De vruchten van een gediversifieerde aanpak

Meer dan posities: hoe handelsactiviteit het rendement van een trendstrategie bepaalt.

augustus 20239 min leestijd

Harold de Boer

Managing Director / Head of R&D

Dit artikel is alleen in het Engels beschikbaar.

The fruits of a diversified approach

De vruchten van een gediversifieerde aanpak

Download PDF

At any given time, our Diversified Trend Program (DTP) holds a wide variety of directional positions in different futures markets. Some of these positions are longs in equity markets. Sometimes, potential investors ask us if we can offer DTP without long equity, because they already have a lot of exposure to stock markets. We understand why they ask this.

Similarly, some of DTP’s positions in currency markets are ‘long carry’; that means, longs in high-yielding currencies against shorts in low-yielding currencies. Many investors also hold similar positions through other strategies. This leads to requests if we can exclude this part as well, and offer DTP as a ‘pure trend’ program. We understand this request too. But in what markets should we trade this ‘trend’?

The thing is, most of the positions in DTP are also held by other, more specialized, investment strategies. And to make things worse, many of these specialized managers are much better at their specific strategy than we are. Most long equity managers are better at investing in equities than we are. Most specialized energy traders are better at trading energy markets than we are. Most long vol managers are better at trading volatility than we are. And okay, we do trade synthetic market combinations, but that’s not very unique either. Some of these synthetics are for example aimed at trading the yield curve in bonds; there are many specialized managers that are better at that than we are. Why would anyone want to hold positions through a CTA trend program, when they can have the same positions through investments with more specialized managers instead?

Why would anyone want to hold positions through a CTA trend program, when they can have the same positions through investments with more specialized managers instead?

We can’t deny it. If you only look at the positions we hold, investing in a trend program doesn’t make sense. However, positions in futures are not simply held, if only because these contracts expire. Futures contracts have to be traded actively. And the returns of a managed futures program come from selling the contracts at a higher price than buying them, or vice versa.

There are of course more important reasons for the high trading activity in DTP. It may be true that the positions in the program at any given time look like a mix of positions also held by other, more specialized, investment strategies, but the effective ‘allocation’ to these strategies changes a lot over time. For example, in 2021 and the first part of 2022, when inflation was a big theme, DTP had large long positions in commodities, including crops. But as we explained in our April 2022 article “Investing in commodities as an inflation hedge?”, most of the time DTP is mainly short in crops. Also in other ‘asset classes’, DTP can be long or short. And even the allocation to these asset classes, regardless of sign, is not fixed. When positions in different asset classes are part of the same risk concentration — like it was with long commodities and short bonds during the recent inflationary environment — we take that into account when sizing the positions in each market.

This is probably why, already before 2000, a former trend following CTA started to promote its investment program as a tactical asset allocation program, without actually changing its investment approach much. But again, you might wonder, wouldn’t it be better for professional investors to do their own tactical asset allocation to the various investment strategies? That way, all their money would be in the safe and reliable hands of specialized managers.

Drifting on capital flows

We agree it probably wouldn’t be too hard to replicate the returns of CTA programs in a simple spreadsheet experiment by dynamically allocating to a mix of specialized investment strategies. However, the hypothetical returns of such a combined strategy would depend a lot on the assumed implementation costs. These include the assumed market impact of the required capital flows in and out of the various strategies. It seems to be widely accepted that these implementation costs would be very low. One common argument for that is that the investment vehicles used would be liquid. This is of course irrelevant, because the liquidity of an investment vehicle only refers to the ability to add or withdraw, not to the market impact of the implied capital flows. Regarding the impact of these capital flows, a common argument is that it would be marginal, regardless of the timing of the execution, because of the liquidity of the markets traded. Such claims basically ignore the fundamental fact that order flows are the one and only force that makes markets move. Order flows are the gravity in markets; one can deny gravity, but that doesn’t stop one from falling.

We believe that in real-life trading, these implementation costs would be far from low, even if the various investment strategies were managed by different teams of the same investment house. Capital cannot be moved out of one market and into another market instantly without leaving traces. In the markets we trade, we see these traces almost every day. They are left by market participants who think they can move capital without impact. This includes the most liquid markets in the world. These traces are the largest in volatile environments.

The performance of a CTA trend program can (probably) be explained well in correlation terms by a dynamic mix of specialized investment strategies. But its performance in absolute terms is largely determined by the trading activity.

From this, we conclude that the performance of a CTA trend program like DTP can (probably) be explained well in correlation terms by a dynamic mix of specialized investment strategies. But its performance in absolute terms is largely determined by the trading activity. That’s why DTP is a trading program and why Transtrend is aptly called a trading advisor. Again, the returns of a managed futures program come from selling futures contracts at a higher price than buying them. Given that capital flows are the only force that makes markets rise and fall, we can only succeed in that if we avoid buying when everyone else is buying and avoid selling when everyone else is selling.

