TTU127: Ryan O'Grady, CEO & Co-Founder of ROW Asset Management

00:00:07] Ryan

Trend is more interesting than just beta, especially the way we think about it where the markets are made up of cycles of different frequencies. So, if you’re employing trend, really what you’re trying to capture is the meat part of those cycles. So, there’s some skill in understanding should I be short-term? Should I be long-term? Should I be medium-term? Should I adjust that over time? Why do I adjust that over time? It is definitely more interesting than just a very straightforward, “Here’s the beta, anybody could do this.”

[00:00:45] 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, and welcome to another edition of Top Traders Unplugged, where Alan Dunne and I are joined today by Ryan O’Grady, Co-Founder and CEO at ROW Asset 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, Ryan, it is wonderful to have you on the podcast. Thank you so much for joining us today. We really have been looking forward to our conversation. How are you doing? Hope you had a good start to the new year.

[00:02:08] Ryan  

Yeah, doing great and thanks to you both for having me. Excited to be part of the podcast. I’m a fan of the podcast. So yeah, good start to the year. Good end to the year, obviously. So again, thanks for having me.

[00:02:21] Niels  

Absolutely. It’s our pleasure. 

Now, before we dive into the different topics that we’re going to discuss today, maybe just to set the stage for the conversation, give a little bit of context for the audience, maybe you can just say a little bit about the background of your firm, and in particular, the kinds of strategies that you run and where the business stands as we enter into 2023.

[00:02:49] Ryan

Sure. I’m a quant macro lifer and that’s actually true of all of the… ROW originally was a group of people that had worked together for quite some time. ROW stands for Ryan, O’Grady, Weiser. My partner Jeff Weiser and I have worked together since 1995, and a lot of the core people at ROW have been together as a team and as a group since that time. I started in 1992 and other people in the team started around that time. So we’ve really been doing this our entire adult lives. 

ROW started in 2010 and our goal when we started the firm was to create a true all-weather return profile that had the divergent characteristics that one would expect from a CTA. We are a member of the SocGen CTA Index, which I believe has the 20 largest managers who are open. For us, it’s always been about having a divergent strategy, having something that adapts. So, we really only run one strategy which all of our best ideas flow into and then we have subsets off of that. 

But yeah, the idea is sort of when you think of the spectrum of active to passive, where if your position is wrong, how do you react? Do you just let it ride? Do you add to it? Do you react by reducing it? The role that we fit within the investment world is we’re on the reactive side. So the other side would be equity beta or private equity, things like that, that’ll hold positions for years, whereas we tend to react pretty quickly to changes in market environment. So, that ends up looking like trend following. 

We do correlate to trend followers. So that’s how we see ourselves. And then as far as business goes, we have great relationships with our clients. I think that’s really what makes this fun is that we created a return stream that we feel like is useful to people and it has worked out to be useful to people. So, we’re very happy with how that interaction has gone.

[00:05:16] Niels

Sure, absolutely. Now, we have created a long list of topics that we want to talk about. So we will essentially go back and forth between Alan and I to dive into some of these. As usual, Alan, maybe you can kick off the first topic for us.

[00:05:36] Alan  

Sure. 

Ryan, you touched on a few features of your approach with being all-weather, having this divergent characteristic, being reactive, etc. If you were to describe the investment philosophy which underpins your approach to markets and the reason that you believe that this is a good way to approach investing, what would you say about that investment philosophy?

[00:06:03] Ryan  

Beyond what I said, trend following is an approach where you are reacting in a pro-cyclical way. If your position is working well, you are either keeping it or adding to it, and if it’s working against you, you’re reducing it. So that is a feature that our products have. We do have a philosophy where we are agnostic to style. So we don’t need a model to be a trend following model for it to be useful to us. In fact, we do quite a bit of research on non-trend strategies. 

It depends just on how much weight you want to put on the different strategies, but no matter what the approach is, whether it’s a carry approach, earning a yield on a position, whether that’s backwardation contango or that’s an FX trade or an interest rate trade or something in the long term value arena or short term mean reversion, any approach where a set of systematic models is going to yield a durable output of profitable returns that you can depend on over time, we’re going to find a way to use that at some weight. 

The philosophy also is that we’re continually trying to think about new ways to add models to our system. To a certain extent, this is a constantly changing game. I know we have some competitors, they feel like, “Well, we’ve got something. It’s worked for a very long time and so this is the commitment we have to this approach so we’re going to stick with it.” Whereas we’re maybe a little different where we feel like the markets are constantly changing. So, it’s not so much that it’s changing and ruining stuff that we built. It’s changing and creating opportunities to come up with something new. 

[00:07:57] Alan  

It’s not so much the case that you have a fundamental belief in any particular model or a fundamental belief that trends will always be evident in markets and that will be the weights. It’s more belief that there will be opportunities to capture via research and that you can do that.

[00:08:19] Ryan  

Yeah, that’s exactly right.

[00:08:21] Alan  

Okay. In terms of the investment objective, you mentioned all-weather. So, you’re aiming then for a strategy that’s not just a diversifier. Again, some CTAs might come from the perspective of generating an uncorrelated return profile to equities. It sounds like your philosophy is you’re aiming to generate positive returns regardless of the market environment. Is that fair to say?

[00:08:50] Ryan  

Yeah. Our approach is not designed around a greater portfolio necessarily. It does work out that way. The first investors in this program were Jeff and I. We designed it to make money. Actually, we’ve had interesting conversations with our clients where we talk about, “Well, we underperformed here but whatever index did worse.” They even say to us, “Look, we really just count on you guys to make money.” We’ll certainly point out when the return stream, that we’re throwing off, works in a portfolio in a particular way. But yeah, the core goal is that our performance stands on its own.

