TTU143: Guillaume Jamet, Principal Manager and Co-CIO at Metori Capital Management

[00:00:06] Guillaume

This is clearly an environment where lots of things might change in the coming months or years. Actually, I think that the benefit of a strategy like CTAs is that they’re going to adapt. They have no BS or constraints that force them away from those potential opportunities. That’s the great thing about our strategies. If you’re a long-only equity manager, there are situations where there’s nowhere to hide. On the opposite side, CTAs might be much more flexible and able to navigate.

[00:00:46] 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.

[00:01:42] Niels

Hey everyone and welcome to another edition of Top Traders Unplugged, where today, Alan Dunne and I are joined by Guillaume Jamet, Co-Founder and Co-CIO at Metori Capital Management 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, Guillaume, it’s great to have you on the podcast. Thanks ever so much for joining us today. We really have been looking forward to this conversation. How are things in Paris this morning?

[00:02:12] Guillaume

Fine. We have some protests actually, but that’s life in France as we discussed.

[00:02:16] Niels

Exactly. That is life in France. Now, before we dive into all the topics, Guillaume, that we’re going to be talking about today, I would love to set the stage a little bit for our conversation so that the audience knows a bit about your firm’s background. So perhaps you could share just a few highlights from its history, evolution, what type of strategies you focus on, and also where the business stands today.

[00:02:41] Guillaume

Okay. So clearly, we are a quantitative CTA manager based in Paris. Metori as a company started in ’16 as, I would say, a spinoff of the alternative fund platform of Lyxor. The program is a bit more older because it started in a subsidiary, the Epsilon Program, the program we manage, it started in ’94 in another subsidiary of SocGen. So it’s a long story.

Me and the co-founder, we started to be in charge or connected to the Epsilon Program as far as ’11, then a bit later, there was a spinoff. Today, we have a bit more than $800 million in AUM and a subsidiary also in China, so we’re close to 25 people and we are fully committed to this one strategy which is trend following.

[00:03:41] Niels

Which is great, that’s what we love to hear on this podcast, Guillaume. Now, as you probably know, today’s conversation is going to be a list of different topics that we often get asked and also that we find interesting, but we may not have time to discuss all of them. So what we plan to do is just to pick a few along the way and just take it from there really. As we normally do, Alan, why don’t you kick off with the first topic?

[00:04:10] Alan

Great. Good morning, Guillaume. Thanks for joining us. We normally like to start by getting a sense of the investment philosophy of the manager. How was it that you came to focus on a trend following strategy? What is your belief about markets and how they operate that brings you to focusing on trend following as a core strategy?

[00:04:35] Guillaume

Okay. So I can make a long or short answer because how I ended up being a CTA manager is a long story and it was not clear from the start that I would do that. I’m a pure mathematician, at first, more directed to an academic world for many reasons, some connected to money, I moved to finance around the 2000s and I was more in the derivative business; the complex models, stochastic calculus, and all those things. It was a very interesting period. I witnessed the GFC, by the way, from a nice point of view.

Later, around 2010, I moved to fund management at Lyxor and quantitative fund management, obviously, given my background, and in ’11, I had the opportunity to connect and be included into the Epsilon Program as the responsible head of research. At that time, Nicolas, our Co-Founder, was CIO of Lyxor, and we had the same (I would say) engineering, mathematical background, and we were very interested in strategies in general.

Maybe it’s something that’s obvious, I don’t know, when you come from the world of quantitative finance, interestingly, you’re not interested in the expected return. You’re only interested in understanding how risk changes, the correlations, forward volatilities and so on. Well, obviously, you know that people come to markets because of the expected return. So there is something very strange that in the derivative world, you don’t want to predict expected return. You don’t care because you hedge them actually.

So we came to the world of asset management with a stranger background and so we have a very fresh look, obviously. I think that with this scientific background, something like trend following makes the most sense because you need to have a risk premia. An expected return in front of a risk is a necessary condition for a market to exist. Nobody would take risks without expected return.

The next stage, when you’re mathematically inclined, is how should I measure that because I have the full thing in the market. I have the expected return, I have the trends plus the risk, and I have to kind of filter them out. It’s a very natural and scientific problem, actually. It’s called stochastic filtering or whatsoever. When you discover the market with this angle (and I know it’s not the majority of market participants that sees the world like that), you understand trend following as this sort of filtering problem.

[00:07:26] Alan

Okay, interesting. Many people would say trend following is a fairly simple idea. You’re basically buying price series that are going up or going down. Obviously, you come from more of a complex mathematical perspective. Do you think that is necessary, or is that just an additional perspective to bring to the risk management in particular, or also to the signal generation?

[00:07:56] Guillaume

I think the intuition behind trend following, you don’t need all the math to see that sometimes markets are consistently going into one direction. Maybe it’s even a model that says I need to last long enough so that when I see that trend, I will benefit from it lasting a bit more. The problem is that obviously, when you want… So, you can have that very intuitive or basic observation. The name of the game is going to measure things and deciding when it is a trend and how strong it is.

