TTU144: Brian Proctor, Managing Director at EMC Capital Advisors

[00:00:07] Brian

The real key is to be able to put a position on and resist the urge to liquidate it, to take profits real quickly. You want to trade actively, cut your losses quickly, and let your winners evolve and run over time. Your winners will turn out, with good systems, to be two, two and a half times bigger than your losing trades.

[00;00:34] 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;01:30] Niels

Hey everyone, and welcome to another edition of Top Traders Unplugged, where today Alan Dunne and I are joined by Brian Proctor, who is the Managing Director at EMC Capital Advisors, as part of our mini-series focusing on the one investment strategy that beat everything else in 2022, namely, trend following in managed futures more broadly.

First off, Brian, it’s great to have you back on the podcast. It was all the way back in April and May of 2017 that you were last here along with your fellow turtle, Jerry, as well as your mentor, Richard Dennis. So thanks so much for joining us today. We really have been looking forward to catching up. How are things on the shores of Lake Michigan this morning?

[00:02:09] Brian

Everything is going fine here. Happy to be on the show once again and look forward to getting grilled by you guys.

[00:02:17] Niels

Oh, we’re not that tough, are we? Now Brian, before we dive into all of the various topics that we’re going to discuss, we always like to set the stage so that the audience knows a little bit about your firm. So perhaps you could share a few highlights, because the history is a little bit interesting. Maybe you can talk a little bit about that and the evolution, and where you are in terms of strategies and business as of today.

[00:02:46] Brian

Okay, happy to do it. The firm was founded by Elizabeth Cheval. Liz got her start on the floor of the Board of Trade back in the early to mid ’80s. She was chosen to be one of the participants in the Turtle Program, which was a mentoring program run by Rich Dennis and his partner Bill Eckhart. Their goal was to try to see if they could train traders using set rules to become profitable traders.

Liz was one of the initial turtles chosen to be part of that program and did very successfully. When the program was over, a few years later, she founded EMC Capital in 1988. Liz passed about a decade ago and the firm was taken over by a guy named John Krautsack, who’s our president now. John joined EMC in 1995 and he also had extensive experience on the floor working for some very big CTAs at the time, helping them run their investment strategies before joining EMC.

Our Director of Research, Dave Polli, joined the firm in 2002, so he’s been here over 20 years. You can see a lot of continuity in our firm. I joined in 2002 after starting on the floor of the board and the Merc in the early ’80s, eventually running the floor operations at each of those exchanges for Rich Dennis and his company, C&D Commodities.

I also joined the second wave of turtles and traded for Rich in that program before branching off, running my own CTA. I worked in New York for a while for a long-short equity hedge fund and then came back to Chicago and joined up with Liz in 2003. So that’s it in a nutshell. Obviously, being part of the Turtle Program was really beneficial to learn from two of the real pioneers in our business and the importance of managing risk, the importance of building robust systems, and just the emotional side of trading too. Try to remove the emotion from trading so that you can run programs without trying to make subjective decisions on a daily basis.

[00:05:35] Niels

In terms of strategy, how has that evolved since the early days of Liz?

[00:05:40] Brian

It’s evolved considerably. Back in the early days of EMC, she was incorporating more traditional breakout systems that she learned in the Turtle Program. Over the next decade, I would say, in the late ’90s, she incorporated different types of trend following and momentum-based systems. She got away from buying and selling around breakout points and really overhauled not only the trading systems that were in the program, but the amount of leverage used, the risk management protocols that got put into place. The industry really evolved from a period where the risk-free rate was very high when we started out as CTAs.

The target, our targeted returns and volatility from the investors who were putting money with us were geared to be maybe three times the risk-free rate. They were looking for higher returns, higher volatility. Over the years, the appetite for those types of programs changed and so we had to evolve. There’s still a considerable amount of risk that we take when we trade, but it’s nothing like we were doing in the late ’80s and early ’90s.

[00:07:11] Niels

Sure. Absolutely. We’ll certainly dive into both of your strategies during this conversation. Now, Alan and I have created a list of topics and we’ll sort of interchange between us. But as we normally do, I’m going to kick it over to Alan to get it going. Alan?

[00:07:32] Alan

Thanks, Niels. Brian, thanks for the introduction. That gives us a great sense of the background on EMC. Obviously, a lot of the people there are steeped in the turtle trading experience. So if you bring that initial experience to the present, how would you describe the philosophy in the sense? Does it continue to be purely focused on trend as the primary source of return? I suppose, are you adherents of the idea of using multiple sources of return? How would you characterize the philosophy more generally?

[00:08:13] Brian

I would say our investment philosophy has stayed pretty much the same, and that is that all the relevant market information is reflected in the price of whatever market we’re looking at. The quantitative analysis is the most effective method to identify predictable trends in price. Historical data shows that markets consistently trend and they yield identifiable and repeatable profit opportunities.

