TTU123: Russell Korgaonkar, CIO at Man AHL

[00:00:00] Russell

We always think clearly as investors, ourselves, about what’s most likely to work going forward. If you take a very basic model from 20 years ago, is it likely to work as well going forward as models that have been refined, and have had new features added, and where you’ve changed your execution profile – all these different improvements? The answer is that it’s easy to replicate the past, it’s quite difficult to replicate relatively sophisticated strategies going forward.

[00:00:42] 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 product before you make investment decisions.

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

[00:01:35] Niels

Hey, everyone. Welcome back to another edition of Top Traders Unplugged, where today, Alan Dunne and I are joined by Russell Korgaonkar, who is the Chief Investment Officer at Man AHL, 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, Russell, thanks ever so much for joining us today. We really have been looking forward to this conversation. I hope you’re doing well. How was New Year’s for you? Quiet, I hope.

[00:02:02] Russell

New Year’s was sufficiently quiet. I was with the family. But yeah, it was the end to quite a hectic year so I’m glad we had a quiet New Year’s.

[00:02:17 Niels

Yeah, I can imagine. Of course, we are going to be diving into a lot of that. But because we do have quite a few different topics that we’re going to discuss, I would like to set the stage a little bit for our conversation so that the audience knows a little bit about Man and the background.

So without taking too long, do you mind sharing a few highlights about the types of strategies that you are focused on and also where the industry or the business, I would say, sits before we head into 2023?

[0:02:47] Russell

Yeah, sure. Man Group is a large alternative manager. We run many different types of funds, many different types of strategies, broadly speaking, split between systematic and discretionary. We don’t have too many things that do both. So AHL is in the systematic part of the business, and even within that, we run many different types of funds. Heritage is a trend follower, but we run quant equity, we run long-only risk-managed strategies.

The key thing about everything that we do is it’s systematic 100% and we’re active, so we trade a lot. The common things are we trade and we do it systematically.

[00:03:43 Niels

Yeah. That’s great. I appreciate that. Now, for our conversation today, both Alan and I have created a list of different topics. We won’t have time, probably, to get through all of them, but we’ll pick a few of them as we go along. So maybe Alan, I could ask you to kick things off with the first topic of today.

[00:04:05] Alan

Great, yeah. We could speak to Russell. You’ve touched on that everything is systematic, and you’re active, which gives us a sense of your philosophy around markets and trading. But if you were to describe that philosophy, what would you say that is in terms of what’s the belief that underpins everything about the trading strategies?

[00:04:27] Russell

Yeah. I started in this business 20 years ago. I couldn’t actually believe that there was an edge in trading at all. I always thought it was likely to disappear fairly quickly. That was 20 years ago. I’m still here so there must be something that’s working. I guess our belief within AHL is that across markets, there are small opportunities and small inefficiencies, and in order to capture them, you need to be able to move positions around to trade in an efficient way. Because obviously, trading is always going to be a drag on performance if you’re not doing it sensibly. Then you need to be somewhat sensible and pragmatic about where those opportunities exist and also when they stop existing.

A big part of being successful in the systematic space is not actually just knowing what to trade, it’s also knowing what not to trade. So, we spend a lot of our research process thinking about that, thinking about the drivers of return, thinking about where inefficiencies might exist. Then, when we find them, we want to exploit them as efficiently as possible and in as diversified a way as possible – so, in many different regions, many different markets as we can. We find that putting all of those things together – efficient trading systems, very well-diversified, you can then take those small edges that you find and turn it into something that makes an interesting return stream.

[00:06:18] Alan

You mentioned trend following as the heritage of Man AHL and how you were surprised 20 years ago that those types of strategies work and they seem to be still working pretty well, judging by the industry’s performance last year. What’s your perspective on trend following? Why do markets trend? Why does trend following work?

[00:06:41] Russell

Trend following, in some sense, is one of the most straightforward of strategies, but also in other ways, one of the most elegant. The reason I think it works in the first place, the reason why you tend to get autocorrelation in price movements is pretty much down to three reasons.

The first is that information doesn’t flow immediately. It takes a bit of time for information to diffuse into a marketplace in a sense. Markets, therefore, aren’t 100% efficient, they take time to absorb. That means that a piece of information that you might expect should move a price 5% doesn’t instantly move it 5%. It moves it maybe 3% and it takes time to propagate. So that’s number one.

Number two is that markets themselves and drivers of market returns tend to be autocorrelated. So if you think about the Fed and levels of interest rates, or inflation readings, or economic data, all of that tends to move fairly slowly. It doesn’t jump hugely from one place to another. It does sometimes, but most of the time, that tends to move in the same direction. So, the world, to a degree, and the drivers of markets are autocorrelated.

The third, and probably for me, the most compelling and the one that kind of stands the test of time is behavior. If any of us are trying to sell a property and property prices are going up, we’re keen to wait for a better price. If they’re going down, we’ll take the first bid. That’s just human nature. Human nature, actually, that goes back as far as we have data for prices. It goes back to ancient Sumeria and some of the ancient scrolls that have been discovered about wheat prices 5000 years ago, and they tended to autocorrelate. So I think that’s the main reason.

