TTU139: Simon Judes, Co-CIO at Winton

[00:00:07] Simon

What we’re focused on is what matters and what matters is original research, to find new alpha sources and improve our existing ones and on risk management. Those are exactly the two things you mentioned. We’ve got a really exciting research department and research process and so there’s a huge amount going on all the time both in our trend, in our systematic macro and across our multi-strategy hedge fund with the long-short equities and the credit as well.

At the same time, our approach to risk management is the one I mentioned of hypervigilance and monitoring the whole array of possible scenarios and situations that could come up. So that’s what we think is important for 2023 and that’s what we’re going to continue to do.

[00:00:53] Intro

Imagine spending an hour with the world’s greatest traders. Imagine learning from their experiences, their successes and their failures. Imagine no more. Welcome to Top Traders Unplugged, the place where you can learn from the best hedge fund managers in the world so you can take your manager due diligence or investment career to the next level.

Before we begin today’s conversation, remember to keep two things in mind. All the discussion we will have about investment performance is about the past, and past performance does not guarantee or even infer anything about future performance. Also understand that there’s a significant risk of financial loss with all investment strategies, and you need to request and understand the specific risks from the investment manager about their products before you make investment decisions.

Here’s your host, veteran hedge fund manager, Niels Kaastrup-Larsen.

[00:01:48] Niels

Hey everyone, and welcome to another edition of Top Traders Unplugged, where today, Alan Dunne and I are joined by Simon Judes, Co-Chief Investment Officer at Winton as part of our mini-series focusing on the one investment strategy that beat everything else in 2022, namely, trend following and managed futures more broadly.

First off, Simon, it is really great to have you on the podcast. Thank you so much for joining us today. We really have been looking forward to our conversation. How are you doing? I guess you’re in London.

[00:02:14] Simon

I’m in London. Yeah, I’m doing very well. Thank you. Thank you very much for inviting me on.

[00:02:18] Niels

Yeah, absolutely. Now, before we dive into the various different topics, maybe it’s a good idea to always set the stage for our conversation so that the audience knows a little bit of background, so to speak, about Winton, even though I’m sure most people will be familiar with Winton. But maybe you could share just a few highlights as to the type of strategies you focus on nowadays and also maybe where the business stands as we’ve entered into 2023.

[00:02:44] Simon

Absolutely. Winton is a systematic hedge fund founded 25 years ago by David Harding. We employ about 200 people and we manage $10 billion in AUM. We’ve got three types of products. There’s a quant multi-strategy hedge fund, there’s a diversified CTA which is three quarters trend following and a quarter other systematic macro strategies, and then we have pure trend products as well. So that’s where the business stands right now.

[00:03:14] Niels

Yeah. Just out of curiosity, when did you join Winton?

[00:03:17] Simon

Almost 15 years ago actually. I joined in July of 2008, which was a great time to start in the industry and learning about markets.

[00:03:27] Niels

Yeah, absolutely. It’s funny, my first touch with real finance was around September 1987 and I think some people have said that you kind of get shaped a little bit in terms of the timing you enter an industry. So in my case, probably a little bit more risk averse than other people because of what happened. I don’t know if you ever thought about that, actually, whether the timing of joining the CTA industry sort of impacted you in some ways?

[00:03:57] Simon

Definitely, yeah. It affected David himself too because he co-founded AHL, first in the time of the ’80s, (when you mentioned) and traded through 1987 and that was obviously a formative experience. I started in 2008. You can look at other firms that started after 2008 and you can see that that led them often to make different decisions around how they parameterize their trends strategies and how they organize them. I definitely agree. People have been very influenced by their initial experiences.

[00:04:37] Niels

Yeah, absolutely. That’s a great point. I’m sure we’ll dig into that a little bit later. Yes, I do remember actually, David talking about that when I sat down with him and his co-founders of AHL a few years ago in Abbey Road Studios, and it was quite fun to celebrate their 30th anniversary of founding AHL by bringing them together. So people can go and find that on the website, and they should definitely listen to that three part conversations, certainly one of the most memorable that I’ve had in the last eight or nine years.

Anyways, enough of this. We do want to get to the meat of it. Simon, as we normally do, we share different topics between Alan and me, and as usual, he always has a first choice. So I’m going to send it over to you, Alan, and see where we’re going.

[00:05:25] Alan

Thanks, Niels. Simon, good speak to you. In your opening description, you touched on a couple of features of Winton’s trading, systematic diversified quant strategies with an emphasis on trend. So if you were to wrap all of that up into an investment philosophy that underpins everything, what would you say that is?

[00:05:48] Simon

Our philosophy is that you can discover and improve successful investment strategies by applying empirical scientific research methods. The key thing there is to approach markets with an open mind rather than with preconceived ideas and theories about how markets are supposed to behave.

Obviously, this is very strongly related to David’s initial experiences in founding AHL and then Winton where the preconceived idea that many people had was the efficient market hypothesis. That was a bit of a headwind for many years because many people just believed that implicitly. Clearly, the foundation of our strategies and our research processes is that that hypothesis is not right and that you can, through very hard work and scientific analysis, make good predictions about what markets are going to do.

[00:06:39] Alan

When you say scientific analysis, is the process then formulating hypotheses and rejecting them or not or more data mining? How would you describe the application of that scientific approach?

[00:06:56] Simon

Well, data mining is kind of a pejorative term very often, because what it suggests is that you sort of route around looking in data until you find something and that thing you find is probably at that point just a random fluctuation which doesn’t reflect reality and isn’t predictive of the future. So obviously, one of the key scientific techniques you have to think about is how you’re going to avoid falling into that trap. A lot of our processes and thoughts are about having a collective understanding of the relevant statistical techniques that help you in particular test multiple hypotheses at the same time and sometimes also sequentially one after the other.

So that is a technical level of competence that’s required in the organization and that affects things like who you hire and how you train people and so on. Some of it is also about process, avoiding situations where researchers privately overfit ideas and data mine and then only show you a subset that looked good in a backtest. So there are different aspects to it, but that’s what it looks like in practice.

[00:08:03] Alan

To put it more simply, I guess some people will have the idea that when they develop a strategy, they have to have an idea about how markets operate and why that might work, and you’re kind of saying that may not be your approach if you have sufficient technical, I suppose, or maybe statistical evidence that something is working, you’d be happy to go with a strategy like that. Is that fair enough?

[00:08:26] Simon

Well, look, what are you trying to do at the end of the day? You’ve got a strategy and you’re trying to improve it. You can improve it by adding a new signal. Suppose you come up with something, you want to have some confidence that this thing is actually going to improve your strategy. What is it you need confidence in? You need competence that its Sharpe ratio is good and it’s what you think it is and that its correlation with your existing strategies is low enough.