I have a father that is an architect, and obviously people have been building houses and bridges and stuff much before engineering and physics and they were quite okay. It doesn’t mean that I don’t think engineering adds a bit to the world of building but it’s as simple as that maybe.

[00:08:52] Alan

Yeah. Obviously, there’s plenty of sources of potential return, but have you focused more on the pure trend following or is the philosophy trend plus other strategies?

[00:09:05] Guillaume

Yeah. We are pure trend in the sense that yes, we don’t want to add other sorts of carry and stuff like that. This being said, the pure trend thing, I’m a bit uncomfortable because again, it adds to the idea that trend following is simple, and I think that doing it right is not simple. So I don’t like the word simple. It’s almost synonymous to cheap, and I truly think it’s not simple. But yes, no carry to be clear on that. But we believe also that taking into account correlation (which is an important ingredient of trend following) adds a loss of complexity and value if you do it right.

[00:09:49] Niels

Yeah. Obviously, you clearly come with a scientific background so you will probably dismiss the next term, Guillaume, that I’m going to bring up, and that’s the Sharpe ratio. We’ve referenced the paper that Cliff Asness and his colleagues wrote last year and it kind of raises a few questions such as what are the true mandates really of CTAs, etc.

But to make it very simple, what it comes down to (and this is probably something that you don’t think is a good idea in general for investors to do), but most investors look at the Sharpe ratio as some kind of measure of quality, even in the CTA world. But the real question comes (at least in our minds), with this thing about whose Sharpe ratio should we then be concerned about? Is it the Sharpe ratio of the managers or is it the Sharpe ratio of the overall portfolio of the client once you include these strategies? I’d love to hear your thoughts about that.

I’d love to hear maybe how you steer people to look at other things than the Sharpe even though I have to admit it is pretty difficult to get away from that specific ratio in our industry when people do some kind of analysis on the strategy.

[00:11:09] Guillaume

Yeah. Maybe I’ll surprise you on that because actually, I think that the Sharpe ratio is quite a good measure of a strategy. You can add other metrics which, would be important, but when you only have the track record of a firm, a strategy, I find it reasonable to start from the Sharpe ratio. Obviously, the Sharpe ratio is a good measure of return adjusted to risk. So the first thing is, are the risks well reflected into the volatility? I’m thinking about carry strategy.

In carry strategy, you have a problem. If you sell the wings (an option game where you sell far out of the money), for a long time you may have no volatility and collect premium. Obviously, you would have a wonderful Sharpe ratio. Here, you would say, well, it’s because the risk of this kind of activity is not reflected into the volatility. But for a directional strategy, using a linear instrument like trend following, it’s sad to say that you should not look at Sharpe ratio. I think you should.

The other interesting thing, where do the returns come from? So it’s more about the correlation with equity markets or some structural correlation with other markets, then yeah, it would be the question of beta and alpha, which is not an easy one. But if you have no structural correlation to the main risk of markets and you have a good Sharpe ratio, you are doing a good job.

[00:12:58] Niels

You’re right. You are surprising me a little bit here, Guillaume, which is great. So I’d love to explore this a little bit further, in the sense that some people that I speak to and I think they kind of have a point and that is, well, we all know, of course, that the Sharpe ratio kind of penalizes the upside volatility. So that’s always been a little bit of an issue.

But the other thing is, of course, that the Sharpe ratio doesn’t really say much about the risk of big drawdowns or the risk of a strategy blowing up, as we’ve seen so many of these high Sharpe ratio strategies do. So how do you adjust for that in embracing the Sharpe actually?

[00:13:42] Guillaume

If I understand well your question, the point about Sharpe ratio is supposed to be a measure return adjusted to risk and when you use the Sharpe ratio, you make the hypothesis that the past realized volatility is a good indication of the risk. Obviously, we know that there are very specific strategies, the carry strategy, where you are selling insurance, selling strategy, all those sorts of strategy, you’re going to collect the insurance premium very consistently. So there’s an almost linear return and you hope that a tail event or a more exceptional event does not occur.

Obviously, when you look at the track record, if the exceptional events have not been occurring into the one year, two year, three year period, and you believe that the risk is included in the volatility for this track record, you’re doing a big mistake. But I think that typically for a strategy like a CTA, I don’t see what you could do better. I understand that there are more subtle sorts of risk, but for a strategy using linear exposure to markets…

[00:14:54] Alan

If I could jump in, if you were to measure maybe the Sharpe of various trend following programs for the period say 2010 to 2019 (and that would have been a decade where you didn’t have say a major equity drawdown), and then you were to rank those trend following programs on the basis of their Sharpe, some of the managers or programs (the best in terms of Sharpe) may not then do that well once you go into maybe a 2020 type scenario or a 2022 type scenario.

So I think that was the idea behind the Cliff Asness paper that obviously Sharpe is a good measure of risk-adjusted return, but it doesn’t tell you about this second possible mandate of crisis alpha.

[00:15:39] Guillaume

Yeah. Okay. Thank you for reconnecting me to the initial question. There is this problem is with trend followers. You are talking about the crisis alpha thing. When I look at the trade of our strategy or peers, you see this crisis alpha thing that’s been occurring obviously, and it’s very present, and we understand why. It’s about convexity.

