- Commodities trading offers a number of advantages for investors. Commodities can contribute to portfolio stability, serve as a hedge against inflation and have the potential for high returns. And “alternative markets” like commodities might just be the ideal place to deploy the strategy of trend following.
- Scott Kerson, the head of the quantitative trading arm of commodity-focused investment management firm Gresham Investment Management, says that commodities are particularly conducive to quant tactics.
- Are alternative markets superior to traditional markets? Do alternative markets have better-trending properties and higher Sharpe ratios within a trend-following context? It's time to find out.
Amid the complexities of the financial markets, trend following offers a pathway to navigate the ever-changing currents of investing — and commodities have long been popular among investors looking for stability and tangibility.
New York City-based Gresham Investment Management pioneered commodity investing via exchange-traded futures and forwards, predating both the S&P Goldman Sachs and Bloomberg Commodity Indices. Its Alternative Commodity Absolute Return [ACAR] Fund was the first alternative markets trend-following fund with an exclusive focus on commodities.
Scott Kerson, who runs GreshamQuant, the quantitative trading arm at the firm, says he built the ACAR Fund not just because Gresham is a commodities house. It’s because research shows alternative commodity markets are the best way to apply quantitative strategies. The ACAR is “rooted in what we talk about as the conviction versus concentration and trade-off,” he says. “I understood why commodities had the right properties for trend following, and I was willing to trade off concentration for conviction.”
Another driver of the ACAR Fund? “Commodities actually aren’t one asset class,” Scott says.
“If I trade coal onshore in China and I trade electricity in the West Coast of the U.S. and I trade liquefied natural gas delivered to Asia or the EU, those are structurally different things. They’re different points in time. At a molecular level, they’re different things. That lends itself to what we call structural diversification. It was really that combination.”
Scott joined host Moritz Seibert on an Open Interest episode of Top Traders Unplugged to discuss why Gresham believes alternative markets to be superior to traditional markets when it comes to trend-following trading strategies — as well as the firm’s approach to research in this space.
Read on for Scott’s take on the evolution of quantitative strategies in the commodities space and the future of the commodity trading advisor (CTA) industry.
Systematic investing and Sharpe
Scott’s introduction to quant trading, and to the genesis of the ACAR strategy, was “somewhat circuitous and it took a while to get there,” he says. He started his career as an academic economist, which led to commodities research and ultimately into fundamentally based discretionary trading.
He spent time at Barclays Global Investors before joining GLG, Man Group’s discretionary investment management arm, as its head of commodities. At the time, Man Group’s business was split between discretionary investing at GLG and systematic investing at AHL. He quickly realized that the two should be combined, which happened in 2012.
Scott says he owes a great debt to the team he joined, which had the insight, not long prior to his arrival in 2011, to launch what is now known as AHL Evolution fund. As he sees it, the underlying investment thesis of that fund was that classical trend-following techniques uncovered a fundamentally sharper signal (in certain classes of markets) than others.
Back when the fund was new, commodities brokers “looked at us like we had two heads” when the AHL team expressed interest in proving that thesis.
“That actually got us really excited — because that was the clear signal that nobody else was doing this,” Scott adds.
“Now we’re not the only ones doing this, but 10 years ago, that was a fairly radical insight,” he says. “Because if you rewind to that period of time, the prevailing wisdom — [which was] virtually pervasive across the space — was that, within medium-term trend following, the individual market information and individual Sharpe or information ratio at the market level was so low that really it was almost impossible to distinguish between markets.”
That’s critical because from a portfolio optimization standpoint, max diversification is key.
And “if you have relatively little information about any individual market, max diversification leads you to essentially as many markets as you can possibly trade,” says Scott.
Let’s talk about alternative markets
Before they launch into the nuts and bolts of quant strategy, Moritz asks: Exactly what differentiates an alternative market from a traditional market? Something that might be alternative to one person may not be alternative to others.
“I’m going to give you a partially wooly answer to that, and that’s because I don’t think there’s a single definition,” says Scott.