[00:09:41] Alan 

Just delving into why there should be these opportunities in markets, why should a research process be able to unearth inefficiencies or anomalies? Is it market microstructure, behavioral biases? What’s your perspective on that?

[00:09:58] Ryan

I guess it’s all of it. This is a closed system of market participants and there are some people who play well or try to play well and we sort of fit in that category if we are in this system. By system I mean the markets that we trade in trying to earn a positive return. There are other people who are in this system for some other reason, for risk mitigation. Then there’s just some people who have behavioral biases or the greed/fear cycle. 

There are even people who run strategies where they’re forced to do a trade that they may not want to do, like some sort of risk parity system that when vol goes up, they’re forced to sell their positions whether they believe that’s a good idea or not. So, in that mix of participants, it’s really about trying to make the most good decisions versus trying to avoid mistakes. So, that’s on the active side, how someone who has a solid approach is able to succeed. 

There’s some participants out there who really are not interested. They’re not part of the market for profit. In the foreign exchange world, there are central banks who have no interest in profitable trading. Their interest is in smoothing monetary policy or something like that for their country. And so sometimes they’re doing things that they know are not necessarily profitable trades. They’re doing them for some other reason. So that’s what creates opportunities for firms like us.

[00:11:44] Niels

Yeah. I want to stay on some of these topics. You mentioned that you have this all-weather approach, but you also mentioned trend following, and so on and so forth. It kind of reminds me of an article that Cliff Asness wrote a few months ago. I guess he was taking it from a standpoint of trend following, but he was essentially saying that many of those who were identifying themselves as trend followers, a few years ago, changed along the way, adding more stuff to their strategies maybe as a response to fewer trends for a period of time, but maybe also (and I wanted to hear your perspective of this), maybe also because we are all being influenced to some extent, and maybe therefore, we’re all becoming too concerned about our Sharpe ratio, so to speak. 

What are your thoughts on that? Should we be concerned about our Sharpe? 

[00:12:54] Ryan

That paper was forwarded to me by a few of our clients, actually, and they said, “What do you think of this?” Because they know that ROW would be the opposite side of what that paper was arguing. So I guess the one difference between us and the foil of that paper is that we’ve always been committed to having a multi-strategy approach. But the other thing about the paper was it kind of implied that managed futures CTAs have a specific role and we don’t necessarily agree with that. The role of what we’re doing, that’s up to the purchaser, maybe they have a different idea of what it’s supposed to do. So that was part of it. 

I think the other part is that when we think about what is trend following, trend following is not long vol. I think that’s not necessarily what the paper was about. But I do think it’s important to remember, when you’re putting together a portfolio of strategies, and you feel like you need to stay committed to trend because trend is long vol and you need to throw off a long vol P&L. But trend following isn’t long vol. In fact, if you look at the daily skew of a trend following strategy, it’s negative. That’s been true for a very long time. It’s because on huge shock days, usually trend followers have the wrong position initially, and then they eventually get to the right position. But really, what trend followers are is that they’re long low frequency vol and short high frequency vol and they’re also short liquidity. If a trend follower could execute its entire strategy instantly for no cost, the profitability of the strategy would be much higher. 

So in those initial shock moments, going back to the paper, that’s kind of the implication of the paper is that this is why you’re in the trend following space is because you’re worried about these things. It’s not even necessarily trend that performs the best in the first moment or the first day or the first week of those sorts of events. 

So I think it’s better (and that’s obviously our philosophy) to think more holistically – think of, well, there’s low frequency vol, there’s high frequency vol and markets are cyclical. If you think, what is a trend following model? If you believe that trend is the right model to use to replicate behavior that we see, then you would expect prices to reach infinity or zero with certainty eventually. That isn’t actually what happens or at least not to the extent that a true trend following data generating process would imply. 

We think there’s long cycles where they’re long enough that trend following approaches can exploit them, and then there are short cycles. The nature of cyclicality leads to longer-term or even shorter-term value mean reversion-type approaches, and then in between, you have the ability to hold things and get paid for them. 

So to us, if you limit yourself to trend following, given the way that we see how markets behave (I compared it to a haiku), you’re sort of limiting yourself. You’re saying, “Well, I’m going to approach. I’m going to exploit this part of how things work and not the rest of it.”

[00:16:29] Niels  

Yeah. Interesting. Also, just staying on the paper a little bit because I think Cliff introduced this idea that trend following has a dual mandate. One was, I think, to make money, but also to make money in particular when equities were in a crisis. I certainly remember when the term crisis alpha got introduced after the financial crisis. I thought it was great. I’m not in the same camp anymore because I think we just end in a debate about what is a crisis, instead. 

I know that you sit maybe a little bit outside of that so maybe this is a question where you look in at the maybe more pure trend followers in our space. I’m curious about what you think about maybe investors’ expectations of what “we” (maybe I shouldn’t say “we” in this case), but trend followers, what investors should expect from a more pure trend following strategy, if we take a step aside from what you’re doing specifically. Clearly, trend is part of what you do.

[00:17:42] Ryan  

Yeah. I think, from the point of view of investors, pure trend and what we’re doing are pretty close. You see the differences more when you’re right down in it with another participant, but from the outside, our correlation is pretty high to pure trend. It is actually something you have to be careful about when you’re talking to people because you certainly don’t want to make any promises or guarantees of how you’re going to perform an inequity drawdown. 