When you do trend following, you’re sort of implicitly (so it’s again maybe option background), you’re replicating a basket of straddle; some convex profile. So obviously, when things start to move quite consistently, quite strongly in some direction, you’re going to adapt and catch that.

So this is convexity, and that’s a great, really great property of CTAs. But maybe there is a type of crisis or situation where I would not bet that, in any case CTAs would procure. They are not a hedge. You can’t claim that you are uncorrelated and claim that you’re a hedge to something. That’s inconsistent.

[00:16:51] Alan

I think the general point is around, obviously, Sharpe. As you say, it’s capturing the risk-adjusted return. But then as you say, you’ve got the higher moments of the distribution, the skew, etc., and then obviously, whether you believe in crisis alpha or not, the characteristic of how well the program might deliver in a significant crisis period or equity drawdown.

[00:17:19] Guillaume

Yeah, but again, I think what we try to achieve, at least I speak for ourselves, have a CTA, a trend following that behaves as best as possible in all market regimes and not only waits for the next equity drawdown. I would only say that we don’t manage the fund in this perspective. We believe that there are many macro surprises, lots of different sources of risk in the market, which may have their trends and we try to catch all those and diversify as much as possible.

[00:17:55] Alan

It makes sense. We have spoken to other managers, on this podcast, who have emphasized more of defensive nature to trend following and others have balanced absolute return versus crisis alpha characteristics, and others like yourself who are more focused on the absolute return. Sorry, Niels.

[00:18:15] Niels

It’s fine. Obviously, there aren’t that many ratios as such. For example, and I don’t know if you’re familiar with this one, Guillaume, which is to Serenity ratio. The Serenity ratio also takes into account the actual drawdowns and the path dependency, and so on and so forth. It’s too complicated for most people to calculate anyway and it’s not something you would universally look at. But for say a CTA strategy, it’s kind of an interesting ratio. But let’s not dive in too much into that. Let’s jump into one of your favorite topics or wherever you want to go, Alan.

[00:18:52] Alan

Obviously, Guillaume, you mentioned you’ve been involved with the program since I think around 2010 or so, 2011. How has the program evolved over that period? I guess from a research perspective, what have been the things that you found where you can add value as a researcher in a trend following program?

[00:19:22] Guillaume

So over this period, quite long now, the program did evolve a lot. Actually, we made a major change in the evolution of the program that was in place at the end of ’12, where actually, we completely moved from a more maybe traditional approach where you had lots of signals and an optimization of a combination of signals which are, in fact, technically analyzed and coded into a computer, more or less, and then a very robust but maybe too basic equal volatility or equal risk budget allocation.

We moved to something different where detecting trends became statistical or you can say senior factoring of problems. My conviction was that correlation should be more taken into account both at the portfolio construction stage (which people will easily agree even if not so many people really account for correlation), but also at the trend detection part, because actually, if you believe that markets are connected, then why behave as if they were not at the first stage. So something more statistical and I would say joint multivariate for all of the investment process. This was the major change.

When you do these kinds of systematic counterintuitive things, there’s always a question that you should come back and improve. There are continuously questions that you could dig further, design choices that you have made that you may challenge or improve and so on. So it’s a continuous thing.

[00:21:10] Alan

Okay. You mentioned a couple of interesting things. When you say signal filtering, how would you explain that to the layman? Secondly, embedding correlation. I guess from that perspective, you’re trying to avoid making bets on what is effectively the same trend in multiple markets. Yeah. Okay.

[00:21:34] Guillaume

Yeah. So trends, you’re right, it’s a bit buzzword and not very clearly defined. So it’s expected Sharpe ratio. We want to think about expected return because in a CTA universe, the volatility of the markets which are so different from less than 1% volatility to maybe sometimes close to 100%, it’s obvious that when you want to look at this large investment universe, you’re going to factor the volatilities of markets quite systematically. So, signals mean expected Sharpe ratio. It’s as simple as that, and then you choose your weapon on how to predict or estimate.

[00:22:13] Alan

In that sense, are you saying that for every signal that you’re getting in each market, you’re assessing the expected risk adjusted return for that individual trade? Is that it?

[00:22:25] Guillaume

Absolutely.

[00:22:28] Alan

From the perspective of correlation and, as you say, not wanting to reflect the same broad trend and themes in multiple exposures, how do you go about doing that? People have different techniques for looking at that. Some people say you use principal component analysis or group markets or manage the risk by different sectors or by similar instruments. What’s your philosophy around that?

[00:22:59] Guillaume

The philosophy is to listen to market data first, avoid any predefined sector groups or predefined budget across (I don’t know) anything because you’re going to create bias in your strategy and your bias may come from past experience. So yeah, this is about estimation of correlation, and actually, it’s a difficult thing. So lots of people talk about it and it’s very intuitive.

But when you want to have a dynamic estimate of correlation, you have to accept that this estimate is going to be somehow noisy. But interestingly, people get afraid of estimating correlation, and in particular, short-term correlation. There is something very funny that everybody’s almost perfectly okay making a prediction on expected return and very afraid of making a prediction of correlation. But if I could earn my living by predicting correlation, I think I would be much richer than I am today.