[00:09:04] Alan

Obviously, you guys run more than just trend following strategies, so it’s trend and non-trend strategies. Would you say it’s those same factors that are the source of inefficiency in markets or the source of opportunity for the non-trend strategies as well?

[00:09:22] Russell

Not necessarily. I think each strategy has its own inefficiency that it’s trying to exploit. Sometimes that’s trading inefficiency or liquidity inefficiency, which is less behavioral and more liquidity driven. If there’s more demand for a certain asset that the price moves don’t justify, then that price move tends to revert. It’s quite a different inefficiency.

Likewise, for many of the fundamental strategies that we would run where it’s more of an information inefficiency, if you like, there’s some piece of information that’s not fully reflected in the price. So I think it depends, but trend following, to my mind, is very much a behavioral phenomenon.

[00:10:10] Alan

Fair enough. So maybe just the last one on this topic. So, If I was to describe the Man AHL approach. It seems to be trend at its core, but trend plus other strategies. How would you see yourself relative to other peers in the industry? Obviously, the classic answer to this is you don’t know what everybody else is doing. But I mean, versus a pure trend follower or other types of strategies or managers, do you see yourselves as doing something particularly unique that you would highlight when you’re presenting yourself to investors?

[00:10:48] Russell

Yeah. I think it’s difficult in our industry to claim anything’s truly unique. There are a few things that I’ve come across in my time that I would hope are unique. There’s certainly plenty of IP that gets created in the teams at AHL that is novel IP. But thinking about trend, for example, in all trading systems when it comes down to it, what you’re trying to deliver is the best return that you can for a given risk target or risk profile.

Trend followers have a particular risk profile and risk target. What all of our work leads us to try to do is to maximize our expected return for that risk profile. That comes in many different forms. It’s the set of markets that you trade, the set of signals that you trade, how you group those signals together, how you execute various risk limits, etc. So, there are many things in parts of a system.

Our belief has always been that you can take basic trend following, which is a fairly low Sharpe strategy but with nice risk properties, and then you can improve the expected returns of that strategy without changing the risk properties; make it not a trend strategy or not having defensive properties and everything else. You can make it more digestible for investors, because as much as last year was a very good year for trend following and very welcome in the context of a bad year for markets, most investors don’t really have the stomach for years and years of really poor performance. That means that they need to be in place, they need to be there with capital invested for when markets are more conducive to trends.

[00:12:51] Niels

I want to just interject here with a small question. I know Alan has a few other things to follow up on. But you kind of mentioned that for you, yourself, it was a gratifying year in a year where other things didn’t do so well. It kind of takes me back to, I don’t know, 2009, 2010, to talk about crisis alpha when Katy Kaminski coined that term. Back then, I was quite excited about it because I thought, “Wow, here’s a term that probably most members of investment committees would understand if you came in and said we’re going to allocate some money to a crisis alpha strategy.”

But then I’m also thinking, well, I don’t really know any trend follower who specifically put anything into their models to make money in a crisis, per se. So I’m just curious what your thoughts are on this and if you do or don’t do anything specifically to try and have a certain performance payout, so to speak, in periods of equity stress.

[00:13:54] Russell

Yeah, it’s interesting that you say that. I would agree. I think trend followers going into 2008 didn’t consider, in particular, performance during negative equity markets. I think they thought here’s a system that tends to work, it tends to have nice return properties, and that was the focus. Then 2008 happened and probably in the aftermath of that, with performance not being so brilliant, actually, directly after 2008, then the focus is on if you don’t get quite so much return, at least you get the good performance in internal markets.

I think if you talk to most institutional investors, they care about both. Clearly, they care about the return side, but they also do care about the crisis alpha or genuinely diversify, because there are things or strategies that are diversifying in the sense that the average correlation is low. But that’s actually different from strategies that tend to do well when the market does really badly.

Trend following by its nature is one of those things because when the market does badly, it doesn’t… I mean, COVID was actually an exception. It was a very, very short-lived market sell-off. But most of the time, when you get market panics, it happens over the course of weeks or months. That creates a nice environment for trend followers so they tend to do well. I think it’s one of those things that was never the original intention but, as time has gone on, and perhaps what people value has evolved and changed, it has become part of the lexicon.

[00:15:46] Alan

Yeah. It’s interesting. It does get to this topic that came up in a recent paper from Cliff Asness where he said there was a dual mandate effectively, absolute return plus crisis alpha. It sounds like you’re suggesting, maybe, that’s not strictly part of the mandate but more of a feature of the historic performance of trend following.

You also touched on how the evolution of your systems has been towards making trend following more digestible for investors to ensure people can stick with it and then participate in those periods. But I guess, is there a bit of a trade-off in terms of getting that bit more digestibility of the strategy? Are you giving up a little bit, then, some of that either crisis alpha or some of the convexity or some of those ( you mentioned maintaining) risk properties, broadly speaking? But is there a trade-off in those two objectives?

[00:16:50] Russell

I don’t think so. I think that the key point about the crisis alpha is that if you run a very, very slow trend following system ( so if I did something that said if the price is higher than it was 12 months ago, I’ll buy; and if it was lower than it was 12 months ago, I’ll sell), and you look at the properties of a system like that, it actually doesn’t have particularly brilliant tail properties in a crisis, because it can take quite some time to switch around. It might have okay properties, in terms of return, and it will certainly have good properties, in terms of diversification, but in terms of crisis protection it tends not to perform quite so well in equity-stress environments.