So how can you get that confidence? Well, one way is by looking at a backtest and analyzing it and seeing if, statistically speaking, it does actually make the improvement that you think. But there are other ways that you can also get more confidence. If there is, for example, an economic rationale behind a particular idea, then you can use that. You can test the intermediate economic relationships and that can add additional confidence to your assessment and that’s great if you can do that.

There’s absolutely nothing wrong with doing that and you should do it. But if you can’t do it, that doesn’t mean that you have to throw away the idea altogether. It just means perhaps you have slightly less confidence about it.

[00:09:27] Alan

If you were to describe the sources of opportunity you are capitalizing on now, I think it probably started in the trend following space. But in terms of where these opportunities arise, is it down to behavioral issues or economic relationships or market microstructure, people trading at certain times because they have to, that type of thing? Basically, who’s on the other side of the winning trades for Winton?

[00:09:54] Simon

It’s a real mixture and it depends a lot on exactly which type of strategy you’re looking at. As I mentioned at the beginning, in addition to our CTA and trend strategies, we have a quant multi-strat hedge fund where we trade long-short equity strategies, and more recently, we trade long-short credit strategies as well. So there’s an extremely diverse collection of alpha sources. It’s very hard to generalize about who in general is on the other side.

[00:10:21] Niels

I want to stay with the investment philosophy a little bit. I think our listeners would be a little bit disappointed if I didn’t address one particular question in terms of the findings that you came out with a few years ago or specifically David came out with a few years ago, where he, as I remember it, it was along the lines that perhaps you didn’t feel that trend was necessarily as strong a strategy as it had been in the past.

So I’m curious about if you can talk a little bit about how that came about in a sense. What was it that alerted you to draw that conclusion or led you to draw that conclusion at the time? But also, as I understand it, you’ve changed your view on this, and so I’d love to hear what changed your mind coming back into the full trend fold, if I can phrase it like that.

[00:11:21] Simon

Sure. Yeah. I don’t think we’ve really changed our view in any important sense. Obviously, David is perhaps the preeminent trend follower in the world so there’s a slight tendency to sometimes overreact to comments that he makes. I think what he actually said, in 2018, is pretty uncontroversial actually. He said that the long-term Sharpe outlook for trend following is perhaps around a half, and that on its own, that’s scarcely good enough to run a hedge fund on and if you really want to maximize the Sharpe ratio, then you want to look at adding other things too.

In terms of what analysis led up to that, it’s pretty straightforward actually. You look at the performance of trend following over time. Lots of people when they do this tend to focus on the period after the financial crisis and they say it didn’t work as well after the financial crisis as it did before. But actually, if you look over a much longer time horizon, starting in the ’70s which is at the point where a strategy like that became tradable in the modern form, then you actually find a consistent decline in the performance of trend, of the kind of trends strategies that people were trading at the time over the years.

So we first observed that in our research in around 2007 before I even joined the firm. It was already well understood. So the comments that they were making were based on that research that was done long time ago and that trend of decreasing performance had just continued and it had got to the point where we felt that the long term outlook was at that level, around a half.

Nothing has changed in that recently. Trend has always been our biggest strategy and it still is. The main thing that’s changed for us is that instead of having one product dominated by trend following, we now have three products, some of which are pure trend and some of which have a much lower allocation to trend following.

[00:13:20] Niels

Okay, that’s a great segue. I appreciate that and I completely agree that sometimes things can get a little bit misinterpreted by media. Although I will say, Simon, that, I have to say, it certainly caused a little bit of a challenge for me and I’m sure for others, because as soon as someone like David (with his stature) came out saying, well, maybe trend is not as good as it used to be, we had to work a little bit harder to work to keep them interested, I have to say. But anyways, you talk about different strategies, you talk about pure trend versus let’s call it diversified trend.

One of the topics that we’ve been talking to all our friends about is really partly captured in this paper that Cliff Asness came out with last year where he talked about, what’s the mandate? Is it dual, absolute return as well as quote-unquote “crisis alpha”, and also whether we as firms have kind of become too focused on the Sharpe, or at least Sharpe as a line item, because of course, it was not really created for that purpose and maybe we need to… I’m not sure they said that, but my opinion is we probably should change the conversation to talk about the improvement of portfolio Sharpe that these strategies provide and so on and so forth.

So I’d love to hear your thoughts about some of these comments and observations that AQR came up with last year.

[00:14:45] Simon

Well, on the one hand, you can see that we have both products in our stable. We have a pure trend product and we have a diversified product. So in a sense, we’re being open minded about that. I think that the focus on the mandate, or the Sharpe, is probably not quite the right angle to look at. I think we prefer to look at the investor experience. You look at situations where investors really needed diversification in 2008, last year, and trend has done very well there. It’s provided that diversification. Obviously, that’s not guaranteed for the future. But nevertheless, it has done that in the past.

But what’s also true is that in the intervening periods, there have been large drawdowns and it’s been quite hard for many investors to remain allocated in those periods. It’s important, I think, to understand that that’s not a criticism or a sign of weakness or something like that or an inability to keep the faith. It’s totally natural and understandable because the people who are making these allocation decisions have to justify each line item to their investment committees and it becomes very difficult to do that.

So there’s great value in adding signals which smooth out those drawdowns and obviously still preserve the behavior of trend following at other times like in those moments like 2008 and last year. It’s not definitely a trade-off in the way that I think Cliff Asness describes in that if you add other signals, you necessarily reduce the effectiveness in those periods. You just have to be a bit careful about what kind of other signals you add and also how much weight you give to them.

So from our point of view, that’s why we stick with three quarters trend following because we think that’s the right level where you get this drift that helps with the drawdowns but you retain the classic trend following profile.

[00:16:41] Niels

Yeah. We recently published a conversation with Katy Kaminski where she kind of reminded us a little bit about how this crisis alpha term came about. I don’t know if you listened to it. Maybe I’ll ask you, did you happen to listen to that one, Simon?

[00:16:56] Simon

I didn’t. I thought it came out when she wrote a book with crisis alpha in the title.

[00:17:01] Niels

Okay. So then I have to give you a little bit of context here because I’d love to hear your thoughts about it. Initially, when you just hear the word crisis alpha, you think about it’s a strategy that can create some kind of positive performance during a crisis. But then after a few months of people gravitating towards this term, in 2009, it became this constant debate about was it a crisis or was it not a crisis, and did you perform, did you not perform.