So it’s very strange that people are very confident on predicting return and very shy on the use of correlation. The principles we put in place are really around challenging all those rules that were more, I would say, custom than scientific truths and check what is really true typically. You have heard that Markowitz is not working. This is something you hear everywhere. Actually, what we did is let’s look at how is it not working and maybe there are things we could fix. This is our approach.

[00:24:49] Alan

Interesting. You’re right. Correlation is something that maybe doesn’t get a lot of time and everybody knows you can basically plug in the correlation into the Excel spreadsheet to figure out what the correlation has been over a time period.

[00:25:05] Guillaume

It’s not going to work well if you do that.

[00:25:07] Alan

Yeah, so I guess there’s two things. One is, what’s a better technique? The second thing is, what’s the right timeframe?

[00:25:14] Guillaume

There is this thing, Eisenberg uncertainty principle, you heard of that, I think, which is that you want to estimate on instantaneous quantities that you don’t observe. So you would like to make the shortest timeframe. The problem is the shorter the timeframe, the higher the uncertainty and the noise around it. The trade-off is going to be with this observable quantity, how fast does it move?

If it moves slowly, then you can take a longer average to consider noise. If it moves fast, if you believe it’s moving fast, you have to accept the noise because you want to follow the changes in this fast-moving thing. So this is where, at some point, statistics won’t be able to answer all the answers. You will have to add an assessment of I think this is a fast-moving thing. I think so. I’m not directly answering your question. I see that volatility of correlation, the shorter the timeframe, the better at first, but then you will have lots of noise.

[00:26:25] Alan

Okay, that’s exactly it. I guess that’s the nature of correlation that it can be, even taking the bond equity correlation, for periods it will be negative and then suddenly switch to positive and then switch back again. So I guess that is the nature of correlation.

[00:26:40] Guillaume

That’s the true difficulty. That’s where the true difficulties will come when you do this simple trend following thing is when you really want to find the best way to estimate and then if you accept some noise in the estimate, and how to limit the impact and how to limit the overreaction and the noise and so on. That’s where you start to see that it’s not that simple.

[00:27:07] Alan

In terms of estimating volatility, is it the same philosophy of shorter lookbacks and being more market sensitive is better as well?

[00:27:15] Guillaume

Yeah, this is the approach for the volatility of correlation clearly, which is very challenging in terms of portfolio construction. That’s why people say that Markowitz is not working and that’s why they’re undertaking one year correlation. Obviously, when you do that, you may miss a lot of change of regime.

[00:27:39] Niels

Just staying a little bit on the strategy of the program, the design. Obviously, most strategies have to select parameters to go into the model. I’m curious whether this is also something you’ve managed to completely systematize in terms of selecting the parameters and also changing parameters along the way from a systematic point of view or is there some kind of decision-making process where you have to do it by committee?

[00:28:09] Guillaume

Okay. Thank you for this question because, yeah, I think that when you present yourself as systematic or quantitative, there is something unclear about it may be quantitatively assisted trading or quantitatively assisted portfolio management, which is fine. So you have lots of systematic and quantitative things. But on top of that, you have an expertise, and yeah, discretionary shots. I’ve met lots of great people in this industry doing that.

So here when we say systematic, we really mean systematic. The model we put in production, when we put it in production, this is because we believe that it is the best model with the best parameterization we know of. It would have been the best in the past and so it’s not part of day-to-day or even of recalibrating, changing things. That’s very important. The parameters are the really meta parameters. At some point, you have to make some design choice.

[00:29:17] Niels

Some universal truth that you believe in.

[00:29:19] Guillaume

Yes. I know it sounds very naïve, but it has more to do with trying to have a scientific methodology because at the opposite, if you have a model with parameters and your approach is that every month or every quarter, you recalibrate for data, you will recalibrate from investment committee, then actually, you cannot use backtest if you do that because you cannot backtest your investment committee.

If the automatic recalibration just from data, if you will always do it the same, then you should automate it. It should be part of the model, and then actually, you’re not changing things. But usually, you know that this recalibration, they are somehow discretionary. They are on demand and on filling and so on. Why not again? But then if you do that, you really cannot use a backtest and it’s going to be quite difficult to use historical data in some way.

[00:30:20] Niels

Yeah, absolutely. Now, you’ve mentioned a few times, which of course, we all agree with, and that is that CTAs and trend following, it’s not as easy as people may want to make it sound like. Of course, this has been an on-and-off debate on our side and that’s this thing about can CTAs be replicated. There are different approaches to that. Some people try and develop trend following models that they’re happy to make reasonably transparent if they want to put it in an ETF product, etc. Then there are people who will just analyze historical returns and try to replicate.

Since you have a much more mathematical background than I do, I’m just curious and also because although this episode will come out after my next conversation, my next conversation is actually going to be with a replicator. So do you have any views on replication, and maybe in particular, the type of replication where they do some kind of regression analysis of returns to try and estimate what exposure CTAs might have?

