I was going to fob you all off with a “Happy Thanksgiving” and retire – and indeed I do wish all our American readers the complements of the season – instead, however, I thought I would creak the brain into action one more time and discuss the problems in the CTA space.
Regular readers will know that while I am “always certain”, your correspondent doesn’t always get it right, however earlier this year I went up against P&L’s editor Galen Stops in a debate about trend following. I basically said it had had its day and Galen took the other side and, as things stand I am being proven right (and am equally taking every opportunity to remind Galen of that fact – as you would expect of me). The Societe Generale CTA Index paints a horrible picture and things are not much better when you look at the BarclayHedge Indices.
I actually think matters are even worse for the CTA industry than these numbers suggest, for this has not been a good year for equity markets, which are generally lower, although as my friend David Clark pointed out at our Singapore conference this week (and into the bargain reinforced his financial “anorak” credentials), the Lahore Index has been booming!
Back to the argument, though, because historically one of the big selling points for CTAs has been their ability to provide a hedge in times of equity market declines. This clearly hasn’t happened this year and I think it is because of modern market constructs. Things are happening much quicker than they used to which means that while trend followers would have had plenty of time to get into the appropriate position in years gone by, in today’s markets the move has already happened.
There is also, as I argued in our debate, a lot more mean reversion around than there used to be, and that can often lead to the trader’s nightmare of identifying the range…the wrong way round!
I don’t think the problem is, as some would argue, style drift, I just think that CTAs – if they are to thrive again – may have to take another look at their models. Not only does the decision making process need to be either quickened up (or, perversely, slowed down by taking positions in a much longer time horizon), but their execution of the strategies probably has to be speeded up as well.
It is interesting to me that in the latest data from BarclayHedge, currency traders were doing OK year-to-date (look away now though if you’re a crypto fan – just the 51% down on the year!) This is probably because, unlike many other market models, the OTC construct allows for market makers to thrive – thus meaning we avoid the endless emotional rollercoaster seen in other markets.
It is also harder to discern the model constructs for FX funds because while we may know the style, no one knows what currency pair the strategy is being expressed in – it’s altogether harder to “read” the funds. This is important because there are undoubtedly models out there that can replicate – and exploit – the trading patterns of many CTAs, this is where a fair bit of the mean reversion probably comes from (that and the rather chaotic nature of geo-politics).
Whichever way you look at it, CTAs are struggling because their strategies are either too visible or their models are inappropriate for the current geo-political, macro-economic landscape, not to mention the changing market structure. It is hard to know when, or if, things will improve for them, but there may be some good news out there.
I mentioned the need to improve the models and how they operate, and the answer may lie in one of the current buzzwords – AI. I do not yet subscribe to the theory that AI will change the trading world, I still happen to believe it creates crowded trades (in many ways it reflects the CTA world currently). Going forward, however, if CTAs can deploy Adversarial AI in particular, it may help them identify the weaknesses in their models.
It must be pretty depressing being a CTA at the moment because the largely abject performance is there for all to see and it is not clear how things will change. This is not to say that CTAs have had their day, of course, more it is the suggestion that “business as usual” no longer cuts it. As so many others in financial markets have discovered in recent years, one must evolve or die. The CTA model is not dead, but it could do with a seriously rigorous health check.