But in our conversation with Katy (which I encourage people to go and listen to) she actually says, well, that’s not exactly what was said back then because it came from a question of a pension fund saying, why do these strategies perform well during stress? She basically said, well, because there are three elements or characteristics that you need. One, they need to be liquid so you can in and out of your exposure efficiently and quickly. Two, it needs to be opportunistic so that you’re not stuck in the same asset class or in the same market, you can move around to where the action is, where the opportunities are. Three, it should have no bias. It can be long as easily as it’s short, and of course, a good example of that would be that in 2020, during COVID, the best trade was long bonds and short energy, but in 2022, it was exactly the opposite.

So I’m just curious whether this term crisis alpha, you think it’s helpful, what value you give or how would you explain it, if different, just because I think it’s something that has really stuck in our industry but maybe some of it has been kind of lost in translation, so to speak.

[00:18:47] Simon

Those three characteristics that you mentioned are prerequisites for something to function in that way, but they don’t guarantee that it’s going to work. For example, a strategy that had precisely the opposite positions to trend following would also be liquid, opportunistic, and not have a long bias, but it wouldn’t work in those situations. So it’s right, but it’s not enough. Obviously, the historical record of trend following is what it is. It has performed well in those situations and you can see why.

You can see that if there is going to be an extended trend, then trend following will latch on to it and do well in that environment. Equally, you can imagine scenarios where there’s a crisis, where trend following gets it wrong. For example, if the equity market is steadily moving upwards and then has a very sharp downturn, that’s the situation like perhaps in February 2018 where trend following is not going to do well. So how do you then react to a term like crisis alpha as a characterization of the strategy?

Well, obviously, it’s not very good as a pure description of the strategy because sometimes it will work in that way and other times it won’t. So from our point of view, the best thing is to really try and describe the strategy, how it’s going to react, what it’s going to do in different environments. We don’t then need to apply labels like that which have the potential to mislead.

[00:20:09] Niels

Okay, that’s a fair point. Now, before I turn it back to you, Alan, sorry, I’m taking up a little bit of extra time with this segment, but I do want to just also ask a little bit of good advice from you, Simon, and that is, I’ve been out traveling quite a lot so far this year and a lot of what I hear from investors, not necessarily current investors, but certain potential investors is they kind of look at trend following a little bit like a trade. They try to time it, and then also, they have this notion that it did so well in 2022, it’s either too late or certainly can’t do well in 2023, so to speak.

So I’m just curious, you guys are the masters of putting out great papers and write ups and so on and so forth. If you had to, how would you change those observations from people looking at the strategy? How could we not make them look at this as a trade or even have a preconceived notion about what we’re going to do going forward in an unpredictable future with a strategy that just adapts to whatever happens? How do we change that?

[00:21:19] Simon

That’s a great question. It’s obviously what’s on people’s minds. Again, I think it’s useful to be sympathetic to the investor concerns because even if you can’t time trend following, even if we agree that you can’t time when it’s going to work, there’s still a risk for the allocator that they make a decision and then immediately afterwards, it does badly. That carries a risk for them personally over and above the long term performance of the strategy. It’s important to be aware of that and not just reply with a series of statistical observations which don’t really address what’s on their mind.

Our view is, as you say, I think that there’s no reason to think that next year will be bad because last year was good. 2008 was a good year, but every year of Winton’s history prior to that had also been a good year. So you often get these long sequences of good years. But to the extent that people are worried about the risk of allocating at a particular point and then subsequently having underperformance, one mitigant is that you can average your way in, over a number of months, and that’s sometimes quite helpful as an idea because that often addresses the real underlying concern.

[00:22:44] Alan

Just to pick up on a few points that came out of that discussion, and maybe firstly, just in relation to you mentioned you’re a diversified CTA, three quarters trend, one quarter non-trend. We’ve got to get a sense of what’s in the non-trend, and specifically the point you mentioned around the Cliff Asness article, it matters which types of strategies that you add. One thing from that perspective is obviously, if you’re using strategies that are more economically related and slower moving, is that a risk when you’re thinking about reactivity in crisis periods?

[00:23:25] Simon

Yeah. Often, the question of this convexity and whether you lose it with these other strategies centers around carry because that’s one of the most common additions that people add, and we trade it too. We don’t trade it on equities though, because often, that does just add a long bias to your equity portfolio which definitely reduces that element of convexity on the trend side. The other point about carry is that a lot of people’s ideas about its properties come from its history as a currency strategy.

It’s certainly true that specifically focusing on currencies, you get small gains followed by large losses. But that actually is mitigated very significantly when you do it in commodities and in fixed income as well. So that’s our approach there. The other strategies in there are a real mixture of other things. Some of them are purely statistical nature, some of them have economic rationales, some of them are other technical signals like trend following but different from trend following, some of them are a bit faster.

There’s a lot of research going on there as well. So we’ve added several new signals in the last year.

[00:24:40] Alan

In terms of the three quarters, one quarter mix, why have you landed on that as the sweet spot? I guess it’s to deliver more consistent returns but maintaining the profile, but any reason why it’s not two-thirds, one-third or 80/20? Is there any observations around that?

[00:25:03] Simon

Well, look, those things are not in a sense very different. They depend on the opportunity set you have at the time with the non-trend strategy. So this is the level that we feel is appropriate now, but it’s also important to fix a level so that investors know what they’re going to get and we’re not messing around with it.

[00:25:22] Alan

I wanted to touch on speed. I saw a paper on your own website recently which goes into what you’ve alluded to a little bit. You talk about how returns in trend following degraded over time, but that was very specifically, in this paper, for short or fast trend following. Obviously, you could consider fast trend following. Some of the other guests have alluded to the attraction of having some faster systems from the perspective of those turning points.

What’s your perspective on how fast to apply trend following from those two perspectives of absolute return versus being a bit more reactive at those equity turning points?

[00:26:04] Simon

Well, the analysis that I think you are alluding to looks at the performance of trend following over time for different speeds, different lookbacks of trend following, and you’re absolutely right that the largest deterioration is seen in the faster trend strategies. This is what everyone was trading, it was what Winton was trading when it when it started in 1997 and many people traded after that as well for interesting reasons sometimes, and you definitely see a marked deterioration there.

With the slower strategies, like you say, you don’t see such a strong decline in performance. But on the other hand, they started out with a lower Sharpe ratio to begin with. So if you are always taking the speed that you believe has the highest Sharpe ratio, then you see that decline as a result of those two effects. In terms of what’s the right speed to choose at any given moment, there are different objectives. If you’re trying to purely maximize the Sharpe ratio, then what our analysis has shown is that you really want to be very slow.

But it’s not the case that in general, investors are trying to maximize the Sharpe ratio of their trend allocation in isolation. That’s particularly clear when you think about what people want from it in a situation like last year or in 2008. You don’t want to hang on to long equity positions for a really, really long time and lose money just because it’s a very slow strategy. So there is some benefit in being faster there and it ends up fitting the investor objectives better even if the overall Sharpe ratio is slightly reduced over time.