I think the right way to think about it is to sort of work from the other direction and say let’s imagine a scenario where (I mean, this one’s going to be kind of self-serving because it’s just happened), but let’s imagine a scenario where both stocks and bonds go down 20%. What strategies could possibly make money in that environment? Whether the trend followers or the CTA community do or don’t, they could have or it certainly created a window of opportunity for them. 

So when you’re trying to put a portfolio together, certainly you did your job in including a divergent strategy even if sometimes the setup of it won’t work out the way you would have hoped. Equities in the past 12 months had been pretty difficult even though they were in a downtrend. If you look at the really strong performance of CTAs in the last 12 months, it had nothing to do with equity. It was all about interest rate moves and commodity moves. 

But I think it’s still correct to think of it more as, let’s imagine different scenarios, different possibilities and just think about who could do well in those types of scenarios and then who probably won’t do very well and then balance the approach that way.

[00:19:34] Niels  

I think that makes a lot of sense, Ryan. Actually, what’s interesting also is that even though, on an industry basis, no doubt, 2022 was an important year (was a great year), but it was also a year, actually, where there was a lot of dispersion in returns at the same time. So again, a little bit about how you implement your rules and risk management so to speak. 

Speaking of that, Alan, maybe it’s time to dive into one of your favorite topics.

[00:20:10] Alan

Yes, research. Ryan, you touched on how you started off in the early to mid ‘90s. You’ve been doing this ever since and you’ve emphasized how your view is that there are opportunities to be extracted from research over time. Where do you get the inspiration for research? Is it academic papers? Is it just observing the markets? And then how do you go about starting with an idea to bring it into a possible trading model in the portfolio?

[00:20:49] Ryan

There is definitely a connection to academia that’s important. Our research team is pretty small. That’s always been true for us even at the firm that we were all at together previously, the research team was never more than four or five people. The team here is the same thing – four or five people doing work together. So, academic research becomes a force multiplier where if we’re canvassing one or 200 papers a month, then you can jokingly say that those are unpaid assistants for us or interns or things like that. 

But I think the strength (or what I perceive to be our strength) is in being able to read a paper and have some intuition about that this person has figured something out versus this is just sort of overfit nonsense. I think there is a balance where we have interns come in and one of the first things that we give them is  (I fit the whole thing on like one piece of paper), it’s like here’s some moving average crossovers and here’s a really simple carry concept and here’s just a basic moving average (two standard deviations go the other way). We give that to them and within a week, they’ve produced a return stream that is pretty compelling. It’s not bad. It would be an average CTA when you put them all together. 

So that’s the simple side of the spectrum and then the super aggressive research would be like having a team of 100 people and all this AI technology and trying to create the perfect model. There are firms like that too. There are firms with really simple models, there are firms with incredibly complex approaches with huge research teams full of machine learning specialists and PhDs and things like that. We find ourselves very comfortably in the middle of that. 

The issue with the super complicated approaches is that maybe the intuition isn’t there. In our industry, there’s not really statistical certainty on whether a model works or not until it’s run for a pretty long time, and then when it’s run for a pretty long time, there’s no certainty that the conditions that lead to its success still apply. An example we give all the time is the GSCI roll trade where there was an inefficiency in the market created by a lot of passive money on a commodity index that if you had an AI program, it would have found it, and a lot of people in the middle of that range I described found it and profited from it. Then the inefficiency was fixed and it went away. 

So, that’s where intuition would come in and say they fixed it, the trade is not there anymore so we’re going to stop trying to exploit it. But I guess the struggle, the rigor of our approach is making sure we’re in the right spot where we’re doing as much work as we can do but staying within the realm that everything we’re doing makes intuitive sense.

[00:24:03] Alan

You’re basically highlighting the ongoing challenge of doing research and deciding what’s going to be persistent over time. What does that look like in terms of when you have your system and then presumably go and reassess lookbacks and things like that? What does it take to make changes to the system? Is it significant deviations in terms of performance with the new system relative to the old system coupled with some kind of intuitive assessment that it’s the right thing to do? How do you think about that?

[00:24:44] Ryan  

Hopefully, if we’re adding a model, it’s because we’ve pre-understood a weakness. Let’s say we’re just sitting around chatting, and we’re like, “Well, what if X, Y and Z happens, how would our mix of models handle that scenario?” And we realize, maybe not so well. An example might be like crowding. Let’s say there’s a lot of talk now that our industry is going to see an influx of capital over the next few years because that’s pretty standard for when there’s a big run in the industry that the inflow comes 12 to 24 months later. 

So then it’s like, how will that impact our strategies? Is there an obvious negative to that? Is there a model that, if that negative scenario presented itself, this particular model would actually do really well in that spot? So, then if we all are in agreement? Yeah, okay, this type of model would do well in that spot. Then we start doing this lit review and trying to develop that model before it actually becomes necessary or useful and then have it ready. So, that would be the best-case scenario for how things get added to the portfolio. 

Then the worst-case scenario (not worse, sort of the other) would be like when we had a model that’s core thesis was globalization. It was pretty clear on 2016-17 that that trend was reversing. It had nothing to do with the model’s performance. We didn’t do a t-statistic or this or that or the p-value or any of those things to say this model should be turned off. It was just the thesis behind it didn’t really match reality anymore and so we turned it off.

[00:26:39] Alan  

Okay. In the whole research world, the big topic in the last number of years has been machine learning and that there has been great strides in what can be done from a machine learning perspective. My sense is mixed success, some managers seem to have done well with it, while others less so. Is that something you’ve extensively used in your approach? What do you think are the merits and pitfalls of incorporating machine learning?