[00:31:32] Guillaume

Yeah. This is a very interesting and tricky question. I think you mentioned two very different directions, which is one oversimplifies and open sources CTAs in a way, and the other which is reverse engineering the existing manager and finding what they are doing. I think the problem has to do simply with a fiduciary duty. Because if on one hand you say I commit to things that everybody knows and it’s going to be very simple because I want to be very transparent on everything, it’s a limitation.

So maybe the argument is going to be less fees, but it’s a limitation. It’s a strong limitation because obviously, you don’t want to work hard on improving something and then give your portfolio almost every day or display the formulas and anything. No company would do that. So maybe some investors find their reasons on that. For the replication again, the problem with fiduciary duty is quite high because how do you feel responsible for the portfolio and the decisions you take when you are just trying to filter the returns of others?

[00:32:53] Niels

They’re valid points, they’re good points. Before I hand it back to Alan, let me then turn in a different direction, namely, let’s just say that people agree with all of us that it’s better to choose a manager. I have two questions for you really. One is, if you were on the other side of the table looking at managers, knowing what you know about design of systems and so on and so forth, what would be the things that you would look at in order to evaluate?

Because actually, I think our clients generally have a difficult job in deciding between managers, because in many ways, we kind of look the same unless you say, okay, well, he trades for 400 markets, he trades 50 markets, okay, that’s the difference. But the performance may not be that different, frankly. So looking from the inside with all the knowledge you have, Guillaume, how would you advise investors to evaluate different types of strategies within the CTA space or trend following space?

[00:34:06] Guillaume

I think that’s a very difficult job to select a manager. I don’t pretend to have the silver bullet to select a good CTA from the outside honestly. For a manager that tends to be quantitative and systematic, obviously, I would have more suggestions of questions to ask maybe. Yeah, typically your question of how frequently you reparametrize is in fact that is a good indication because I would not be very comfortable with that. Then maybe the track is not the result of the clever reparameterization or reallocation across a strategy and all those things, but just the fact that maybe it’s not so systematic as it pretends to be.

This is what we do as a company is to be as transparent as possible on providing data on the history of the fund because it’s obviously going to help investors understand where the returns come from. I think the question (we are investors with our own money by the way) is you don’t want to put money in things that you’re not sure you understand what they’re doing. It’s as simple as that. So we try to be as transparent as possible with investors on providing data, explaining what were the risks, where did the returns come from. I think it’s necessary to reassure yourself on where you are putting your money.

[00:35:51] Alan

I know you mentioned you have a subsidiary in China, so presumably, trading the Chinese markets is a key focus. A big topic and debate we’ve had here on the podcast is in terms of what’s the right number of markets to trade and market selection and is there a benefit of trading more and more markets, and particularly from the perspective of the alternative markets, which I guess the Chinese markets would be in that subset. So what’s your perspective on the benefit of creating those markets?

[00:36:29] Guillaume

We have a version of the strategy on Chinese markets which is for onshore Chinese investors, by the way, currently, which dates from ‘14 and which has had incredible returns. Also, from our research, obviously, those returns are very complementary to the Western trends and the Western returns. So the strategy is quite obvious.

We want to have a larger investment universe, Western market and Chinese market, and more generally, the first order is, the larger investment universe, the better, obviously. But then some difficulties come with this first order answer; with the fact that if all those markets are uncorrelated, then clearly, the larger the better and your Sharpe ratio is going to improve as square root of the number of markets. This is this rule of thumb. So it’s wonderful. But if you start to compute square root of 10, in fact, those markets are highly correlated and that’s why CTAs do not have a Sharpe ratio of 400% or 500% because markets are correlated.

So this is where, again, we have to dive into the details that you should have a larger investment universe, but not for the sake of having a larger number of investments, for the sake of getting a grasp on new risk factors, truly new risk factors. But obviously, if I add more and more European equity indices to my investment universe, it’s not going to change things much. We know that. The question is going to be the marginal benefit of adding a new market, which might be very small, and maybe sometimes the operational cost of adding this market because you start to go to non-deliverable currency, a non-listed investment. So you add some specific risk and difficulties.

The question is going to be the data quality maybe sometimes, at some point, is going to be an issue. So, in practice, we will not jump on any new markets because we are very doubtful of the true addition it will add.

[00:38:42] Alan

So basically, more markets can be better if they’re genuinely picking up new risk factors, and you think that’s the case in China.

[00:38:50] Guillaume

Yeah, and do not add specific risk or cost to the strategy. Liquidity is obviously one of the questions, because if want to have scalable returns, then there are markets where you’re going to have some issues on trading them significantly.

[00:39:15] Alan

So your flagship, now, is incorporating the Chinese exposures as well. Is that right?

[00:39:22] Guillaume

Not currently. We are close to starting a new flagship with the largest investment universe, including China. This is the big project for ’23.

[00:39:33] Alan

Okay, interesting. Obviously, you’re talking about risks. How do you think about the risk of trading in those markets where you may have geopolitical concerns or tail risks of markets shutting down or capital controls being brought in, etc.? Obviously, these are all kinds of risks that are not, as you say, captured by volatility and the distribution of returns.