[00:27:51] Alan

Fair enough. Just moving on into the research process more generally, you give a sense of this at the outset in terms of the application of scientific techniques. Maybe can you give us a sense of how that research process starts and ends from the perspective of how might a new idea get researched, where would it come from in the first place? Is it from market observation, academic journals? What are you looking for in terms of assessing, testing that idea and bringing it as a potential strategy to the portfolio?

[00:28:29] Simon

Sure, yeah. It won’t surprise you to hear that we’re pretty open-minded about where ideas can come from. The one place we haven’t looked so much recently though is in academic journals actually. We just don’t think that’s the forefront of research into alpha sources. But this is why we’ve got a very large research department. It’s because we intend to do original research ourselves.

Often, it has come from surprising places. We have an office in China, and sometimes when they’ve looked at Chinese commodities, it’s turned out that an idea, that they’ve applied there and come up with themselves, actually works in the global context as well. So it’s often not what you’re expecting or looking at any given one moment.

[00:29:15] Alan

Interesting. Is that anything from observations around supply demand dynamics in individual markets to reassessing speeds? Is that anything at all? Can you give a bit more color as to specific improvements and strategies that have come in?

[00:29:44] Simon

Sure, yeah. There are two different types of research projects there that you mentioned. One is finding a new alpha source where you’ve really got to think of a new idea or a new data input that you could potentially be using. The other is you’ve got your existing alpha sources and you need to make sure you’re doing them as well as you can. That involves reviewing things like the speed of trend following or other aspects of the way you construct your trend signal, your volatility models, and so on.

Very often with these things, you learn from the market environment that you’re living through. Often, adjustments to our volatility models, for example, have been prompted by new types of events. The events in February ’18, for example, were not so bad for us because we had lived through a very similar event in February 2007 which actually resembled the 2018 behavior quite strongly with the rising equity markets on relatively low volatility followed by a sudden sharp reversal, which then although people often focus on equities, was replicated across the space of markets that CTAs happened to be trading.

So it looked like a crowding effect because it was a reversal. They hit CTA positions specifically. But we learned from that and we adjusted our models in 2007 and that protected us to some degree in 2018 when a very similar thing happened. So that’s one of the most important ways that we end up getting the idea of changing our models in a particular way.

[00:31:20] Alan

When you see a market event like that, or maybe if you look at say March 2020 (as another example) where, really fast trend actually did very well, breakout models did well. The kind of models which had struggled maybe in the previous decade did well in that period. So when you see something like that, how do you assess that? What would you need to see to say maybe something’s changing with markets, maybe we do need to learn something as opposed to saying, well, that was just one occurrence in the distribution of possible occurrences, so nothing to change?

[00:31:58] Simon

Well, that’s a great question. You did have that period in February and March 2020 where faster trend following temporarily did better than slower trend following and it’s not hard to understand why. Markets fell for exactly a month, pretty much, between that last week in February and the point where the Fed intervened towards the end of March. If you turned around your positions and were short two weeks in, then you spent two weeks making money as markets continued to fall.

If, on the other hand, you took a month to turn around your positions, then you lost on the way down and then you lost again on the way up when you were short and the markets were rallying. So that was a moment where faster trend following outperformed slow trend following. The question is, at that moment, has something changed? Is faster trend following going to continue to perform slower trend following? We didn’t see any evidence that that had taken place, that there had been any kind of change there. In fact, that has turned out to be right. Since then, slower trend following has once again done better than faster trend following.

The question that then comes up, when you follow this line of thought, is, okay, so let’s suppose that’s right that still slower trend following is the better bet in the long run, but occasionally, you get these moments where faster trend following does well. Is there anything you can do in the moment to be fast just then, say, something about the regime? Our point of view is that it’s not the role of a systematic manager to do that and you can sort of see why if you imagine yourself going through this. We all did go through this so it’s not too hard.

Imagine that you’re long equities and you’re a slow trend follower and now you’re two weeks into the market falling. So you’re still long, but you think to yourself, should I speed up for this regime? Should I speed up? I’ll be short now so if it continues for another two weeks, I’ll do well. Obviously, what happened in reality was that the Fed stepped in after two weeks.

But now think about yourself at that moment. You don’t know that that’s going to happen. So you’re two weeks into this fall, and here are some possibilities. One is that the Fed could step in now. Another is that the Fed could step in in two weeks. Another is that the Fed might not step in or they might do it in a very long time. In the first situation, if they step in now, then you’re making the wrong decision if you speed up because the markets will rally and you’ll now be short.

In the second situation, which is what actually happened, it would be a right decision. In the situation where they don’t step in, well, it’s probably all about the same, because eventually, everyone will catch on to the big downward trend. So, the point there is that you can see that this apparently technical statistical decision about which is the best speed at that moment amounts to just a call on what the Fed is going to do and when.

Hopefully, it’s clear that that’s not the right call that systematic managers should be making. There are managers who are experts in that, in predicting what the Fed will do. Go to a discretionary macro manager for that kind of decision-making. But it’s not what you should expect from a trend follower, in our view.

[00:35:02] Alan

Okay. That’s probably a good segue into the research topic which is around machine learning. I guess the hope with machine learning is that the machine can learn from different environments and assess what kind of regime and if markets change over time, adapt to that. Is that a reasonable summary of what machine learning hopes to be and how successful do you think it has and can be?

[00:35:31] Simon

I don’t know. I think different people have different hopes for it. It’s obviously a collection of quite fascinating techniques which have been applied all over the place in the world and so it’s extremely tempting to try and apply them here too, because after all, they rely on having lots of data as an input and they do apparently amazing things that are very difficult for humans to arrange themselves. So why would you not want to apply it?

The thing is that machine learning algorithms require a lot of data, a lot of training data, to train them and to make them function well. How much training data do you really have in this context? The answer actually depends on the speed at which you trade. If you are trading relatively slowly, then there is only a few examples of the kind of events you’re living through because you don’t care about what the markets do from one minute to the next, you care about what it does over the time horizon in which you can change your position. There usually isn’t enough data to make machine learning a useful tool.

On the other hand, when you look at much faster trading strategies, as we do separately in our research process, then it becomes a very useful collection of techniques because suddenly, you have much more data for these algorithms to work with. We put a paper out a while ago where we distinguished different styles of scientific analysis. In a way, you have observational sciences like astronomy, and then you have experimental sciences like particle physics.

The difference is that in observational sciences, you can’t arrange a new experiment. If you want to study black holes, you just have to work with the ones that are there that you can see. You can’t create a new one for your own purposes. Whereas in particle physics, you can create a new particle accelerator and you can repeat the experiment as many times as you’ve got money for and then get more and more data. Something like that is true in this kind of research too in financial markets.