[00:27:09] Ryan

Machine learning is obviously really, really good at certain things. We just haven’t seen that it’s good at this yet. There’s probably people out there who have figured out ways to use it. But I think the issue is your data set coming in is a constantly changing data set that is resistant to you solving it. This isn’t quite machine learning but let’s say someone writes down a series of rules on a piece of paper and says, “Well, I’m going to manage $10 billion on this and make 10% a year.” It’s like you expect the market to lose a billion dollars a year to you with rules you wrote on a piece of paper. 

The market will adapt, and so the market is constantly adapting. That’s where I think machine learning might struggle. The spam email is the classic example of how machine learning has gotten better. Spam is kind of a good test for machine learning because spammers are constantly trying to adapt themselves to avoid getting caught. But that is the nature of financial markets. They’re resistant to being exploited and they adjust to you. 

So the GSCI roll trade, again, a machine learning model would have found that but it doesn’t know when it stopped working. It’s not aware of how the adaptation happened. Linear regression is a machine learning technique, just a very simple one. I’m not going to say we don’t use it at all but we certainly don’t use it to the extent where we have models that work and we don’t understand why they work.

[00:28:52] Alan  

Okay. You touched on maybe the example of the possible influx of capital into trend and that might be a model for consideration. In broad terms, what are the new areas for research that you’re excited about? Is it alternative datasets or is it just dynamic opportunities that might present themselves as markets evolve?

[00:29:23] Ryan

Flows definitely affect markets and so that will become more important if the number or if the capital behind strategies like ours increases. So that is kind of a big topic for us right now. You mentioned alternative data. Obviously, if you’re an alternative data provider, you’re knocking on every single door so we get to see all these different things that people are collecting and doing. I think a lot of it doesn’t apply to us because it’d be more for equity managers or long-short equity, things like that, because we’re trading big things. The one source of alternative data (not really alternative but just maybe new to us) would be some sort of statistics on what people are holding. But I wouldn’t say we’re big consumers of alternative data.

[00:30:17] Alan  

Custody holdings, things like that, positioning and flows, that’s what you’re getting at there? 

[00:30:22] Ryan

Yeah, but even then, it’s tough because individual large custodian-type names want to sell that sort of data, which we find to be kind of not kosher, not cool, I guess, like they’re sort of profiting from what their customers are doing. So that sort of data we shy away from. But regulatory agencies, if they’re reporting data like that then we’re interested in that.

[00:30:52] Niels  

Yeah, just staying on that just before I move on. I wanted to ask you a little bit about different types of diversification. There are two topics that often come up on the podcast, one is the view about how many markets you should trade. Also because this is my understanding of what you’ve been talking about so far is that the diversification across different types of models is also something you apply and find very valuable. 

So I’d love to maybe ask you to talk about what you think the differences between these two types of diversification, within a systematic strategy are, but also specifically, maybe in terms of, do you think there is a limit in terms of how many markets you really need to focus on likewise, how many types of models you should focus on?

[00:31:47] Ryan  

Yeah. Both have the same risk. Let’s do it with markets. That’s easier to explain. So, let’s say you have 10 markets and your expected return or expected Sharpe… Let’s just stick to trend following. So, you have a very simple trend following system, you apply it to one market, you assume the sharpe of that is 0.2. If you could apply it to nine uncorrelated markets, now your Sharpe is 0.6 (if those markets are really uncorrelated). Then as you move higher and higher and higher, you theoretically should be able to achieve a higher Sharpe ratio and you should theoretically benefit monotonically from adding markets. 

The caveat to that and the risk to that is that, as the number keeps going up, you’re probably targeting a return. If you trade one market versus 100 markets, you’re probably going to put more than 1/100th of your NAV on each market because you’ll say I’m getting diversification, I still need to target a vol of whatever your target vol is and so I actually have to trade 2x 3x or 4x what I would have traded if I only trade one market. 

Now you’ve introduced the risk that your correlations are wrong on those different markets. You can make the same exact claim about strategies that are supposed to be different. So, within the markets, the number of markets, we feel like there’s kind of a right number and we’re always thinking about what that number is and to the extent that it can be expanded, we obviously would love for that to be true. 

One of the markets in China that are not really open to non-Chinese speculators (or maybe they are but the ways to approach them are sort of unfavorable at the moment), when you load them in, they look pretty interesting. They look like they could definitely add something. So there’s always going to be an opportunity out there for adding markets. It’s the same thing with strategies. 

One of the things that you can look at for how to mix strategies and how to try to give yourself some confidence or some assurance that you haven’t made that over deleveraging mistakes (so, you sort of consider them at extremes like what is the max position long or short these different strategies could get). That’s really where it matters in terms of making a big mistake – having strategies that you think are different that all of a sudden become similar. That only matters when huge things are happening. 

You can do a lot of scenario-type analysis both like simulation and just thought scenarios of how models should react. But again, in both cases, in theory, there should be nothing but benefit to adding, but in practice there is a backward bend, at some point, when you’ve added too many.

[00:34:57] Niels

Yeah. Given your AUM now, how many markets do you actually trade? I don’t know how diverse. 

[00:35:06] Ryan

I think it’s 97. 

[00:35:09] Niels

Okay, about 100 markets. 

Now, I want to shift gears a little bit to something that became a big topic (at least in my little bubble) last year, maybe partly because suddenly these strategies caught a lot of attention, and this time around some good attention because everything else was losing money. Suddenly replication of these strategies became a thing and we saw, obviously, one particular ETF do really well in terms of growth in AUM which replicates the SocGen CTA Index that you are a member of and most of the people we’re speaking to in this series are a member of. 