In order to have a system that does better, you need it to be a bit more reactive. That’s the key point about crisis alpha. If you look at different types of funds over different periods in the past, actually, that can often separate style. So most of the time (same for trend followers), most of the time they’ll look fairly correlated in periods in which the market moves fairly quickly. Actually, they can be positioned different ways around.

So things like COVID is a good example or Q4 of 2018, when we went from a low rate to an attempted higher rate environment, and then they gave up on that – those periods where you have got markets that are really changing. In the last year, we’ve had markets changing in the sense that you had volatility coming into rates and rates going up, and that caused all kinds of knockout effects, obviously, because of inflation. But that then led to very long, protracted trends.

Actually, if you were a slow trend follower last year, it’s the best thing you could have done. But the crisis alpha bit can be different because there’s different types of crises. That’s the most important thing. Typically speaking, the more reactive you are, the more likely you are to deliver a positive return in a crisis.

[00:19:10] Alan

Fair enough. That’s absolutely true. I guess the other typical classic strategy that people might add to trend, in a multi-strat kind of portfolio (that Cliff Asness alluded to) would be something like carry, which would tend to be uncorrelated. Carry is the classic kind of opposite return profile of trend where you have consistent returns and then the periodic blow-ups. By combining carry type strategies, would that be blunting that kind crisis alpha and that convexity, I guess, you might call it?

[00:19:52] Russell

Yeah, I think it depends how you construct it. So if you just had a carry return stream and you had a trend return stream, and you said I’m going to blend them 50/50, then clearly the properties of that combination are going to be that you might expect a slightly higher return, possibly, for average risk, but you may expect not quite so good return during equity stress because it takes you longer to build into those defensive positions or maybe you pick up some of that negative skew that you get from FX carry, for example.

One way you can try to get the best of both worlds is saying, most of the time when markets aren’t trending, they aren’t in a particularly stressed state, then perhaps carry makes a bit more sense. That’s the kind of time that you’d like to be invested in carry. But if markets get into a stressed state, and they enter either negative trends or they enter a higher volatility environment, then trend becomes a dominant feature. So that’s not exactly the same as adding the two things. It’s like saying, “Carry is okay except when trends are very negative, in which case, I should have zero carry and focus on the negative trend.”

[00:21:12] Alan

Yeah, maybe one final one on this topic is do you philosophically leave it to investors to choose? I think a lot of the larger firms now have a solutions business to see from an investor, do you want a pure trend, do you want trend plus, do you want trend plus other strategies? Is that the way it’s gone? Do you guide investors as to what’s the appropriate program or strategy for their circumstances?

[00:21:46] Russell

Yeah, it’s definitely gone that way a lot more in the last few years I think for a couple of reasons. Number one, I think the offerings have become more diverse. You’ve got core trend, and then you’ve got alternative trend, and then you’ve perhaps got some trend plus or multi strat, etc., and you’ve got all these various blends. We’ve certainly noticed that over the last few years, more investors like to have a bit more discretion in terms of which elements they allocate to and what features are most important to them.

So it’s certainly gone in that direction. That’s not the case for all investors. You’ll get some who will say, “What’s the highest Sharpe ratio you can give me?” and that’s fine. But it’s definitely, for the larger institutions, become more customized than it used to be.

[00:22:47] Niels

Before we jump to the next topic, which I’m quite interested in, I do want to just follow up on something you just mentioned there. You mentioned the Sharpe ratio and I’m curious on your thoughts as to whether you think our industry has become too concerned about Sharpe? Is that actually a limiting factor for us in order to deliver what we think is the best way we can deliver things like trend following?

[00:23:18] Russell

That’s a good question. I think Sharpe can be a fairly crude measure at times, especially for something like trend where, what do you care about? The volatility or do you care about some other downside feature? So whether it’s return over downside deviation or it’s return over some average measure of drawdown or it’s some other measure that’s correlated to the tails, I do think that matters. We do, when we’re having those conversations with institutions, we do try to run through that way of thinking.

Don’t just think about the Sharpe ratio. It’s okay. It’s a useful measure, all else equal. But it’s useful to think about those other return properties as well. I agree with you. If you think just too much about Sharpe ratio, in a naive world, if you were given something that had a Sharpe ratio of 2 and something else that had a Sharpe ratio of 0.8, let’s say, I would always take the 2, wouldn’t I? Not necessarily, actually, especially not when the 2 can come with a nasty tail that you’ve not yet seen.

[00:24:31] Niels

Absolutely. I appreciate that. Now, next thing I wanted to talk to you a little bit about is something that has become quite a big business in 2022, but it’s not new as such. It’s been around for a while. Lots of people, or some people have tried, some people have failed. There seems to maybe be one group out there, right now, that seems to be doing a good job, but we don’t have that much track yet.

What I’m talking about is trend replication, or maybe CTA replication where essentially the argument is we can do it for a lot lower cost and therefore you’re going to be outperforming just because of the savings of those costs. How do you feel about replication in general? Maybe I’ll give you the second follow up question even now, and that is if you have any concerns or if you see any weaknesses with replication in our space.