If you’re not talking about what happens in slower timescales, how you deal with events like a market crash, well, you can’t just create a market crash on demand. You have to study the ones that happened and try and learn from them as best you can. But if you’re looking at something like a very fast execution algorithm or something that’s operating very quickly intraday, then you have lots of opportunity to try out different things, see how it works, and then iterate over and over again.

So different styles of analysis are appropriate in different cases, and machine learning is definitely appropriate for the faster intraday strategies where you have more data to work with, and it’s not appropriate (at least we haven’t found a way to make it so useful) for the slower types of strategies that we’ve been discussing.

[00:38:17] Niels

Yeah. I just wanted to stay with this topic a little bit because it’s fascinating. I wanted to ask you, Simon, something that I think we probably haven’t asked enough so far in our series, and that is, the parameter selections that you make across the markets you trade. Do you trade universal parameters, meaning the same for all markets, or do you have different parameters depending on the kinds of markets you trade?

My second question would be, how do you select parameters in general? Meaning, is there some kind of human element in it or is it completely automated? How often do you change them? I’m just curious about that side of things.

[00:38:59] Simon

Yeah. We’re fairly universal, but not completely. There are practical differences often between markets which you have to account for. Some are more expensive to trade, some have different cost structures. In some, the market impact has a bigger concern. In others, they have intrinsically wider spreads and that has a bigger component to costs and sometimes commissions are a bigger component of the costs. So this sometimes requires you to choose different parameters in different cases.

In terms of how you select the parameters, that’s a fascinating question because obviously, we’re a systematic firm. Almost all of our processes are highly automated and we try to automate everything we can. But there’s always this boundary of processes where you think should that be automated or should that be done by a human. Parameter selection is one of them. The overall leverage a given strategy takes is another one. Choosing to put in a new signal, that’s another thing.

There’s a temptation, as a quant, to say everything’s automated. I didn’t do anything. The whole thing is done by computer. That’s definitely the wrong thing to do. That’s an abdication of responsibility over the investment management process. You can sort of see that because if you don’t decide what the parameters are, then something else is going to. It’s going to be an algorithm for deciding the parameters.

Is there only one algorithm for designing parameters? No. There’s an infinite number of possible algorithms for designing parameters. Which one are you going to use? Well, you’ve got to decide that. Are you going to decide that or are you going to have an algorithm to decide the algorithm which decides the parameters? How far back does this go? How far back should it go? The answer is not entirely obvious.

The question can be rephrased as, at what level do you want humans to take responsibility? Do you want to say I’m responsible for choosing this parameter or do you want to say I’m responsible for choosing this algorithm which chose that parameter? Now, which is the right answer depends a bit on precisely what we’re talking about. Sometimes, it’s a good idea to use an optimizer which does some of the other selection for you, and sometimes, it isn’t.

It depends on what impact that’s going to then have on the portfolio and how much it’s going to impact your ability to understand what the portfolio is doing. You obviously don’t want to have a black box where you delegate everything and you have a hard time understanding why the system is taking the choices it’s taking.

Our point of view has generally been that we try to push responsibility on the human fairly early. So things like the trend following speed, things like the allocation to individual assets, we say are human decisions. They’re made with input from analysis with the usual scientific rigor, but they are things that humans sign off on.

[00:41:53] Niels

Interesting. Actually fascinating because I think it also gives insight to why trend followers are not the same, right? We all have our own little interpretations of what is the right approach to some of these seemingly maybe, to some, small parts of a bigger puzzle, but actually, even the small parts can be quite important. So I appreciate that great insight.

Another thing that came around last year that I’d love to hear your thoughts about a little bit is CTA replication. From my memory, I don’t think you run any CTA replication type strategies, but a lot of people do. Some people for internal purposes, other people as their flagship program, so to speak. So I wanted to just hear your thoughts about, in particular, the replication type that has come out recently which is not a trend model but rather just (I say just but it’s a kind of) a regression analysis of returns of managers, because I want to better understand what the risks might be.

I’m not a quant so I can’t really judge that, but I’d love to hear different opinions about what the risks might be so that at least investors can be informed about these choices. Because clearly, you could say, yeah, there’s always going to be a manager risk, we can get things wrong, and occasionally, some of us do. But there must also, in my view, be some kind of replicator risk somewhere. So I’d love to hear your thoughts about this?

[00:43:33] Simon

Well, I know that they try to put out the view that they mitigate single manager risk. I don’t really agree with that. I think they are a single manager. If you want to mitigate a single manager risk, then obviously, the fund of funds approach is the way to go. That obviously has its own pros and cons. So it’s not that that’s a universal good, but it has its role. I haven’t looked too closely at the ETF you’re talking about though. How much do they charge? Is it very cheap?

[00:44:07] Niels

It’s about just shy of 1%. It’s a little bit different depending on where you look. Sometimes it says 85 basis points, sometimes it says another 10 basis points for dividend payments, so then maximum 95. So let’s call it just shy of 1% as a charge, and I guess some managers out there offer trend even cheaper than that, which of course, I’m not necessarily in favor of giving our great products and strategies away at very low cost because I think it’s not as simple as people may think it is if you give it away cheaply, but that’s a different discussion. So yeah.

[00:44:44] Simon

I’m sorry to tell you we offer our major markets trend strategy with daily liquidity for 80 basis points.

[00:44:51] Niels

There you go.

[00:44:52] Simon

It’s not clear then what the advantage is of a copy from a regression when you can invest in the original.

[00:44:57] Niels

Exactly, yeah, sure. Anyways, one other thing that I’d love to hear before Alan jumps in and that is just I guess you are in the camp of believing that more markets are better. Is that a fair conclusion? If it is, I’ve often sort of looked at different managers where I kind of know that some managers trade the classical liquid markets, maybe 50, 60 of them, and then all the managers clearly we know are out there trading hundreds of markets. You could be one of them. I don’t really know, honestly.

But I don’t really see a big difference in long term performance. I see it performs different at different times, but I don’t really see a huge difference in long term performance as such. What are your observations and why do we need to trade 200 or 300 markets, for example?

[00:45:55] Simon

Well, we’re definitely in the camp that more diversification, more opportunity to capture trends is better. I don’t really agree with you that there’s been no difference in the performance. When you look at the performance in the alternative markets over the last decade, it’s outperformed major markets by a long way. Last year was really the first year that that wasn’t substantially true. That decade after the financial crisis where trend following didn’t do so well, that was not shared by the alternative markets portfolio. They did extremely well in that period.

I think one of your other guests pointed to the fact that the Sharpe ratio increases like the square root of the number of markets you trade as evidence that the marginal benefit of adding another market is not very big. The reason that’s not right is that the number that goes into that square root is not the number of markets you trade, but the effective number of markets when you take into account all of the correlations between them.