So, I wanted to ask you a little bit about replication, maybe how it feels to be replicated but also maybe more importantly, what you think of it. Because if you’re familiar with the product, in this case, it’s not a trend model. It’s a linear regression of the data that SocGen delivers, to your point earlier about linear regression and its limitations. But I’d love for you to talk about how you just feel about replication. 

I don’t really want to put my opinion into your thinking, but just how you feel about replication of these types of strategies. Obviously, you know the nitty gritty of them so you probably also have some thoughts about what the pros and cons might be if you want to try and replicate them using linear regression of sorts.

[00:36:45] Ryan  

Yeah. In fact, we have a program that replicates. It replicates whatever you put in.

[00:36:53] Niels

Can it replicate Medallion Fund, Ryan? Because then I might be interested.

[00:36:59] Ryan  

Well, not well, that’s the thing. So it replicates out of a set of really simple toy models. So it says, “I’ve got this set of little models and very simple rules. Let me see if I can put together a mix of these different things to match the returns that I see.” It can get pretty close. The more deterministic a strategy is, especially if it’s limited to the same sort of markets (the biggest markets), then the closer it’ll get. So that’s reality. 

Going back to the project that we give our interns, they can produce a pretty okay return with a series that would have done really well in 2022 with very little knowledge and understanding and really just essentially fitting what other people have previously done. So, our job is to be better than that. We’re obviously very aware of not just replications. It is kind of interesting to replicate a strategy that is (granted, the CTA index is not pure trend, but trend is a strategy, so it doesn’t really need to be replicated) fairly straightforward. 

That would be the only thing that’s maybe a little bit surprising that someone would even want to attempt to replicate a strategy that is, in itself, rules-based and somewhat publicly rules-based. But yeah, we see that as a competitor. We have people that we talk to who have hired managers like that and that’s what they want.

[00:38:47] Niels  

Okay. So we certainly have seen that, as you say, performance can (certainly during periods of time) be very similar and maybe even sometimes higher than the index it’s trying to replicate. So that’s fine. But since you know so much about this field, what are the risks that you see these types of approaches present that investors may not see? They may just be focused on the headline saying, “Well, we can, by analyzing the data essentially, pretty much get the broad exposure that these managers have and therefore we can do it at a lower fee.” 

That’s a compelling narrative, frankly. I personally think that there must be a trade-off here. So, there must be some risks in there. I’m sure you have a much better view of those risks, what they may be, so I’d love to hear your thoughts about it if you could.

[00:39:51] Ryan  

Going back to our approach, we think about how to be as anti-fragile as possible. I guess the drawback to a formulaic strategy like that is it’s not thinking in a forward way about what might impact it negatively and what sort of models could be added to the library that would mitigate those periods. So really, it’s all about how you perform in periods when (let’s just stick with trend following), how do you perform when trend following is not performing well? 

Let’s go back to the spectrum of complexity of a system. People would ask, “Why do you think you’ve grown and other firms haven’t grown when returns prior to 2020 were pretty mediocre in the entire space?” You make this analogy about a bear chasing a group of people and it’s like I don’t need to outrun the bear, I just need to outrun you. So having a mix of strategies that still maintains that exposure that people are hoping to get from you, but not dependent on it, is always forward-thinking. [It’s] not even just about strategies, but also about markets and various risk management techniques where all we ever think about is mitigating drawdowns. 

Because if crude goes to $200 (assuming it doesn’t go there in an instant), every trend follower will profit from that. That’s not the hard part. The hard part is, what if it just stays between 70 and 90 for the next six months to a year? How are you going to handle that? So that’s where somebody like us and somebody doing, whether it’s a replication or just a strict here’s my rules and I’ll even show them to you on a piece of paper – kind of low fee data trend, that’s the value proposition there. Who’s managing the off periods better?

[00:42:00] Niels  

Good point. Speaking about managing the off points, Alan, do you want to dive into some risk management?

[00:42:11] Alan

You talked about managing the drawdown which I guess there are a couple of aspects. Obviously, if we were talking to maybe a trend follower who then moved into multi-strat, I suppose that’s been the journey for some people that they started off with trend and then they wanted to mitigate those drawdowns and they added other strategies. It sounds like your journey has been certainly driven by a philosophy of being multi-strat from the start.

I guess the drawdown will hit when everything is struggling at the same time. So what’s your approach to dealing with that? Is it deleveraging in the first instance or is this just sitting through it? How do you think about that?

[00:43:09] Ryan

There is drawdown management where, as you lose money, you reduce risk and as you come out of the drawdown, you sort of put risk back on. We’re not believers in that, specifically. I think where we’re trying to combat, when we think about managing drawdowns, we’re thinking about having a mix of models that we feel like should respond well. If you look at our performance during our drawdowns, it does tend to be less volatile and that’s just a function of the approach. Obviously, being systematic, we’re going to look at anything that is potentially a systematic value add to what we’re doing. So, you look at these setups where, if you go down one standard deviation, you cut to this and you have this whole rules-based setup of how you top-down change your risk based on where you are versus your high watermark. Try these for yourself, and maybe you already have, it doesn’t work. 

You actually end up getting trapped in these death spirals where you keep cutting and then your performance is good, but you don’t make as much as you would have. But then you put money back on because you’ve come out of the drawdown, and then you go back into another drawdown, you cut risk again, and so you get stuck. You can’t get out. I’ve had colleagues that we’ve talked to that ended up down more than they would have been if they had no drawdown management. 

So that’s part of it, and the other part is we have no idea how we fit into someone else’s portfolio. It would be off-putting if we were doing stuff like that when that might change how we’re balancing something else that someone’s got in their portfolio because of just our personal cuts based on drawdown.

