGen Trend Index to the analysis.
As can be seen above, the SocGen CTA Index returns have a correlation 0.55 with the number of assets in crisis. Admittedly, correlation on annual data with only 10 data points is a weak measure, but for our purposes, it is still illuminating—and looking at the SocGen Trend Index, the correlation increases slightly to 0.62. However, a Short-Term Counter-Trend basket on various global equity indexes showed an even higher correlation, at 0.71. Finally, a blend of 50% trend and 50% counter-trend comes in with the highest correlation, at 0.79.
While no approach outside of dedicated bear market strategies can really guarantee “crisis alpha”, it’s clear that both trend-following and counter-trend strategies can serve as diversifiers and return producers during times of market stress, while delivering positive expected returns during more sanguine environments. As I’ve stated many times before, when the inevitable change in volatility and market direction occurs, investors in trend following and short-term counter trend strategies should be well positioned to benefit from the low to negative correlation provided by those allocations.