Obviously, if you trade 100 markets that are identical, you haven’t really got 100 markets, you’ve got one market. The truth is with the 50 or 60 markets that you’re talking about, there are very large correlations between them. Most of the equity indices are very, very highly correlated with one another even if they’re in geographically different regions. The same goes for the bond futures, many of the commodities and so on. So there aren’t really 50 or 60 independent markets in that group. There might be 8 to 10 depending on how exactly you calculate it.

So the marginal benefit from adding more, of which are genuinely different, is really very great. In that way mathematically, you can see it more straightforwardly if you just look at what happened in the last couple of years. Last year was great everywhere but the year before that was okay for major markets. Look at the gigantic trends in the European gas markets and the European power markets.

These are things that there was no way to access. UK, natural gas (where we’re sitting) went up by a factor of 10, which I think is the largest trend I’ve ever seen in a futures market in that space of time. So there definitely are huge diversification benefits and huge trend opportunities from adding this particular collection of markets.

[00:48:06] Niels

Yeah. I appreciate that and I think it’s fine, absolutely, to pinpoint these differences in that. I do agree with you, there certainly have been a couple of products out there that are more focused on the alternative markets that for a period of time did better. So no doubt about that. But I just want to for clarification purposes say when I was talking about a 50, 60 market portfolio, I was actually talking about the classical markets where you have a lot of diversification because it includes a lot of the commodities.

But then if you go from 100 to 200, except if you go into really alternative Chinese, etc., type of markets, a lot of the additional markets you see often are, as you say, financial markets that may have correlations to other markets that you may already have. So I also appreciate that.

[00:48:58] Simon

We agree on that point, I think. Certainly, this is one of the points of disagreement we have with some of the other people who also trade a lot of markets, which is that the markets that you add have to genuinely be alternative and additive. It’s easy to get fixated on the number and try and win the competition of who has the most number of markets. But like you say, you don’t want to do that by adding things like, I don’t know, equity sector ETFs or developed market interest rate swaps, which are really very highly correlated.

They’re either super easy to access or they’re very highly correlated with things that are vanilla and that you already trade. So our effort there is to focus on things that are genuinely diversifying. You mentioned Chinese futures, we trade a lot of them. We’ve traded them for many years onshore in China and we trade them now via the internationalized markets, via swaps and via the QFI route for our global investors.

So that’s definitely one area where you see a lot of diversification benefit and we’ve seen a lot of trends. But like I mentioned also, the regional gas markets, the regional power markets, freight markets, marine fuel, all of these things have been tremendously additive over the last few years.

[00:50:11] Niels

Yeah, absolutely. Just on that topic, final question, because I think since definitely people who have been in those markets very early on, our industry came from the roots at least that we trade on liquid and exchange traded markets. I imagine that some of the alternative markets are not on exchange, per se, and you have to access them in a different way.

How do you balance the choice of maybe an additional benefit from performance, but maybe you could say (I don’t know what the word I’m looking for right now is), maybe the drawback of not being on exchange and taking on counterparty risk and other things. How do you look at that when you decide to add more markets? How important are these things?

[00:51:03] Simon

That’s a great question. I think there are several parts to the answer. The first is that although it’s true that most of the liquidity of these markets is not on exchange, most of the trading is not on exchange, there’s still markets where you can clear your position on the exchange. So you can really reduce the counterparty risk close to the level with major futures markets, which is a big point. You can’t always do that, but very often, you can. But there’s another way of phrasing or another aspect here you could have asked, which is that these markets, by definition, have a barrier to entry.

That’s one of the reasons why they have extra alpha because it hasn’t been so easy for speculative trend followers to just extend, automatically, their strategies into them. So it places a burden on the organization and you have to structure your trading team and your middle office team and so on to absorb these without requiring a huge scaling off of the number of people there. This is a big effort in software and in automation and it’s something that we’ve paid a lot of attention to. It’s work, but it’s also the reason why it often works, so it’s worth doing.

[00:52:21] Niels

Yeah, those are great points. I appreciate that. Alan, where do you want to go next?

[00:52:26] Alan

Maybe just to talk about the role of these strategies in the context of a broader portfolio. Obviously, a lot of what we’ve heard about talking to Simon and also other managers is the difficulty of keeping people invested in these types of strategies. So that’s why smoothing can be beneficial. From the perspective of having an alternatives allocation within a multi-asset portfolio, be it a 60/40 or something else, would that be more trend focused?

Would that make sense to be more trend focused for that allocation? Would it be pure trend? What about considerations that you see with other managers that it might make sense to, say, cap the equity beta of that trend program when it’s definitely been used as part of a multi-asset portfolio?

[00:53:21] Simon

Those are great questions in the sense that they’re the key questions for investors, in a way, rather than managers, because the answers depend on what the investor’s objectives are, what their tolerance is for drawdowns in particular parts of their book, what their process is for reassessing their existing allocations. If you can say I know that trend following has these characteristics, these drawdowns, and your process allows you to withstand those through time, then a pure trend allocation is really good. People who’ve made pure trend allocations and held on to them for the long run have done really well.

But like I said before, that’s not always possible and it’s not always possible for entirely sympathetic reasons. It just depends on the nature of the organization so it’s very hard to give a hard and fast rule and say trend following deserves this or that allocation in a portfolio. You have to have those individual conversations.

[00:54:20] Alan

What we’ve heard from some other managers is as you say, that there’s a recognition of both these trade-offs, but increasingly, the investor has kind of being asked to choose themselves and managers are in some ways becoming more like solutions providers. Is that the approach you’ve taken? Do you customize portfolios specifically for clients based on requirements or is it that you have a discussion and then decide what’s the mix of your existing programs?

[00:54:50] Simon

We do customize programs. People have all kinds of different requirements. It’s often less about the amount of trend than about the amount of alternative markets for example, or the amount of the Chinese markets and so on. Obviously, having multiple products makes it very straightforward to do this because people can just allocate some to each product and then it doesn’t require you to add whole new bespoke elements to the setup.

[00:55:23] Alan

Fair enough. You mentioned how a lot of these markets are now, particularly in the alternatives, that you can kind of clear centrally, mitigate the counterparty risk. So maybe just to talk a little bit about the big risks that you worry about. Obviously, you’re monitoring the risk profile of the program. You already touched on the leverage, etc. But what would you say are the qualitative risks that you think are to the fore of your mind when you’re running these types of portfolios?

[00:55:56] Simon

The way we think about risk management perhaps is a bit different from others as well. There’s a strange confusion that afflicts the CTA industry particularly, which is to equate risk with volatility. To the extent that very often, when a manager is asked, “What is your risk level?” They will just reply with their volatility level. It can start to feel natural almost. But when you step outside of this for a moment, it’s totally bizarre, right?