[00:45:18] Alan  

You mentioned that you’re obviously multi-strat but then present, in terms of return profile, very similar to trend following and you’re a part of the index, etc. So is that just a function of the risk allocations and that although it’s multi-strat, that trend following does tend to dominate either to risk allocations? Why is that?

[00:45:39] Ryan

Well, yeah. It’s the presence of trend and the weight that it has, first of all. So let’s say you’re in a carry position, and with carry you’re not necessarily paying attention to direction. The whole point is it’s a yield. But certainly, if you have a reactive philosophy, then if volatility goes up, you will cut the position. So now you’ve taken on a trend-ish concept.

So we’ve noticed that our carry (not just actually ours), but in general, carry tends to have a positive correlation to trend because when it’s working, it tends to be left alone versus when it’s struggling, then reactive strategies will reduce, certainly, the notional value. So, then it will sort of pick up a correlation to trend in that way. In general, when you’re reactive, when that’s your philosophy, then you will always have a correlation to trend since trend is a reactive strategy.

[00:46:45] Alan

Okay. In terms of managing the risk in the portfolio, as you said, it’s multi-strat so you have trends, you’ve got carry-type strategies, and I think you mentioned value short-term strategies. Presumably, then at times, everything will line up and you will have higher exposure. So it’s managing gross exposure, the philosophy or is that the long side of vol target? Maybe give us a sense on how you think about the particular metrics you’re looking at.

[00:47:17] Ryan

Just as a failsafe, there is a pyramid of various concentration and limit checks. So you’re just trying to make sure you’re not over-concentrated anywhere. Because we’ll get like, what if all your strategies say the same thing? They actually can’t. You can have situations for sure where, let’s say, two of them are similar, like trend and carry certainly will spend periods of time being similar to each other. 

The whole point of mean reversion is (going back to the cyclical way of thinking), think of a sine wave – it starts to move, your trend is profiting from it, and as it approaches the upper bound of where it’s going to go, that’s where you would hope your value and mean reversion would start to lean against it. With the pure trend approach, one of its drawbacks is it tends to have its largest position when the trend is about to end. 

So by thinking of markets more cyclically than in trend, that’s where you can design your combination of systems so that, in fact, you’ve started to reduce as the cycle starts to exhaust itself. But certainly, we spend a considerable amount of time thinking about and checking and have systems in place to make sure there isn’t over-concentration.

[00:48:47] Alan

That kind of idea of not getting too large at that point when the trend might reverse is a choice in a sense that at the same time, some of the big months in trend following are when you get the big position on and the trend extends. So is that just a choice to try and deliver more consistency of performance as opposed to having that really convex return profile? Do you think it’s just a better way to run a portfolio over the long term?

[00:49:21] Ryan  

There are people who are just pedal-to-the-metal. If something’s going, they’re not getting off that until they’re told that they’re wrong, and it’s going to be full. So it’s not necessarily true that we will not have that. If we are prematurely reducing positions due to a mean reversion model, that means that model is losing money and so we keep track of how our different models are doing. That is different than us. 

When there are people out there who that’s their philosophy, you can make an argument that there’s a place for both of us -that philosophy might be useful. If you’re only allocating let’s say, 1% to the space, maybe that sort of concept is preferable. So I think, yeah, we’re aware of that approach and it has its merits. But for us, it doesn’t quite fit with what we’re doing.

[00:50:21] Niels  

Yeah. You mentioned something earlier, Ryan, where you said, people might ask why have you grown during a decade where the industry has basically stayed without growth, if we take away the performance growth we’ve seen in 2022, maybe to some extent in 2021. It also brings me back to, I think, what happened after the great financial crisis, you also alluded to that after these periods of time like 2008, maybe 2022, we may see an inflow of money to the industry with a lag of 12 to 24 months. Now, how do you think we as participants, we as managers in this industry, how should we deal with capacity in general? 

I guess one of the things that I wanted to explore a little bit more, which maybe we haven’t done in the other conversations, is this balance between those who have essentially structured at least their trend following strategies as relatively low fee, maybe even management fee only, and where you could say that the only way for them (from a profitability point of view) to make a big impact, that is just to take in money from clients. 

Whereas other firms, where they focus perhaps more on the performance fee, they might be much more sensitive to the point about capacity. Where does it start to hurt performance? I’d love to hear your thoughts about this both as a manager but maybe also as an industry because I think we got a little bit of a bad rap last time from 2008. To some extent, people, rightly or wrongly, perhaps blamed it on some issue with overextension of capacity even though I don’t really think that that was the reason. But, at some point, strategies do hit capacity. How do you think about that?

[00:52:24] Ryan

Yeah. 

Just a general statement about fees, I think people charge as much as they can. So if a strategy is simple, then the fee is probably low because that’s what the market will bear. And so it really just comes down to what is a strategy worth to somebody. Performance fees, in general, that has been considered favorable in our industry because it creates the alignment of interests that the manager is on the same side as the client. 

A management fee-only structure, while that might be in a Gangbuster year, then you feel proud that you hired somebody with a management fee only because you didn’t have to pay that big performance fee. But on the other hand, that tends to be (and I think this kind of what you’re hinting at) more of an asset-gathering exercise. 

To the next point about if there is an influx of money into the industry and how will that affect things and did it affect things? What’s interesting about ’07 and ’08, the money wasn’t there yet in our space. So, those moves had nothing to do really with what our industry was doing, and so then the money came afterwards. Obviously, there was money in our industry, but the inflow didn’t create the disruption, the disruption created the inflow. 

When thinking about it, going forward again, I come back to this again that if assets make large moves, then we’ll be fine. If we see another bout of inflation like we did in the first four months of this year, that will be universally good, whether there’s the amount of money we have now in the industry or double, those trends were pretty much unmissable. The whipsaws liquidity, how do you handle it? 

