FM: But most of the allocations to managed futures are concentrated in a few institutional managers, like Winton. Isn’t there a lot of capacity in emerging and non-institutional CTAs?
JS: Yes, that’s an old problem. I consult as a portfolio manager and the managers I recommend are smaller managers. We have very few large managers. Ironically, the one large manager we have has by far been our worst performer. You’re better off [investing] in smaller managers. It is very difficult for these giants to deliver great returns to risk. Also, a lot of the big guys are very correlated. So for those reasons, I personally gravitate away from the larger managers.
FM: After two tough years for CTAs, we are hearing once again that trend-following is dead. This is the third time since I have been following CTAs that I’ve heard this. Do you think the markets have changed in a material way or has it just been a poor environment that will change because everything in trading changes?
JS: The markets have changed. You have a tremendous amount — a much larger amount — of money using trend-following techniques than what was the case earlier. The fact that there is so much money in this strategy inevitably hurts returns. That makes trend-following more difficult. However, there are some very strong fundamental reasons why trends should exist. The key reason is the way markets fundamentally work. Central banks may make major policy changes. You get those all the time; countries make major changes in financial policy that will change the direction of interest rates and currencies. Those policy changes that are supported by action will cause trends and will cause trends to persist. Commodities aren’t [affected] by government as much but you have surpluses that cause mines to close down and with the lag; world demand then goes up. Supplies are not increasing so that surplus creates a new shortage and starts a new process so you get these cycles in commodities where the actions to rectify extremes can only be affected with a lag; so you get trends. However, if too many people are trying to exploit those trends then those trends become much choppier. It doesn’t make those trends disappear but it makes them much more difficult to capture.
FM: A point I have made over the years is that trend-following is a pretty broad description of a group of approaches used to find an edge in futures markets and there is more diversity in that space than many people realize. Do you agree?
JS: Actually, I would almost disagree. If you look at different strategies, one of the strategies that has the most correlation among managers is trend-following. Trend-followers as a group, correlate with each other in bull markets, bear markets, sideways markets; year-in, year-out there is a much more stable correlation. If you took some other strategy like merger arb or convertible arb and look at them, there will be periods where they will be correlated to each other and periods when they [aren’t as correlated]. But with trend-followers the correlation is very stable. There is some diversification there but not a ton so what I would consider for example in my portfolio is I don’t want all trend followers or most trend followers because it would kill diversification. But I have a certain segment of trend-followers and within those trend-followers I am looking at for doing qualitatively different things to capture the trend. When I look at the average pair correlation if it is something like 0.4-0.5 that is still significantly correlated but it is a lot better than 0.7 or 0.8 so you can find managers who are doing things differently.
FM: There is a tendency in professional investing to try and benchmark everything and set a beta for various types of investments. While there are better and worse environments for various strategies, this approach tends to eliminate qualitative differences among managers. Is this a mistake?
JS: The CTA space becomes much more differentiated as people like myself look to find strategies that are uncorrelated to trend-following. It has become much more differentiated and there are a lot of strategies out there other than trend-following. In my own portfolio I make up my own sub-categories: Fundamental systematic, trend-following, fundamental discretionary, technical discretionary spreads, mean reversion, intraday, systematic vol arb, pattern recognition, short-term systematic; all of those are different from trend-following.