If you think about 2017 when equities were going up in a straight line and the VIX was at a record low, was there anyone in the world who thought that equity risk was at a record low? No one seriously thought that. If you talked to any equity investor and said, “Why don’t you think risk is low? Look how low the VIX is,” they would have pointed you to the fact that the VIX can increase very suddenly. They would have also pointed to valuations and they would have said valuations are looking extremely rich.

So what does that mean? Why do we end up in this segment of the investing world focusing on vol? It’s a strange story. It probably comes from the fact that actually estimating the risk, the risk of a loss, which is what everyone cares about, is incredibly hard. Whereas, the problem of estimating your current volatility is relatively easy. So we try and avoid that mistake. How do we do it? Well, you look at what equity investors were saying and they were saying valuations were rich.

What they’re really saying is, okay, I can look at the VIX, but then there are all these other numbers I can calculate which tell me that actually, risk is quite high right now. So you have to treat volatility in the same way. You treat volatility as one number that tells you something about risk, but it’s only one and there are many, many other things you should be looking at. You already mentioned leverage as another thing but it’s quite important to look at different scenario tests and different stress tests.

Now, this comes back to your question, well, which scenarios should you be looking at? That’s another way to ask, what are you worried about right now? You don’t want to restrict yourself to just things that you happen to be able to think of at the moment. That’s not a very systematic approach to it.

So what we often do is we look at our current positions and we see how they would have performed in all the previous market environments that we have data for. We say, what’s the worst day or the worst week that we would have experienced with our current exposures? Then we say, well, that’s something that did happen, and so by definition, it is realistic. It can’t happen again. How would we feel if we lived through that? Are we taking an excessive amount of risk when you then put it into that context? Often, this brings up things that you’re not necessarily thinking about.

For example, in 2014 when oil was declining, this was saying, what happened in 1990 when the first Gulf War started? Oil jumped by 30%. Don’t think that that couldn’t happen right now. It could happen. What would happen to your portfolio if it did? So how are things looking right now? Well, right now, most things are short bonds and long equities.

So obviously, the situation that’s really worrisome is a very classic risk-off move where bonds rally and equities decline at the same time. That’s not the behavior that we’ve been seeing recently from equities and bonds. We’ve been seeing modest positive correlation between those asset classes. Obviously, the dynamic with inflation has made that somewhat different from the usual situation where people just buy bonds as a risk off asset.

But nevertheless, it’s not impossible that that could happen. There could be some news which is both disinflationary or negative for the economy, and then you might see exactly that. You might see a bond rally and equities fall and that’s exactly what would be worse for the current portfolio. So that’s the kind of scenario that we worry about now.

[00:59:45] Alan

Okay. Those kinds of scenarios then, presumably, does that cause you maybe to reduce leverage relative to what you would have otherwise? Is that the point?

[01:00:00] Simon

Yeah, it can do. The way that these strategies operate, the leverage naturally fluctuates as the signal strength varies and the correlation between markets varies. The question you have for yourself is, do you want to maintain it in a particular band or are you happy at the moment with it being below that or perhaps above that? That then depends on what these other risk measures are telling you.

[01:00:19] Niels

Can I jump in with a non-scientific question here, Simon? Because what you’re saying makes perfect sense, of course. But I’m thinking, couldn’t you always think of a scenario that would go against all of your positions in the major asset classes at the same time? How do you deal with that in the sense that yeah, it could either force bonds to go down and equities to go up, or vice versa, or move in the same direction? So how do we balance that kind of risk that we are being so cautious, because it could happen?

[01:00:53] Simon

That’s right. We’re not just sort of imagining a scenario that could go against the portfolio, we’re looking at scenarios that have happened in the past and seeing what’s the worst case that we would have experienced with the current portfolio? You’re obviously right. If you start off from the point of view of what’s the worst that could happen, the answer is always going to be that everything goes to zero and you lose everything.

That’s just not very helpful because you need the scenarios you’re looking at to be realistic in some way. By definition, anything that has happened is realistic because it did happen.

[01:01:26] Niels

True. Here’s another non-scientific question. I completely agree with that in a sense that history is a good guide. But I’ve also heard very smart people say what we really should worry about are things that hasn’t happened yet. So I’m wondering, do you also have like a scenario for things that haven’t happened yet but could be quite meaningful in terms of either up or down for that matter in terms of performance?

[01:01:51] Simon

Yeah. Look, we have other scenarios which are more stylized in nature. You look at things like what if all the correlations change in whatever way would be worse for you, and that was actually a very helpful scenario in 2013. Because prior to the taper tantrum, trend followers when long bonds and long stocks, correlation was negative. So people who were (like most of the industry) fixated on volatility were very highly leveraged, and at the time, we were not.

The reason we were not was that we looked at precisely that scenario. What happens if that correlation goes positive and they both go down? That looked like it was an unacceptably large level of risk so we temporarily operated at a lower level of volatility and that really helped us actually in that year.

[01:02:34] Niels

Okay. Well, that’s great. That’s exactly what I was trying to get to. I wanted to ask you a couple of questions that actually were sent into us by a large allocator after having listened to some of the episodes. I’m not going to go through all of the questions but there were a couple of them that I thought were quite interesting.

First off, he asks if the ultimate objective of investing in trend following is to improve the compounding of your multi-asset portfolio while having tolerable drawdowns? He’s asking, should you then cap the trend follower’s ability to go long equities? I’m sure you can work out why that is. Do you think that’s sensible or not?

[01:03:16] Simon

Look, again, I think it depends on what the overall investor objectives are. So you can certainly do that. You reduce the long-term performance, to some extent, depending on how significant the cap is. Some people look at that and think that’s good. Other people look at that and think that’s missing out on an opportunity. In particular, people who are more concerned about the Sharpe ratio and the overall performance and not just how it performs in crisis periods might have that point of view. So we have offered that as a bespoke arrangement to investors that we’ll cap the beater. But it’s very hard to say one is objectively better or worse than the other.

[01:03:58] Niels

Sure. Yeah, I fully understand. The other question he has which I also thought was interesting for maybe a general audience of institutional investors, at least, and that is, how many trend following strategies do you really need in a portfolio? Have you any thoughts on that?

[01:04:16] Simon

Do you mean in our portfolio or do you mean in an allocator’s portfolio?

[01:04:19] Niels

No, I mean, actually, if an allocator, like an institution, was putting together some exposure to trend followers in general, how many managers should they pick?