So if you’re wrong, and you decide you want to get out and there’s a lot more money doing something similar to you than there used to be, then the price you’re going to get is going to be worse than before that money was there. So if you don’t change anything about the way you’re interacting with the market and more money comes in, then where crowding will hurt will be when trend is not working. 

In the old, old days, trend following either made 20% or broke even and then in the newer times, as more and more money came in, trend following either does well or loses a fair bit and that loss is a function of those whipsaws hitting harder because there’s more money chasing on that idea. So, this goes back to what I was saying. It comes down to, can we do something to counteract that? Is there a way to adjust? And that adjustment would be in some sort of non-trend model to account for the loss that’s going to take place. 

So those would be the biggest risks. Markets don’t move in excessive ways because of CTA flows. They move because of something happening in the world. 

[00:55:53] Niels  

Yeah, let me try and then summarize it a little bit and ask it in a slightly different way also, and that is, do you believe (let’s just call it that, the official CTA industry is now at maybe 400 billion), do you think that there is a limit to how much money can be in the official CTA industry (obviously, I’m sure there are lots of people doing trend following that we don’t see). 

If we think about this, if we see a large amount of growth into our industry, does it actually force managers to revisit the way they do things, so to speak? Meaning, as you rightly pointed out, if we all do things more or less the same, then there could be a lot more crowding at the turning points or whenever we identify the trends to have come to an end. 

You, as a manager, does that give you ground to think about, “Well, hang on, if the industry goes from 400 to 600, maybe we need to think about how we get out of markets, and so on and so forth.” I’m just thinking out loud here. But what are your thoughts on this?

[00:57:14] Ryan  

Yeah, well, definitely. We do, and that definitely should be thought about. Another thing, it’s sort of a closed system of participants. Hopefully, if the CTA balance is growing, maybe the other players in the game are also growing and so it’s possible that we could grow, but the whole market could grow, which it has. If you look at statistics of open interest over time, commodities markets have grown tremendously over the last 20 or 30 years. So, that would be the flip side is hopefully there’s room for that extra money on the strategies. 

But let’s say that people don’t adapt, everyone just kind of keeps doing the same thing they’re doing, then what you would believe would be that the strategy should just become a little bit less profitable. Actually, there’s an interesting analogy that there was a poker boom in the mid-2000s from the whole cards being televised in the World Series of Poker and things like that. So, there was a flood of new people playing poker and then there were some people who were really good at it, and they were performing really, really well. 

And then as time went by, the participant pool started to shrink but the pros were still there and so now they were competing. So for them, it was different. it wasn’t that they were growing, it’s that their pool was shrinking, but sort of the same scenario where the percent of participants who are profit seekers was going up. So it didn’t drive them away, but it just changed things to be a little bit more difficult and then the best players had to adapt. Probably what really came from that is that you started to differentiate between the kind of good players and the really, really good players. The really good players could actually exploit the kind of good players and the bad players. 

What could happen is that it would put more pressure on the overly simplistic approaches, then the differential between the good and the just okay would grow.

[00:59:27] Niels

Yeah, I think that’s a great point, Ryan. Actually, also while you’re speaking there, I’m kind of reminding myself that when I started in this industry, there were less than $10 billion in total. Even back then people were talking about, “Well, there’s too much money coming into trend following,” yet here we are. I think that it’s the adaptability and just the smart minds that are in this industry that find a way to deal with that, which of course would be a positive problem. 

But let’s see if investors, as you rightly say, will be smart enough to detect that some types of strategies might actually suffer compared to others depending on “their level of sophistication.” 

Alan, why don’t you go through your last round of questions and then I’ll wrap up with a few more?

[01:00:23] Alan  

I suppose just shifting over to the perspective on the role of the strategies you run more in a multi-asset portfolio. You talked about your engagement with clients. We’ve kind of touched on this a little bit, but maybe it’s slightly different in your case. If you’re running a pure trend model, it’s very much positioned as a diversifier, as something different to what maybe a typical pension fund or an endowment holds in  your portfolio. 

By running a multi-strat, are you in any way giving them some exposures that they might already have or not? How do you think about that? The kinds of strategy, the portfolios you run, do they fulfill the same role or is it just an absolute return allocation in the context of a multi-asset portfolio?

[01:01:19] Ryan

Yeah, it’s sort of this fascinating reality that everybody knows that if you just do stock/bond/cash, and you throw the CTA index in there, it’s going to make the ratios look much better. There’s almost no point in even having that in a presentation because everyone agrees on it and yet the allocation to strategies like ours is way less than what these types of models would indicate. Then the question is, why is that? 

One reason is that maybe these ratios are not necessarily what people are focused on. They’re focused on a return goal and so it doesn’t really matter that much.  We’ve spoken to family offices, and it’s like, “Do you make 20%?” “Well, no.” Well, that’s our goal every year. They don’t mind if they lose 40% because they just figure it will come back eventually, and they don’t really have vol targets. 

That would be the first question is, to the extent that you have a target vol, that you’re saying, “This is how much risk I want to take. What strategy should I employ to feel best about the return that I can achieve?” That’s where having something like us  (any member of the CTA index) would be beneficial to a portfolio like that. That’s always the starting point; that there’s a self-selection. There are some people who do think that way, and some people who don’t, and the ones who do are the ones that sort of come around and they’re going to be looking at all the managers like us. 