[00:25:36] Russell

You guys may have remembered the time during the 2000s when there was this big kind of phenomenon called hedge fund replication. What people did is that they took returns of hedge funds and they regressed them against some simple metrics. And they said, “I can replicate the hedge funds with these simple beaters and I’ll give you the same thing for a lot lower fees.” It’s really easy to replicate the past. It’s really very easy to make things that look good in backtests, that explain variance in backtests.

But going forward, we always think clearly as investors, ourselves, what’s most likely to work going forward? If you take a very basic model from 20 years ago, is it likely to work as well going forward as models that have been refined and they’ve had new features added and where you’ve changed your execution profile – all these different improvements? The answer (we’ve done studies on this for many years, and hopefully, I would like to think proved it through trading as well) is that it’s easy to replicate the past, it’s quite difficult to replicate relatively sophisticated strategies going forward.

[00:27:02] Niels

Let me follow up on that, Russell, because I think what the people doing it would say, yeah, maybe we can’t replicate one particular firm, but when we take the SocGen CTA Index, for example, or now I think, actually, some people are coming out with some kind of AI version of the SocGen Trend Index, then it’s a different story. We can take the return streams and we can figure out what their likely exposure is, or a way to replicate their exposure through a smaller number of markets, and so on and so forth.

Of course, as I said, the data is still relatively short. But so far, it looks like they have done it. I have my personal concerns, but this is not about my concerns today. It’s about your concerns, if any. So I’m just curious whether you think that maybe that part of the business has also evolved and maybe it’s a different case when you do an index of trend followers or CTAs. I don’t know.

[00:28:06] Russell

I think it depends on what you’re trying to do and how you’re trying to do it. But certainly, utilizing some return information, creating a model to try to explain those returns, you’ll always do a good job of explaining the returns. You’ll always come up with a model that looks like it is explaining a lot of why those returns have been as they are. But I guess my point is, in order to make a better system, actually, you need to replicate the positions, and ideally, the position is slightly ahead of time, and be able to trade those positions very efficiently. That is something that I think that it’s easy to overlook the importance of that, the complexity and the importance of it.

We’ve all been around for years in which these things come into fashion and they go out of fashion. But the key from my perspective, the key message, I suppose, to investors, is to think about your expected returns going forward. What makes you convinced that whatever strategy or firm or set-up or platform, whatever it happens to be that you’re utilizing to produce returns, why do you have conviction that’s going to work going forward?

So I think the whole setup of saying I’m going to try and figure out what others are doing, try to replicate it and then implement it in a slightly simplistic fashion, I think it would be difficult for me to have conviction in that going forward.

[00:29:54] Niels

Alan, now we’re heading over to something you love to talk about.

[00:29:59] Alan

I guess we’ve kind of touched on this a little bit. It’s the whole research. You mentioned how models that were developed 20 years ago might not work so well now and how all of the work you’ve done on various models has helped improve things.

Against that, there are some trend followers I’m aware of in the industry who’ve had fairly stable systems, what they might call classic trend followers, who seem to have done very well. The question is, how important have those improvements been? If you run systems that had been run 20 years ago now, on the current market (say like obviously last year), it was a very good market for trend following. How much behind would they be [compared to] the performance of the more up-to-date models? Can you quantify the benefit of the research?

[00:30:59] Russell

Yeah. We have a process here that tries to focus on that specific question. There’s always a challenge with systematic investing in that you have a model that works okay, and then you have a group of researchers. They do lots of research and they come up with some backtests and some theories. Always, those backtests make the system look better in the past, they always made the system look better. But that’s very different from saying, do they make the trading system better in terms of live performance?

So part of our entire process at AHL is to focus on that question: what’s the value of changes been? If you apply that to any single change, it’s very noisy. Let’s say you make a change where you add a new predictor or you change the speed of trading or something like this, it’s very noisy as to the impact that has over one year. But if you’re doing hundreds of changes, making hundreds of improvements, then you start to get a sense of which of them are useful, which are not so useful, etc.

What we’ve found over the years is that tweaking a trend following system, for example, doesn’t make much difference. You can fiddle with the speeds here and there, and you can fiddle with the market weights, but it doesn’t tend to make a huge difference. Adding markets that are genuinely diversified, that makes a big difference. Adding models that are genuinely different somehow, that also makes a difference. So it’s a big feature of what we’ve done.

I would say last year, by the way, the more basic the trend following system, the better. The thing that held on to being short bonds for the whole year and was a bit slower and a bit more basic and didn’t have so much risk controls or risk management in there would have done spectacularly well. But in the fullness of time, in the different types of periods, the research and improving things, for us, has definitely made the difference.

[00:33:17] Alan

You touched on tinkering with speeds a bit. As far as I am aware, if you went back to the ’80s and I think the ’90s, fast trend following did better back then and there’s been a degradation over time and a suggestion that the industry has shifted to be a bit slower over time.

If you notice something like that, that there seems to be a degradation, what would it take for you to significantly change the profile of the overall system in terms of its speed? Would you need some intuition around that something is changing in the market; that the market microstructure might be changing to shift things or would you just be driven by the data?