[01:04:30] Simon

Well, if it’s Winton, then obviously only one, and in general, this comes back to the point of single manager risk and you can definitely see the point that there’s a trade-off, because you can diversify and you reduce the single manager risk and on the other hand, you don’t have a fee netting. So you potentially pay more in fees to do that. I’m sure Alan will have a strong point of view on this particular question. But to us, that trade-off is very clear and people just have to decide where they want to be on it.

[01:05:11] Niels

Alan, do you have a few last questions before I wrap up with the usual quick-fire ones?

[01:05:16] Alan

Maybe just one final one around return expectations. At the outset, Simon touched on how the Sharpe maybe historically has been 0.5 or so in trend following. We have spoken to other guests who’ve kind of done the research of the 100 years or the 200 years of trend following and have found that return to be very stable.

I’m just curious, in Simon’s conversations with investors when they’re constructing portfolios and trying to think about return expectations, is that the default way of saying, well, this is what it’s done over long term so that’s your best guess of what it might do going forward?

[01:05:59] Simon

Yeah, that’s probably right. Those long-term simulations are very interesting. I think the purpose they serve is to persuade people who are still thinking about efficient markets theory. So what they show is that the underlying markets have had this trending behavior over the very long term and they haven’t behaved efficiently. That is different from forming a Sharpe ratio expectation for a real traded strategy because that depends on the number of people who are also trading the strategy in the futures markets right now.

So even looking back to the ’70s, when you can sort of start to have a reasonable strategy including the currencies and so on, because prior to that, you have the Bretton Woods system, you see this change, that we talked about, in the performance over time and it’s clear that that correlates very strongly to the growth of people trading the strategy.

Like every other investment opportunity in the world, the more people who do it, the more they have to share out the returns among themselves. So 0.5 is our expectation based on all of the available evidence that we see, including what’s happened over the last 50 years as the industry has grown, as the performance of different speeds has changed where we think the best speed is to be right now.

[01:07:15] Alan

Okay. So that takes into account all of it. Obviously, I suppose you could say you’ve had lower trading costs now that you might have had in the past, which will be a positive, and fees have probably come down for investors too. So I guess there are lots of moving parts there. But no, that’s fair enough. Niels?

[01:07:35] Niels

Yeah. I just want to round off our conversation, Simon, with a couple of shorter questions not really related to specifically what you do at Winton. One of the questions that I always ask is, what is the one thing you hear about trend following from other people that you disagree with the most? We’ve had a few different answers, some similar answers, but I’d love to hear your thoughts on this question as well.

[01:08:04] Simon

Sure, yeah. I think the thing that we most disagree with is this idea that trend following is a kind of dumb strategy that is just people following each other in a herding crowd behavior. In fact, what’s really interesting is that very often, when trend following is most successful, it’s because it’s actually a contrarian strategy. What I mean by that is, maybe it’s easiest to look at a specific example.

If you think about oil in 2014 (the example I mentioned previously), when it was falling because of the rise of shale production in the US, and you look at what fundamental analysts were predicting for the price at the time, what you see is that no one was predicting the price would fall. Why was that? It was because the narrative at the time was around peak oil supply at the time, a term you don’t hear very often anymore, but that was very common then.

At the same time, demand for oil was growing from emerging markets, from China, from India, from Brazil, and so on. So with that combination of factors, it was felt that the price could never really go down because supply was dwindling, demand was increasing. What do you expect to happen? What you see is that no one really fully assessed the impact of the shale production increase.

The great thing about trend following is that it doesn’t know about these theories about what’s supposed to happen. It just says, well, the price has gone down, let’s bet that it will continue to go down. At that point, obviously, it’s acting as a contrarian because it’s betting against what everyone else thinks, and that was an extremely successful trade. Why was that successful? It’s because a trend is a situation where a price moves from one place to another place over an extended period of time.

If the price jumps immediately, then that’s not a trend. You might get right or you might get wrong, but it’s 50/50. For that to be a trend, the movement has to be extended in time, and often, it can be extended because it is contrarian, because there’s a narrative which people think means that the market needs to actually go in the other direction. A very similar thing was happening at the same time in a totally different area in German fixed income.

So there, what was happening was the yields were getting lower and lower and there was a widespread belief that the central bank could set negative interest rates if they wanted, that’s a bit weird, but they could do it. But who would give money to the government for 10 years and get less of it back? So there was this belief that you could never have long term yields be negative. Once again, trend following doesn’t know about that. It doesn’t care. It just says yields have been going down, let’s bet that they will continue to go down, and that long bunds trade was extremely good.

Obviously, what happened was those yields did go negative. Again, that contrarian attitude, because it was fighting against this narrative, it led to a very smooth trend. So very often, not always, but very often, one of the great things about trend following is that it’s actually a contrarian strategy. So that’s, I think, our point of view on trend, which is different from what we often hear.

[01:11:02] Niels

Yeah. I love that. Actually, that was a great point so I appreciate you sharing that. The final question I just wanted to ask is really kind of, we are at an interesting time, point in time. What are you most excited about when you look a little bit ahead? I don’t mean performance of course, I just mean what are you most excited about in general, and also, is there anything that gives you any source of concern when it comes to our industry or our strategies type thing?

[01:11:33] Simon

Sure. I think it’s very well phrased. We can’t make predictions about short term performance. What we’re focused on is what matters and what matters is original research, to find new alpha sources and improve our existing ones, and on risk management. Those are exactly the two things you mentioned. So we’ve got a really exciting research department and research process and so there’s a huge amount going on all the time both in our trend, in our systematic macro, and across our multi strategy hedge fund with the long-short equities and the credit as well.

At the same time, our approach to risk management is the one I mentioned of hypervigilance and monitoring a whole array of possible scenarios and situations that could come up. So that’s what we think is important for 2023 and that’s what we’re going to continue to do.

[01:12:23] Niels

Great way to end our conversation, Simon. Thank you so much. It’s been fascinating. It’s been delightful. Thank you so much for being on the podcast today and sharing those thoughts and insights with us and we hope we could do this again sometime in the future.

To all of you listening today, I hope you were able to take something from today’s conversation onto your own investment journey. If you did, please share these episodes with your friends and colleagues. From Alan and me, thanks so much for listening. We look forward to being back with you on the next episode of Top Traders Unplugged as we continue our deep dive into the CTA industry. In the meantime, go check out the show notes for this episode and all the other resources you can find on our website, and not least, of course, take care of yourself and take care of each other.

[01:13:07] Ending

Thanks for listening to Top Traders Unplugged. If you feel you learned something of value from today’s episode, the best way to stay updated is to go on over to iTunes and subscribe to the show so that you’ll be sure to get all the new episodes as they’re released.

We have some amazing guests lined up for you and to ensure our show continues to grow, please leave us an honest rating and review on iTunes. It only takes a minute and it’s the best way to show us you love the podcast. We’ll see you next time on Top Traders Unplugged.