But then there are questions. Specifically for us, people say, “I’m worried that your strategy XYZ might be similar to what we’re doing in equity (or something like that), because isn’t that a risk-on strategy? So why are we adding a risk-on strategy to our book of risk-on strategies?” So, that’s where we can get into great detail and just see if that’s true or not. And it turns out to often not be true because we’re just in a completely different space. 

Commodity carry, even if you just looked at a ticker on Bloomberg, the most basic commodity carry system that’s long backward and short contango, that’s not a strategy that has a correlation to equity beta at all. In fact, if anything, it’s got really strong inflation properties and it had a really great year in 2022. When you have shortages, that’s typically a scenario where prices are going to be rising and so backwardation is going to be long and prices are moving higher. 

To the extent potential clients for us or for anybody else are willing to really drill down and see. You can always start with the numbers. Anybody is probably willing to give you at least monthly numbers and you can load them in and see, okay, this makes ratios better. Then you can have that added level of comfort of trying to drill down and see what the strategies are individually and how they match off. 

Our goal and the goal probably of everyone in our space is just to try to get through that first step. There are so many managers. I think that’s also probably the hard part, “Who do I pick? I feel like I picked XYZ manager and then they had a really bad year, and I wish I would have picked this other one. So just forget it, I’m just not interested anymore.” So, just trying to work through that.

[00:1:05:16] Niels

That’s what the replicators say, Ryan. That’s exactly that pitch you gave there. So we have to be careful with that. Now, I have a couple of questions left, one that I’m just curious about it. If there’s one thing that you hear people say about trend following that you really disagree with, is there a of thing that you come across where you say, “Yeah, no, I definitely don’t agree with that?”

[01:05:44] Ryan

So, there’s a question of trend beta. I think what’s kind of neat about trend is that there is obviously no individual beta and the proof of that is people who claim to be pure trend following have really different returns, like way more different than if five different firms have an S&P Index Fund, those performances are very similar to each other, but five different firms could say we are pure trend, and they have totally different performances. I think, really, it’s that there’s trend basis, that there is an exhaustive number of strategies that, in a theoretical sense, any trend follower could be represented by some set of those functions. 

That can be interesting. Let’s say someone, they just want pure trend. I think it could be a really fun exercise for them to canvas that basis and try to construct the strategy that fits best for them. It’s not something that we do. When you’re evaluating how a trend follower did (this goes back to the fitting question), you can find the mix of those things, the nearest neighbor to that return stream and then see, did they outperform it? That’s where the skill comes in. 

Trend is more interesting than just beta, especially the way we think about it, where the markets are made up of cycles of different frequencies. So, if you’re employing trend, really what you’re trying to capture is the meat part of those cycles. So there’s some skill in understanding, should I be short-term? Should I be long-term? Should I be medium-term? Should I adjust that over time? Why do I adjust that over time? It is definitely more interesting than just a very straightforward here’s the beta, anybody could do this.

[01:07:52] Niels  

Yeah, I completely agree. Final question is just whether there’s anything in particular you’re most excited about as we head into 2023 or concerned about? It could be anything.

[01:08:05] Ryan  

Our two best quarters are quarters where bad things happened, but that’s just the nature of what we’re doing (when I say we, it’s the whole space).  So, what are the compelling reasons to be in our space? It would be the argument that people made from 2010 through 2019, is “Oh, well, the central banks controlled everything, so nothing could move.” And that’s definitely true for equities. That created a difficulty for trend following in equities. 

Actually, that fascinating response to that is a lot of the bigger names adapted by becoming much, much longer term, because it meant that they were long a lot. So that adaptation ended up being disadvantageous in 2020. So that’s where you do have to be careful about adjusting based entirely on past performance. 

It’s easy to say that the Fed (not just Fed), central bank control is probably going to be less over the next decade than it was over the past decade; that their ability to control things should go down. It’s more of a political statement that as you’re seeing more and more populist governments, with sort of the old school approach of just let the Fed do whatever they want and give them unlimited resources to do so. If anything, it’s more likely to not be true than it is to be exactly the same. 

So that would be kind of the start. And then (this was actually mentioned in one of your previous podcasts), just the rise of competition, nationalism, protectionism, inflation, versus a decade of globalization. When you have globalization everyone’s cooperating. Why would crude go from 20 to 80 to 20, back again, in a period when everyone is sort of playing nice together? If you think of 2014, the reason crude moved so much was because countries were not playing nice with each other. 

One group of countries was trying to put another country out of business by jamming the price of oil lower, and that was a fantastic trade for everyone in our space. So, to the extent that those forces continue to grow (COVID was obviously a huge contributor to volatility; the Russian invasion of Ukraine was a huge contributor to volatility, especially energy volatility), so it’s those types of events that are generally favorable to adaptive strategies, to strategies that don’t have a bias, that don’t have a benchmark, that react to things and sort of lock onto them and stick with them. It’s tricky, because people come and say, “Well, did I miss it?” It’s like, “Well, try to think 10 years.” It’s a really tough thing where if you try to think of trend following as something that you missed or that you get in and you get out based on a view, that is a very difficult way to manage it because you’re just always wrong-footed. 

To the extent that people have the more steady-state approach, we’ve seen profit-taking, which is great. Opportunities will come up in the future where maybe traditional assets will do well and then you start to see that flowing back. That’s the right way to play it. Anyway, I’m sort of going off-topic there.

[01:11:53] Niels  

No, it was great. Actually, it’s super to hear a different explanation to some of the things we talk about. So I really do appreciate that. It’s actually an excellent note to wrap up this fascinating conversation. Ryan, thank you so much for being on the podcast and for sharing your thoughts and insights with us. We hope we can do this again sometime in the future. 

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 at the website and not least, take care of yourself and take care of each other.

[01:13:02] 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.