[00:39:59] Guillaume

Yes. It’s difficult to make a generic answer because actually, it’s precisely depending on the type of market to understand maybe those specific risks that are not into volatility. If you look at the Western market, we had lots of great surprises recently. I think of some central banks from time to time, so we don’t need to go to China to find a surprise. Obviously, yes, the non-deliverable currency adds a layer of difficulty. We try to recognize all those risks and then I think that our model framework is nice enough so that we can put a price on those.

Typically, we don’t claim that being longer than China is completely economically equivalent to being longer in the US. If we just want to be longer, it would be better to be longer from the US than from China because we acknowledge that there are specific risks linked to Chinese positions. I don’t want to stress on China. It could be true for all those markets, counterparty risk as a general rule simply, or illiquidity risk as a general. You need to put that in your model so it does not freely choose the gold future or a sophisticated ETN in some strange currency. I don’t know.

[00:41:30] Alan

It sounds like very much more markets are better, but it’s very much on a case by case basis. That seems to be the philosophy. Beyond trading Chinese markets, do you see another source of opportunity beyond there over the next number of years, that you think might be interesting?

[00:41:49] Guillaume

When you look at the at the liquidity of the futures market, the biggest commodity markets are Chinese markets. There are few US markets that look big, but many of them are Chinese markets. So it’s just very natural to have a look there. We want to enlarge the investment universe but we want to be cautious at the same time in terms of the operational cost and liquidity risk.

[00:42:22] Niels

Just staying on China, it’s not a market that I have any experience with. But one of the things that we see more and more and that we get asked more and more, and especially to fill out certain questionnaires, is ESG. I have to say, I have personally struggled a little bit with where do you draw the line in terms of our industry, our strategy’s relation to ESG.

But frankly, to me, a country like China might be a little bit different in terms of ESG. So how do you think about that if you open it up to more investors in a new flagship strategy?

[00:43:19] Guillaume

ESG is clearly a big topic. I think that the main difficulty is that when you are a CTA, it’s very difficult to first build a measure of any ESG risk because you aggregate financials of all currencies, all types of financial equity bond rates, foreign exchange commodities and you need to have a measure. If you look at what exists currently in the ESG, you have scores on kinds of instruments; scores on equities, scores on bonds. But obviously, when you go short, it’s not clear how to use those scores if you go short.

So what we have done, in terms of ESG, is first how do you aggregate scores from different sources? How do you mix them? How do you account for the long or the shorts? We came to a methodology that we have been sharing with investors and academics and the feedback is quite good. We had the opportunity to talk, I think, in some conferences a few months or years ago, I don’t remember, and this is where we stand.

First trying to know what you’re think about; having a measure of the thing and obviously, at the end of the day, some investors… The UCITS fund we have does not include commodities, and for many European investors, it’s just like that. They prefer to stay away from commodities. Obviously, adding Chinese markets, you may have the same kind of investor mindset. You just draw a line, obviously.

[00:45:06] Niels

That’s fair. Funny enough, it’s a question that is super difficult, I think, for all of us to talk to. So I appreciate it because you probably would have done some work on this as you’ve indicated, so I really appreciate that.

Just another question that sort of springs to mind as see people’s reaction to what’s happened in the last year or so, in particular, where, as I mentioned in the introduction, CTAs, managed futures more broadly, was really the strategy that beat everything else last year.

But often, people still, I don’t think they truly embrace having these strategies as a core investment at all times. They probably think, well, if it did well, it’s had a good run, let’s just get out or let’s reduce and we’ll do something different. I know you are not necessarily on the client facing side of things but I’m sure you discuss it internally. What do you think are some of the hardest things and how do you think about overcoming that in the way you communicate with investors and potential investors in terms of getting people to embrace what we believe should be a core allocation at all times in a portfolio?

[00:46:25] Guillaume

To have investors getting in with the right expectation, actually. Because the investor getting in after a fantastic ‘22, there might be some disappointment because obviously, all years in a CTA… If you already claimed that on the long run, you will have a 75 Sharpe ratio, which would be great for a manager, I don’t speak about 100, I think it has to do with the statistics of the Sharpe ratio actually.

It’s difficult to estimate the Sharpe ratio. You need lots of time and you have to admit that 75% Sharpe ratio, which again is great Sharpe ratio when you look at the long-term numbers, all the years are not going to be positive and pleasant. It’s as simple as that, I would say. It has to do with, in fact, behavioral finance that we are risk averse, we are loss averse. We focus on the short term and we are fooled by randomness very frequently.

When you practice statistics, it’s kind of helpful because you can put some numbers on that. But when it’s money, it’s difficult because we all hate losing money. I think this is the beginning of the discussion with investors about appreciating that it’s hard to measure a Sharpe ratio. You should accept some volatility to get some return. Trying to time CTAs, I don’t know how to do it, because obviously, I would just put it in my model once and for all.

[00:48:01] Niels

It’s funny you say that because that’s exactly what we say to clients and potential clients. Don’t try to time it. We don’t know. Even us who look from the inside, we have no idea how to do it. Yet, that’s what they do at the end of the day, so it’s kind of an interesting dynamic.

