Dit artikel is alleen in het Engels beschikbaar.

The reasons why we and many others in our industry take pride in trading systematically are known. The systematic nature of the strategy grants it a certain reliability. The historical performance is not the outcome of a series of arbitrary trades that happened to be successful, but the outcome of a well-defined, verifiable and repeatable process based on the disciplined application of thoroughly tested rules. And indeed, if we narrowly focus on the decisions of what to buy or sell, how much to buy or sell, and when to buy or sell, all of these decisions are fully rules-based.

But what if we dig a little deeper? These rules are only applied to the markets traded by the program. How do these markets end up in the program? We manually add or remove them, in a clearly discretionary way, based on our view on these markets. We’ve never used a system that tells us which markets to trade. And what about the applied rules themselves, where do they come from? They obviously did not fall from the sky. We defined them ourselves, again with full discretion, based on our ideas and our understanding of the market. And the past 30 years we’ve adapted these rules whenever we felt it necessary.

When I was a kid I tried whether I could fly. I spread my arms, flapping them like wings while running across the meadow. On tired legs I had to admit that I wasn’t a bird, nor a bat. However, because some far more inventive kids made it possible, people can fly after all: in a plane.

Essentially, our basic approach rests on formulating a view on the functioning of markets and the current market environment, and continuously challenging and readjusting those views. We do so in a fully discretionary way, with the aim to trade the markets in accordance with our views in a systematic way. We are able to trade in such an adaptive way because of the nature of the relationship with our clients: we have trading discretion over their accounts.

When we offer our trading program to investors, we essentially invite them to fly in our plane. A plane that we’ve been operating for many years. Our investors can rest assured they will really fly in this plane. They don’t have to fear seeing me or one of my colleagues spreading our arms, desperately trying to lift the plane in the air. Our program is a fine piece of technique.

This trading discretion of course doesn’t imply that we can do whatever we want. Our clients authorize us to invest for them pursuant to our Diversified Trend Program or a customized version thereof. And although we use trading rules when we implement the program, we don’t sell these specific rules. We offer the whole program, which includes the services of our entire team, who implement the program. In this respect, our service is somewhat comparable with that of an airline. It uses planes, but it doesn’t sell planes. It offers flights. And it’s the responsibility of the people working for the airline to ensure that its planes reach their destination.

And yet, it is us — human beings — who are flying that plane. It always has been. With respect to basic questions like: “Why are we long lean hogs?”, or “Why is the long position in EU CO2 emission allowances that sizable?”, or “Why was the 2-yr T-note contract sold at this particular price?”, we realize that the answer “Because it is in accordance with our program.” may offer comfort. These investment decisions are the result of following rules that have been thoroughly tested, in historic simulations and — more importantly — in real life trading.

However, if I ask these exact same questions to our staff and their response would be: “because the system says so”, we would have a serious issue. We ourselves are responsible for the strategy including the rules that are part of it. It’s essential to be the masters instead of the slaves of our machines. This again can be compared with an airline. To questions like: “Why are we flying over Iceland?”, or “Why are we flying at this altitude?”, we wouldn’t want the pilot to answer: “because the autopilot did so.” If the pilot didn’t agree with the course or altitude, we would expect the pilot to have intervened. Making such decisions is their job, their responsibility.

Fact is, every now and then a plane crashes. And indeed, investigations of such crashes often uncover a human error. Does this imply that the human element can better be eliminated? Absolutely not, because most of the time the human error is not about intervening, but about not intervening. For instance, a pilot who should have taken over the misled autopilot but didn’t (or wasn’t able to, as was the case with the dramatic crashes of the Boeing 737 MAX in 2018/19). But the pilots aren’t the only persons responsible for a safe flight. Whenever a plane has landed, technicians thoroughly check the plane and, if necessary, repair it before it is allowed to take off again. If these people do not intervene correctly the plane could crash as well.

The introduction of the autopilot has been an important technological advancement in the airline industry. And it came with a fundamental discussion about the role and above all the responsibility of the human pilot. (A similar discussion is now going on with respect to self-driving cars.) I know people who prefer pilots not to be able to overrule the autopilot; an important argument is the Germanwings tragedy in 2015. I respect this view, but I don’t share it. I trust people more than machines, if only because people can be held responsible for what they do. And equally important: for what they should have done. The same holds for investing. I believe it is vital that our people can take, and do take responsibility for their part in the investment process. For the things they do, as well as for the things they could have done but didn’t. Eliminating discretion allows us to not intervene in situations where an intervention would actually be the most prudent course of action. It essentially relieves our staff from their responsibility. (Which puts the slogan “eliminating the human emotion from the investment process” in a somewhat different light.)

In our industry, technological advancements have also blurred the view on human responsibility. The move from floor to screen trading has been an important catalyst, enabling execution algorithms to replace human traders. This seems to have removed an important part of the human element from the execution process. But does this also eliminate the human responsibility from this process? And if not, should human intervention in this process then be regarded as a discretionary input? And would such discretionary input make the strategy less systematic? And therefore less reliable? In the age of floor trading, all execution was done manually, but no one questioned whether that undermined the systematic character of the strategy. So the contours have been shifting.

If we listen to the various voices in our industry, we now seem to be at a crossroads. Some systematic managers vocally distance themselves from anything that sounds like discretion, while others are open about their use of discretion when the circumstances ask for it. Having said that, I believe that the fierce resistance against discretion to some extent is just a matter of positioning. (Which, you could argue, is not that different from the “100 percent systematic” answer in our AIMA questionnaire.)

But it isn’t just marketing. It represents a fundamental choice every systematic manager has to make. I believe that responsible investing and the use of discretion are actually two sides of the same coin. You cannot accept one side without accepting the other. Our position is clear: systematic trading is not flying on autopilot. “The system says so” can never serve as an excuse for irresponsible behavior. We have learned to listen to the system in exactly the same way as that millions of people listen to music every day: we only listen when it sounds good. If it doesn’t sound good, it is our responsibility to adapt it. And to clarify this position in the next update of the questionnaire, we will change our answer into “100 percent systematic as well as 100 percent discretionary”.