[00:34:09] Russell

I think often what happens is, number one, you need good analytics for these types of things. You need to understand how those patterns have evolved through time. The second is if you have an intuition as to why, then it really, hugely helps in making decisions. So in that example of fast trading stopping working, or fast trend following to a point, that’s fairly intuitive. As markets get a bit more efficient, you would expect (if you go back to your question at the start about what drives trend following, and I said one of the things was information leakage) that there’s a lot less information leakage through time. Markets have become more efficient and, therefore, you might think that particular inefficiency drives short-term trend following much more than it does long-term trend following.

So, we might reasonably expect that as markets become more efficient (which is not all markets, by the way, it’s some markets), that those types of signals tend to perform less well and you would expect that it’s not going to come back. It’s something that was there when markets were less efficient; they’ve become more efficient; and they’re not going to become less efficient again.

So I think intuition helps. You need to be careful about being too confident in whatever assertions you make because markets always tend to have a way of surprising people. But I think those are the two things that we do: good data, good analytics, and then some intuition as to why something’s happening.

[00:35:53] Alan

Fair enough. For a lot of managed futures managers, in the last number of years, it seems that execution has been a big area of focus. That seems to almost become a source of alpha. Has that been the same for you guys? With some medium to long-term systems, I’ve heard this suggestion for managers in the past that it doesn’t really matter if we trade today or tomorrow. From a medium-term trend following perspective, it’s a bit of a wash. Would you agree with that? Why is execution and being so timely so important for these types of systems?

[00:36:32] Russell

I think what execution allows you to do, what really good execution allows you to do is harvest a richer set of alphas, and that’s a really important thing. So if I was running a really basic, slow trend following system, I don’t think that my execution quality matters that much. You’re someone who said you could do it today or tomorrow, it doesn’t make too much of a difference. I think that’s probably a reasonable statement.

But what a really good execution system allows you to do is introduce many different types of alphas that you can trade around the edges and still retain the overall risk properties of your system but improve the expected return.

An example of that is we’ve got a team here at AHL who are running short-term trading models. (Short-term for us is hours to days. It’s not super high frequency but it’s not super slow.) They were making good P&L, and were very diversified from trend following with a very different set of signals. [They had good P&L, but it was kind of capacity limited. There’s a certain amount that you can put into those types of systems as a standalone, and they were operating as a standalone team.

If we use some of those alphas, some of those skills that they have, and use it to adapt the trade profile of slower systems, then actually what we can do is we don’t change the amount that we trade at all, we just change (slightly) the timing of the way that we trade. That means that we can exploit a lot more of that alpha in a bigger scale. To answer your question, that’s what it gives you. Really good execution gives you more ability to enable different sources of alpha.

[00:38:20] Niels

There’s a couple of things, and you’ve already kind of alluded to them. So, maybe for me, it’s more of a little bit of a statement. You mentioned this thing that it’s much better to trade more markets, especially when you can find diversified ones. I think that makes logical sense. However, I’ve often said on the podcast that when I look at our industry, I don’t really see any strong evidence that the firms that at least I know that trade hundreds of markets, that they are performing better. Maybe they can manage more money. That’s fair. That’s also valuable, of course. We’ll leave that at that.

However, one thing I’ve heard you say in the past, which I would love for you to repeat here if you remember it, and that is you had this great explanation where you talked about why you think it’s better to have dynamic position sizing or dynamic position management compared to static because this has been a big conversation between some of my co-hosts here on the podcast.

I don’t think there is a right or wrong, per se, but I’m always curious to hear different views as to why people feel one way or the other. I don’t know if you remember, Russell, what you talked about, but it had something to do I think with the fact that volatility changes, risk changes, and so on and so forth.

[00:39:58] Russell

Yeah. I don’t. There’s a danger here that I’m completely contradicting the past. Hopefully, I give the same answer. But that bit in markets is hugely important in our understanding of them and probably in our entire philosophy of how we approach trading systems. Returns in markets are fairly difficult to predict. There’s a fairly small edge. But risk in markets is much easier to predict and it’s much more time-varying and it’s intuitively time-varying. There are so many examples of that that we’ll have lived through but perhaps some of the listeners haven’t.

If you remember credit markets back in 2007, you would have been trading for several years and not really seen anything in the way of volatility; a very calm market, maybe credit spreads moving a couple of basis points a day. All of a sudden, in July 2007, some stress started coming into that market and two basis points turned into three and turned into five and turned to ten. All of a sudden, day after day after day, the order of magnitude of the move, so those are the moves of the order of magnitude bigger than they have been.

What position scaling allows you to do is to control risk. If you’re not careful, and you get the wrong side of an event like that, it can have a hugely damaging impact on returns. You might get the right side of it and you might be lucky. We would approach all of these things as that we want to maximize our chance of maximizing returns. In order to maximize our chances of maximizing returns, we need to make sure we’re still around and still functioning and still preserving our capital base.

[00:41:53] Niels

Okay. If I can just interject there, because the other side of the argument would be, well, that’s what our stop loss will do. That’s going to keep us in business. It has kept us in business for decades. By changing positions, there is a risk. I have my own reasons as to why I don’t think this applies but there’s the argument that the money you’re going to make on the outliers, that trend followers are sort of deep down hoping for and hunting for, is going to be smaller.