[00:48:19] Guillaume

Yeah, it has to do with the nature of risk because if you believe that the risk of CTAs is the classic random noise, boring motion (sorry for this word), and convexity, then obviously, you have to accept that there are going to be boring periods with no return and volatility and a rally of performance from time to time. You can’t time that. You have to be on board already.

Actually, if you compare it to carry strategy, where you are going to have long periods of steady returns on sudden bursts, then maybe you could try to time that, time the moment to accept. So many people also are applying to one type of investment something that is just not the right tool. Maybe it’s difficult to time a volatile in a convex strategy.

[00:49:10] Niels

Yeah. I agree with that. I think you’re right. I think investors, including ourselves, of course, to some degree, we are conditioned by everything around us and there’s no doubt that timing investments is something that people spend a lot of time doing, and also the so-called experts on financial media will try to do it as well. So I appreciate that.

Alan, where do you want to go next? Maybe we have one or two rounds left before we wrap up.

[00:49:45] Alan

Absolutely. Yeah. So maybe just to delve into… We’ve talked about the client perspective. So thinking about the role of trend obviously in a broader, multi asset portfolio, it’d be good to get your perspective. The typical question when you speak with clients is how much to allocate and is it more funded out of the bond side or the equity side?

Obviously, if you run historical optimizations of bond equity trend following, you’ll tend to get a much higher allocation to trend following that you typically see in an investor’s portfolio. Do you think the allocation should be higher in line with those optimizations or not?

[00:50:34] Guillaume

Yeah, you can’t deny that if you have a strategy with a significant Sharpe ratio, on the long run, very in line with the Sharpe ratio of equities or bonds (I mean, on the long run, not only the last year), and close to an average zero correlation, I think it’s going to be difficult to find a reason not to put some significant share of that into the portfolio. But obviously, it’s not that easy. It’s not going to be an easy investment decision.

But still, we see many, I would say, highly sophisticated institutional investors that came to a conclusion not so far from that. I’m not an expert in the US market, but apparently, CTAs have a special position among all the sort of hedge or alternative strategies and they’re called also risk mitigation strategies. It says something on the fact that you have a class of investors that understand quite well their benefit in a portfolio.

[00:51:43] Alan

Yeah. I guess adding trend following or managed futures 101 is to say, yes, let’s have an allocation. You then take some of these institutions who say, well, we’re going to use trend within the context of a multi-asset portfolio. So is there something that we need to do to the trend following program to optimize it in that role from the perspective of should we cap the equity beta or should we look at the speed of it to give us more protection in drawdowns, etc. Any perspective on whether they are good ideas or not?

[00:52:17] Guillaume

Again, I think being an investor is a difficult job so I would not want to give advice and claim that I would be smarter than them. It sounds very natural. If I know that I’m going to put this CTA with my bonds and equities, maybe I should tweak them a bit. The problem when you do that also is that you’re going to confuse responsibility, because it’s a problem of adding overlays and delegating overlays. We want to manage a strategy that has to not make money when equity markets are down or whose quality depends on something that you don’t control. It’s going to be difficult to share responsibilities and benefits in this framework. But I understand the reason.

So if you believe on the long-term statistics and say, on average, correlation is going to be zero, Sharpe ratio should be good, risk is quite well managed, this is one of the great characteristic of CTAs. Maybe I’m wrong, but I’ve never heard of a CTA blow up or CTA drama. Correct me if I maybe I am not the oldest and the most experienced, compared to other types of strategies, it’s quite remarkable, this positive skewness or so that you don’t discover the risk when things go really bad.

[00:53:51] Alan

Okay. Maybe just quickly, you mentioned how these strategies have found a niche and a place, particularly in institutional portfolios among public pension plans in the US with this crisis risk offset label or crisis risk mitigation strategies. It does seem maybe in Europe, there’s obviously plenty of investors in trend following but maybe we don’t see that the same extent.

What’s your perspective on why there’s been less of a take-up maybe or less investing? Certainly, if you look across the wealth management space and multi-asset funds, you don’t see it well represented, I would say. I’m curious to get your perspectives on that.

[00:54:31] Guillaume

I don’t claim to have a global view on what investors are doing in US or in Europe. So it’s always quite difficult to answer this question. I think maybe there are still misunderstandings on what it is. The words ‘hedge fund’ sometime maybe you might say it’s maybe too funky, too exotic for my portfolio.

Well, actually, it’s a very liquid instrument and maybe it has to do with that. Compared to the most famous hedge fund strategy like long-short equity, maybe a too simple explanation where were you go long the good one and sell the bad one, and you take cars, you take computers like this one is the good, this one is the bad, there is a story.

People understand that if you’re good at detecting the best carmaker and detecting the worst car maker, you should do the obvious long-short. For CTAs, I know it’s more complex to explain. You’re using correlation, convexity and the first to do it, and so it takes more time to explain.

[00:55:35] Niels

It’s funny you mention these things, Guillaume, because actually, in some of the conversations I’ve had this morning here in Asia, some of those very topics came up and so I think it is interesting. But let’s just imagine people listen to this conversation, Guillaume, and finally, it makes it all clear for them, your explanation makes it all clear and we see a big inflow of money to trend following, which would be wonderful.