Back in the early 2010s, there was a fairly simple, working definition for an alternative market: “If it was not in a traditional CTA portfolio, it was definitionally alternative,” Scott notes. “I think that has morphed over time.” One of the typically quant questions that led him to launch GreshamQuant’s ACAR strategy, was why this is the case.
As a former academic who spent a lot of time on the prop side of investing, “it was never intellectually satisfactory to me to just accept this received wisdom that, Well, if they’re not like this, they’re like that, and they seem to work better,” he says. “Since it goes directly against the grain of decades of conventional wisdom around max diversification, why is it that we have chosen, consciously, to restrict our diversification and trade this other class of markets?”
Duration, drift and diversification
He mentions that Gresham’s ACAR strategy is fundamentally rooted in two observations that are not specific to commodities. (“Commodities just happened to fit this definition rather than the other way around,” he says.)
The first observation is that the efficacy of the classical trend signal has everything to do with the underlying distribution of returns in the markets in which an investor trades.
“More specifically, it has to do with the preponderance of relatively large moves over the right kind of frequency,” Scott points out.
As a “stylized” example, he suggests a hypothetical three-month, half-life duration, average-hold trend signal. For this trend system to work, the market must trend over a three-month horizon.
“What we see, time and again, is — and it doesn’t really matter whether there’s upward drift, or whether there’s downward drift — that it’s not really about serial correlation returns,” he says. “A lot of these metrics quants point to aren’t actually at the root of it. Really … [it’s about] how fat the tails are and the duration of those tail events.”
Observation two was that “this max diversification argument is completely valid but was erroneously applied,” he explains. When people, including Scott and his team, ask why we restrict potential diversification, “it actually belies the point: that you can achieve at least as much diversification. There’s always a caveat, and the key caveat here is capacity constraints.”
However, if an investor can restrict their capacities, such that they can risk budget enough and allocate that budget sufficiently to enough small markets, it’s possible to not only generate more diversification than a traditional market CTA would, but also more consistent diversification.
Conditions and correlations
“One of the bugaboos I have, and I think all quants have … one my team hears day in and day out, is the importance of looking at conditional distributions rather than unconditional distributions,” says Scott.
The unconditional distribution of diversification is essentially about identifying the long-run average pairwise correlation of one’s assets.
“But that actually isn’t really what matters in the real world,” Scott adds.
What does? The conditional distribution. That’s because traditional markets tend to cluster. “So, when things go wrong, they all go wrong together,” he explains.
That means if we plot something simple — such as the rolling pairwise correlation of a multi-asset portfolio including FX, equities, bonds and commodities in the traditional asset sense, we find that correlations tend to spike.
If we do the same exercise with the alternative commodity space in particular, “not only is the average pairwise correlation lower, but the variability of that correlation is lower and more stable,” Scott notes.
The future of alternative markets
Moritz also asks Scott to share his outlook on the alternative markets CTA space.
That question “raises this fundamental tension between staying super disciplined and … retaining consistency in approach tempered with the self-awareness, if you will, that markets evolve, and market efficiency increases over time,” he replies.
Scott says that it would be naive to suggest that any investment style has a clear, consistent, ironclad advantage over other styles — short of proprietary locations like co-locating servers next to the floor of the exchange or something like that.
“In the broad sense, any strategy can only be expected to become more competitive and more populated and alpha [will] degrade over time,” Scott adds.
He thinks quant traders should take that challenge head-on.
“Development of new and interesting alternative commodity markets goes hand in hand with the most basic nature of human progress and technological innovation. We [at Gresham] don’t think that will change. But it also means we have to be willing to let go of stuff that we’ve done in the past.”
An apropos message as 2024 unfolds. Here’s to the future.
This is based on an episode of Top Traders Unplugged, a bi-weekly podcast with the most interesting and experienced investors, economists, traders and thought leaders in the world. Sign up to our Newsletter or Subscribe on your preferred podcast platform so that you don’t miss out on future episodes.