Now, again, I don’t want to put forward my view on this so I’d love to hear your view as to why you think that (even if you may, at times, be reducing position sizes), you’re still not going to lose out from having a static position into a trend like, whatever, Tesla Moderna, whatever (just to pick a few really big ones that we all remember).

[00:43:00] Russell

I think if you’re looking at a very diversified portfolio of positions, you could argue that if I’ve got good stop losses in place, I can control my risk profile that way. I’m sure you could do some empirical tests and illustrate that to be the case. I would say, though, that given that we know that risk changes through time. Intuitively, we know it, we feel it, we see it. There’s as much empirical evidence, as anything in financial markets, that risk predictably changes through time. Then you, in my view, want to be able to trade your system particularly in risky markets.

Stressful markets are the markets in which trend following systems tend to do best. So what position sizing allows you to do is stay in the game during stressful markets. If you just set a stop loss system, it might be that the market swings around and it knocks you out and then you lose your exposure for that period coming up. So I always say good risk management, or dynamic position scaling (whichever way you call it), just allows you to run your systems, whatever the market conditions.

[00:44:17] Alan

Changing gears a little bit, it’s related to this in terms of managing drawdowns, and I suppose that’s what you were getting at in terms of that position size management. But more generally, it’s a philosophical question you touched about at the outset of why markets trend and why trend following works. You obviously have these periods where trend following struggles and we have these long drawdowns.

I suppose if you think about it, it’s not like that speed of dissemination of price has changed in those periods, and it’s not like people’s behavioral biases have gone away in those periods. So why does it seemingly stop working for periods or why is it performing less well in some periods?

[00:45:14] Russell

I think people’s first concern, and certainly, it was their concern for many years during the last decade, was everyone knows about this strategy now and it’s become quite crowded, and therefore, why should we expect it to work anymore? That’s a perfectly logical and understandable hypothesis for any system. If huge amounts of people are doing it, surely it becomes less efficient.

There’s one quite nice way you can test this which is to say split your trend up. So, take a market and let’s say I create a trend signal using moving average crossovers. I have some signal that tells me at some point, if the fast-moving average goes above the slow-moving average, I take a long position. Look at the price evolution of that market. At some point, it will reach a peak. The trend followers never exit at the peak. They need to see the price coming down again before they exit, and at some point, the fast goes back below the slow and you exit your position.

So you’ve got an entry point, you’ve got a maximum, and you’ve got an exit point. In order for trend to be profitable, the exit needs to be higher (in the case of going long) than the entry. But that’s quite a nice way of decomposing the return because then you can say how much is due to my trend and how much was due to the reversal? If you just do the thought experiment for a second, you say I’m just going to dump loads of trend followers into a system and I’m going to keep adding the impact of trend followers to a system, what tends to happen?

Well, it doesn’t make the trend any worse, actually. It tends to push the price up a bit more. If anything, it can kind of overextend the trend. But what certainly happens is they’re all exiting the position at the same time and it makes the reversal much worse. So that’s what happens in theory.

What happened during the 2010s in liquid market (when I say liquid, I mean G10, the biggest markets out there) wasn’t that the reversals got any worse, it was that the trends weren’t as strong. I would say, the reason for that (our hypothesis) is that you had very compressed markets during the 2010s because you essentially had interest rates flawed, you had central banks acting as a kind of backstop behind markets – you had them acting in concert. During this period of very, very loose monetary policy and coordinated central bank action, all of that served to compress volatility and make markets more correlated.

If you think about what’s good for trend, which is a positive convexity strategy on lots of markets, you want low correlation and you want things to move. When you take markets that are kind of less influenced by central banks, those continue to trend pretty well during the last decade. So given all the evidence, I would say that’s our explanation.

[00:48:42] Alan

Niels, that’s probably a good segue into that discussion around capacity and AUM.

[00:48:47] Niels

Yeah, absolutely. We’ll try and hit a few more topics before we wrap up for today. It is something that I think he’s talked about a fair bit. Do people end up with too much money in the strategies? So I’d love to hear your thoughts about capacity in general, but maybe also about fees. Obviously, we know that there’s been a lot of pressure on fees, not just in hedge funds, but certainly in CTAs. We’ve seen a lot of these flat-fee products come out.

I apologize because I don’t know exactly what fee structures you offer so I don’t know if I’m treading on any toes here. But what I’ve come across more recently, which is kind of interesting, is that with some of this pension fund money that came in big time during the last 10 years, fees were beaten down in the firms where they took the money. Now I see some evidence of the fact that some of that money is now being returned to these clients because the fees are simply not attractive enough because there’s more demand. Also, there is a capacity issue that firms know that you can’t continue forever to just add more AUM. How do you think about this? Is this something that’s an issue for you or just in general? I’m curious.

[00:50:18] Russell

My view on that is, first off, most strategies have a capacity limit. It’s different for different types of strategies. If I were to take alternative trend, as we call it, trend on alternative markets, that’s limited by how big a position you can take in various markets. Typically, how big a position you can take is a function of liquidity and how quickly you can exit that position. All of that, when you put it to constructing a portfolio, sets a limit on how much AUM you can run. The same is true of all strategies to some degree.