Trendvolgen — regenjas of kapstok? (2020)

Lees artikel

Resting on a pillow of diversification

This can best be achieved with a diversified portfolio. In fact, over the years, we’ve learned to appreciate that the ability to move independent of (and often opposite to) the crowd is the greatest strength of a diversified approach. To understand this, we have to consider why investors prefer their capital to be spread over different investment strategies in the first place. It’s because they realize that every specific investment strategy can face (periods of) significant losses. What causes these — often short-term — losses? Investment flows. A stock doesn’t sell off because the company publishes disappointing results; it only sells off because it’s massively sold, whether in response to disappointing results or not. Who are most likely to sell in a sell-off? Investors who hold large positions in the specific market. Especially managers of specialized investment strategies can, rightly or wrongly, feel forced to sell in a sell-off. That could be for risk management reasons. But it could also be due to allocators withdrawing investments.

Some people argue that professional investors wouldn’t sell in a sell-off. We believe that many of them indeed wouldn’t do so. But the fact is that a market only sells off when it’s massively sold. If the market recovers after the sell-off, the only participants who lose money are the ones who sold in the sell-off. They essentially paid a liquidity premium. And that liquidity premium was earned by the participants who bought in the sell-off. We consider liquidity premium to be an important part of the risk premium that we are after in our investment program.

We consider liquidity premium to be an important part of the risk premium that we are after in our investment program.

Investors who hold a diversified portfolio are less likely to be forced to sell in a sell-off. This is especially true when the losses from that sell-off are offset by profits in other markets, whether in other strategies or not. That is one of the key principles in our Diversified Trend Program. The most important thing we have to do for that is to avoid getting distracted by losses on individual positions, which could happen for instance when different markets in the program are managed by different teams. What is required is a fully integrated approach focused on the portfolio as a whole. This offers opportunities. For example, to preserve the balance in the portfolio during volatile times, we often need to add to the losing positions and reduce the winning positions. This can be done by buying in sell-offs and selling in price explosions. And that way, we earn a liquidity premium instead of paying it. In fact, most of the transactions in DTP do not result from trend signals to enter or exit trends in individual markets, but from portfolio risk management signals aimed at preserving the balance in the portfolio. And most of these transactions are of the kind ‘selling in rising markets’ and ‘buying in falling markets’.

Navigating rough waters

CTA trend programs like our DTP are often seen as ‘defensive strategies’ in the sense that these programs can make money when other strategies are struggling. Strictly speaking, this can’t be correct, since this would require a trend program to hold positions opposite to those of all investment strategies (potentially) in trouble. In reality, in our more than 30-year history, DTP has faced almost all the same challenges that from time to time hit one or more specialized investment strategies hard. The real defensive character of the program has always come from the diversification within the program, due to which significant losses were offset by more significant profits on other positions and/or at another time. The only thing we have to do for that is ensure that the effective diversification within the program is maintained, which can be a challenge sometimes. What is often called ‘crisis alpha’ ultimately isn’t simply explained by the positions held, but to a large extent also comes from earning a liquidity premium — or at least not paying it — in times of distress.

Alpha onder de loep — deel 3 (2019)

Lees artikel

Let us illustrate this dynamic with two recent examples:

  • In early 2022, the main theme in the financial markets was global inflation. It was initially led by rising commodity prices, which made various central banks raise their key interest rates. DTP was positioned in this trend with long positions in commodities, short positions in interest rate instruments, and long positions in the currencies whose central banks were early in raising their rates. These were basically carry trades. This included sizable long positions in Eastern European currencies: the Polish zloty, the Hungarian forint, and the Czech koruna. In the last week of February, these currencies sold off after the Russian invasion of Ukraine. However, DTP did not sell these currencies. Instead, we bought some more. Thanks to the diversification in the program, these losses were more than offset by profits on other positions, especially the longs in commodities. And as these commodities became more correlated, we sold part of them while they were rising. This way, we didn’t pay a liquidity premium for selling the falling Eastern European currencies but instead received a liquidity premium by selling the rising commodities as well as buying some more Eastern European currencies.

  • Sometimes, significant losses in our program are immediately offset by profits on other positions. But that isn’t always the case. For example, in early 2023, bond yield inversion was a major theme. In March, when the U.S. bond yields inverted more and more, it triggered a banking crisis. Many markets were caught off guard by this. For DTP, this caused sizeable losses across the full width of the program’s largest risk concentration at that time. This included synthetic positions aimed at a further inversion of the U.S. and German yield curves and long positions in bank stocks. In this case, the losses weren’t offset by profits on other positions. As a result, DTP lost significantly in March. However, primarily by preserving the diversification in the program, we more than recovered these losses in April. While in this case the possibility to earn some liquidity premium by reducing winning positions was largely absent, we managed to limit the amount of liquidity premium paid, which helped the recovery.

DTP is short for Diversified Trend Program, which is, as the name suggests, a type of trend program. Some people argue that trend programs are a specific asset class that profits from trends. However, from an absolute return point of view, we believe that the ‘D’ in the program is the more rewarding source of return. In markets tossed back and forth by capital flows, DTP reaps the benefits of diversification. And over the years, we’ve learned this can only be achieved if we really focus on it.