But from a mathematical point of view, from a research point of view, Guillaume, I think people like finding excuses as to why trend following shouldn’t work or should continue to work even though we have 50 years of proof, of real data, right? But is there such a thing as too much money in trend following in your mind? Because we know that the industry as a whole hasn’t really grown for a long, long time, like 20 years. The AUM is probably the same. The AUM is slightly inflated, I would say, because they include Bridgewater, which is not really a trend follower, in my opinion. But then, of course, we also know that there are a lot of people doing it probably internally and they’re not part of any index. So there is more money than what we see, I’m sure.

I don’t want to impose my opinion, but is there in your view, other than on a manager level where clearly capacity can be an issue at some point, but is there from an industry point of view any argument to say I’m concerned about all the money in trend following?

[00:57:15] Guillaume

I think there was an interesting view on that on the FT if I’m correct about CTAs are more minnows.

[00:57:26] Niels

Oh, yes.

[00:57:28] Guillaume

I was very happy to read that because it’s exactly my opinion, coming from just numbers, the number of managers, the AUM of the industry, the magnitude of the leverage. You to the size of the currency market, the size of the bonds or rates market, the size of CTAs presumably is ridiculously small in fact compared to the size of the market participants. I’m skeptical. I don’t know.

But I’m always skeptical when I see those comments that CTAs did move the markets. Actually ,for trend followers, we never move market. We come just afterwards. We’re not that smart to move market. Maybe sometimes, we overreact. But given our size, I really doubt it, but I know it’s a very popular explanation. I cannot find a single figure or statistic that could… All the numbers I see would say that it’s not.

[00:58:33] Niels

We’re far from any risk of that.

[00:58:36] Guillaume

I keep hearing that.

[00:58:38] Niels

Actually, interestingly enough, even though you said you haven’t been in the industry forever, but that’s the other thing that came up in my conversation today was actually this thing about, has there ever been a well-known CTA that’s blown up? Actually, I would agree with you. I can’t think of a single one as well. So I think there is something incredibly robust about what we do.

Now, just two things that we like to ask everyone, Guillaume, before we wrap up. I have a feeling what the first answer might be, but I’m going to throw it at you anyways. This thing about if there’s one thing that you hear people say about trend following, so not your opinion, but if there’s one thing you hear people say about trend following that you really disagree with, what would that be?

[00:59:26] Guillaume

There is not only one.

[00:59:29] Niels

You can give us two or three, that’s fine.

[00:59:31] Guillaume

The first one, I think the reason why I don’t like those things where it’s going to create misselling or wrong expectations. You prefer to have investors that understand what they are in, what they have invested in. So the first one is CTAs are long volatility. That’s plainly wrong. Actually, we churn on volatilities. We are long dispersion maybe or large moves.

I think this one is very toxic because then when you remember (maybe it’s over now) but during that period ‘10 to let’s say ’20, any time the equity market would spike down for a few days, you would have people say, “CTAs did not go up,” and, “That’s too bad.” There were lots of comments on that and it means that there was a complete misunderstanding of what to expect from our strategy. So this one, I think, is very toxic, that CTAs are long volatility.

Also about timing CTAs. This is not a helpful comment.

[01:00:42] Niels

Sure. Sure. You want to throw in a third one, Guillaume, since you’re on a roll here?

[01:00:47] Guillaume

No, I think it’s okay. Thank you.

[01:00:50] Niels

The one I was expecting from you was this thing that trend following is easy.

[01:00:55] Guillaume

Yeah, you’re right.

[01:00:58] Niels

Fair enough. we’re only in the early part of 2023. There’s already been lots of interesting things, exciting things happening. Looking at the rest of the year, I imagine that the launch of your new fund will be exciting for you. But is there anything in looking at it as a whole, and not necessarily just on what you do, but maybe at the industry level or maybe even from a completely different viewpoint, is there anything that excites you or is there anything that you’re concerned about when you look at what’s happening right now and how it impacts some of the things we do maybe as an industry?

[01:01:44] Guillaume

Clearly, since ‘20, we are witnessing so many events, change of regime and I think that we are really into that. It is carrying on, the debate on the inflation and rates and the impact of high rates, the continuation or not of hikes, many people think it’s over. So this is clearly an environment where lots of thing might change in the coming months or year.

Actually, I think that the benefits of a strategy like CTAs is they’re going to adapt. They have no bias or constraints that force them away from those potential opportunities. That’s the great thing about our strategies. If you are a long-only equity manager, there are situations where there is nowhere to hide. On the opposite, CTAs might be much more flexible and able to navigate.

[01:02:46] Niels

Yeah, I agree. I think nimbleness and flexibility will probably be in high demand as we work through this environment. Guillaume, this was really fantastic. Thank you ever so much for a truly fascinating conversation. Thanks for being on the podcast and sharing your thoughts and insights with us and we hope we can do this sometime in the future again.

To all of you listening today, I hope that 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. Not least, as usual, take care of yourself and take care of each other.

[01:03:43] 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.