Now, trend following on the 30 or 50 most liquid futures out there is a huge capacity strategy. You can run a lot in that type of strategy. Pricing really comes down to supply and demand. If there are huge amounts of supply, and lots of people providing you the same thing, here you go, here’s exactly the same return stream and that’s what you’re buying, then clearly, if there’s lots of supply out there, then the price should reflect it, and it should come down. Where there’s limited supply and you’re offering something that’s valuable, either in the sense that it’s limited or valuable in the sense that it has better properties, then the price of that is going to be higher.

I tend to think, in all honesty, the marketplace kind of sorts itself out. We’re all going to believe in free markets, and that is not just true in terms of trading, it’s true in terms of investors as well. The products that we run are closed, which is a lot of what we do, demand a premium, as you would expect, and those where we can run really high levels of capacity and there’s a more competitive marketplace for it, we can run it at lower fees. So we will try and cover both.

[00:52:18] Niels

Okay. No, that’s fair. We only got another 15 minutes or so, Alan, so let’s see if we can hit the last few topics with Russell.

[00:52:26] Alan

Yeah. Well, maybe just talking about the role of managed futures and trend following in the context of a larger portfolio. If I was running an endowment or something like that, coming to think about how much to allocate to trend following, what’s the role. Maybe first of all, how should allocators be thinking about the expected returns for the strategies? Obviously, you can look back historically and say, since 2000 (whatever it is), the SocGen trend has a Sharpe of maybe 0.4 (it might be a bit higher after last year), but whatever it is, is that the starting point or is there any other way of thinking about expected returns for strategies like this?

[00:53:12] Russell

I think in terms of the fit to the portfolio, people need to do modeling of whole portfolios, and they need to think about tail scenarios as well as averages. The perhaps useful position which we’re now in is that after having years of real plain sailing for most static long-only portfolios, they’ve really appreciated the value of a more active approach and a more complimentary approach last year. So that’s something that certainly we would emphasize.

Think about properties during environments which are difficult for the rest of your portfolio, not just average correlation, but in particular, tail properties. Then in terms of expected return, we wrote something (I don’t think we ever published it, but we did it for some of our clients several years ago), kind of categorizing if you take a straightforward trend system, it has an expected Sharpe of about 0.4, if you add alternative markets and you change your execution and you had some different outcomes, you can retain those same overall risk properties but improve your expected return.

So I think where you want to get to in trend (and certainly where it’s been over the years and I think it should be going forward) is you should be able to add trend to your portfolio and it doesn’t detract from the expected return of your portfolio. It should at least keep it as good as it was before you added the trend and very much improve the tail.

[00:55:01] Alan

Okay. From my experience of speaking to other investors and presenting managed futures solutions, once people say “I like it,” the next questions are typically, “how much, and where do I fund it from? Is it from the equity side or from the fixed-income side?” Sometimes, there’s the additional question of, “can I time it?” How would you answer those questions?

[00:55:29] Russell

I think, in terms of where you take it from, trend should be a genuinely diversifying return stream to any static investment, and therefore, really, you should keep the bulk of your portfolio weights the same. Let’s say I had $100 and it was 60 equities and 40 bonds, we’ll just multiply both of those bits by 0.9 and give 10 to trend and that should improve the properties of a portfolio. Your second question?

[00:55:59] Alan

I said how much? I guess if you run an optimization; historically, bonds, equities and managed futures, it would tend to come out with a number that most probably institutional investors would find very high. So, is that still the right answer or how to think about sizing?

[00:56:23] Russell

I think it depends on the mandate of the investor, to be honest with you. If I was answering the question, on a personal basis, where I don’t have any benchmarks to worry about or anything else, then my percentage would be pretty high because I would hopefully think I could access some high expected return trends systems, and therefore, if the expected return of something is higher than it is of stocks, then you’ll likely have a higher allocation than you would to stocks.

I think for most investors, that’s totally unrealistic. They’re very aware of their benchmark, and therefore, it becomes more in that 10%, 20% range. But we’re always going to answer them with more.

[00:57:05] Alan

In terms of (quickly) on timing, can you trend follow trend following? It’s a topic that always comes up periodically. What are your thoughts on that?

[00:57:17] Russell

That I think is a difficult one to the extent that you need to be able to know when to change your mind on those things and how quickly you can change your mind. So I think there are certain metrics that might lead us to be a bit more confident in trend. The average forward-looking volatility tends to be better if it’s a bit higher. The average correlation between markets tends to do better if it’s a bit lower.

But having said that, you need to be able to react. In order for that approach to work and in order to be a bit more adaptive, number one, you need to have other things that you’re going to allocate to fairly quickly, and number two, you need to be able to make that change fairly quickly. Given that practically, for most people, that’s unrealistic, I would err on the side of saying, don’t bother timing it.

In particular, you have to be tolerant of periods of poor performance. It’s not a high Sharpe strategy, and therefore, you’ll go through periods of poor performance. Those bits that tend to do very, very well often follow some period of worse performance. I think most clients these days and most investors, they appreciate that and they understand that and they’ve been through the bad times as well as the good times. That leads to a slightly more long-term outlook.

[00:58:50] Niels

I’ve just got a few different types of questions that I wanted to touch on. One is something that hasn’t really been an issue for quite a few years, and that’s cash management, because there were no returns. In fact, we were doing our best to minimize the drag, I guess, for investors. Prior to that period, of course, a lot of people just left all the money we were having from the funds in the bank. Then we realized, whoa, banks can go down so that’s not a good idea either.

Now we don’t live in a zero-interest rate environment anymore. So I’m just curious, what do you do with the cash that sits in the various funds you manage but that you don’t need for margin purposes?

[00:59:34] Russell

We have a cash management team at Man Group, and it’s a central team that runs across all of the different divisions. We don’t try to do anything heroic with that cash. We don’t want that to be a source of risk whatsoever. We just want to make sure that, in a risk-free sense, we’re getting the best yield that we can. So normally, it’s a combination of T bills and various kinds of very secure government debt.

I know that there are people who try and be a bit more risk-taking on that part of the book. But I just think, for us, that’s not where we should be, trying to take risks or make money.

[01:00:18] Niels

Yeah, absolutely. Now, another thing I wanted to ask you, more like a rapid-fire question maybe even, and that is, what’s the one thing you hear about trend following that you disagree with the most?

[01:00:32] Russell

That’s a good question. I disagree with the most?

[01:00:37] Niels

Yes. We didn’t want to make it easy for you of course.

[01:00:43] Russell

There’s sometimes a sense with people. The reality is at least the noise or what you hear or topics of conversation can be quite short-term. If trend is doing well, there’s not much complaining. If it’s doing badly, there’s much more complaining. I think one kind of problem or complaint/observation that we used to have is you take all of these big, directional positions, and they’re going to go wrong, you’re going to be the wrong way around on the market. Isn’t that going to blow up the portfolio-type thing?

The reality is, actually, if you were to hold those positions, let’s say we took the biggest set of exposures we happen to have, at one particular point, and held them throughout time, that would be a fairly disastrous thing to do. So, instead of thinking about point-in-time exposure (which is important and you need to manage those risks), you need to think about how the system evolves and is dynamic. So, it’s very, very different.

If I go through a period of performance with us being long equities, and the performance being correlated with equities, I think it’s that bit of saying you’ve just got equity beta. We had it for those three months. It doesn’t mean anything about the next six months.

[01:02:00] Niels

Yeah. I completely agree with that.

Now, I’ve just got two questions left. The other one I was thinking about, and that is, if you are on the other side of the table, an allocator looking in at one of your strategies, what would be, in your opinion, a question that they should be asking you? Maybe it’s something that during these due diligence conversations, you kind of get the sense that it’s never brought up, but actually, it’s quite an important question. Is there something that springs to mind?

[01:02:36] Russell

As I said earlier, I think it’s the investor’s job to build conviction that the system or strategy or group of people running the strategy are going to do a good job going forward. There are many parts of a system, many components that build up that conviction. So I think it’s a bit difficult to ask anyone.

But what I would want, as an investor, is to make sure that you’re running a systematic strategy. So in theory, a systematic strategy is built on a set of rules and it follows those sets of rules. I would want some reassurance that your process as you evolve, as you perhaps modify the strategy or improve the strategy, you’re confident in that process, the process is well built, and it’s well engineered, and it protects the integrity of the system as well as improving results going forward. I think that’s the bit that I would focus most on.

[01:03:48] Niels

Yeah, absolutely. Okay. Well, we are recording at the very, very beginning of 2023 so I just wanted to finish off with a simple question and that is, what are you most excited about or maybe what concerns do you have for the year ahead? You can choose whether it’s about system trading, global macro events, it’s all open field.

[01:04:14] Russell

Open forum. We started last year and we thought we’re in a really changing world and the environment is uncertain. We start this year, and it’s even more so. You look back at the things that happened last year and there were so many events.

Take nickel, for example. That was way back in March. We were doing a review of the year just a couple of days ago. Nickel, that huge event in the nickel market, so it was actually in 2022. We were all under work from home guidance in the UK in 2022. So, such a huge amount of things happened last year that, from a macro perspective, I think we’ve all learned to run at this really fast pace. The market seems to be moving so fast.

I’m kind of glad that we are in the systematic space. We have algorithms that try and kind of navigate it for us. The focus for us, as always but particularly now, is making sure technology platform is fit for purpose as you think about next-generation, and what asset classes you’re trying to add, and you make sure your research platform is doing the best job it can for all of the researchers. That’s our big thing for this year. So try and focus on what you can do to improve things and try as much as you can to block out all the noise.

[01:05:49] Niels

Yeah. I think that makes a lot of sense.

On that note, we’re going to wrap up this fascinating conversation. Russell, thank you so much for being on the podcast and for sharing your thoughts and insights with us. We hope that we can 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, and if you did, please share these episodes with your friends and colleagues and send us comments to let us know what topics you want us to bring up in the upcoming conversation with industry leaders in the world of finance and investing.

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 and check out the show notes for this episode and all the other resources you could find on the website. And not least, take care of yourself and take care of each other.

[01:06:40] 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.