Our objective was to develop and implement quantitative systematic strategies that obviously deliver long-term positive returns with low correlation to other asset classes. All the models that we build, they get re-optimized from period to period, are always looking to capture directional price movement.

[00:09:21] Alan

You mentioned absolute return, low correlation. I suppose it can be seen as trade-offs between generating return or generating diversification for equities or even crisis alpha. When you construct your portfolios it sounds like you’re first and foremost thinking from a return perspective, and those other factors are secondary considerations. Is that fair to say?

[00:09:50] Brian

The returns are the first thing that we focus on, it’s what everyone focuses on. They want to invest in something that’s going to generate returns in a certain fashion. The returns that our particular strategy, as most CTAs do, the returns are just naturally divergent and different than the return streams of other assets they have in their portfolio. As long as we have long-term positive returns that are what our investors are expecting, that’s our main objective.

[00:10:33] Alan

It’s always a tough one to answer as a manager, but from your experience of interacting with other CTAs and other trend followers, is there anything you would point to as making you particularly different relative to your peers?

[00:10:48] Brian

Well, we’re all looking to do very similar things. The difference between one CTA, and the next, and the next, every little thing that we do is going to be a little different from other CTAs. I think the main areas that probably differentiate us a little bit are, first of all, we have a significant allocation to commodity risk in our portfolio. That’s approximately 45% to 50% of the risk we allocate across our strategies goes to commodity markets.

Second of all, we have a blend of systems in each of our programs. The systems are based on two core concepts, trend following or range-dependent systems and momentum systems, which are looking to capture, as I said, directional price movement. We try to diversify these systems so that when they’re initiating and liquidating, they’re doing it at different timeframes and under different market conditions. I would say our systems have an average holding period of 35 to 45 days depending on the period.

I would say we fall into a medium-term rather than a long-term trend follower bucket. We have some systems that may hold a position, on average, only 15 days, other systems 35 to 50 days, and typically one or two real long-term systems where we’ll hold a position 100 days or longer.

[00:12:36] Alan

Just before we move on to the next bit, going back to the whole experience and learnings as a turtle, it’s interesting you talked about managing risk, building robust systems, removing emotion as maybe the key elements as opposed to necessarily trend following. Was that experience? Was it as much about learning to manage risk, or was it much about imparting the philosophy of trend following would you say?

[00:13:03] Brian

Well, Rich and Bill were very adamant that if you don’t manage your risk properly you are not going to be a successful trader. That was always one of the first lessons they taught us. Then the second was trend following, that the majority of your systems, the winning percentages aren’t going to knock anybody off their stools. The real key is to be able to put a position on and resist the urge to liquidate it, to take profits real quickly.

I know Bill used to say the one adage that you never get broke taking profits he said was absolutely wrong. You want to trade actively, cut your losses quickly, and let your winners evolve and run over time. Your winners will turn out with good systems to be two, two and a half times bigger than your losing trades.

[00:14:09] Alan

Just one final one. One of the debates that’s come up on this podcast, and more generally is kind of linked to volatility sizing, and maybe all of that stuff. It seems like the treatment of open equity is different to the treatment of your trading stake. It’s like if you have open equity in position, you should let that run and keep those positions open, whereas maybe the more typical managers now don’t make that differentiation. They manage all dollars equally. Is that a fair characterization would you say? What’s your perspective on that whole debate?

[00:14:47] Brian

Yeah, I think it’s a fair characterization. We have changed. One of the things in our risk management, it does take into account, at the portfolio level, open trade equity and the size of unliquidated profits that we’re building when trends are occurring and when the program has a sizable amount of risk on. There is an algo that we have in all of our programs that there’s really two elements to it.

There’s a profit element when we reach certain thresholds, in terms of profit based on open trade equity and the account size. The other element is a time element. You need trends to mature a little bit before you want to start scaling back the positions. That’s a little bit of a departure from what we used to do where we would let our profits run until the trade ended by reversing the long-term trends.

That particular algo kicks in during highly profitable periods where CTAs with our right tail skewness, it’s those periods that this algo cuts back positions, locks in profits, and really gets risk more in line. Usually what’s happening during those periods is increases in volatility in the portfolio. When we’re initiating trades, obviously, the volatility of that market does helps determine the sizing of the number of contracts that we’re going to trade.

As vol increases, even if all of our systems are long in a particular market, there’s a max risk per market when you initiate the trade. Then as vol increases, over the life of that trade, obviously, risk is increasing. We don’t reduce positions solely based on increases in volatility. That’s what we consider good vol. There’s good vol and bad vol. Good vol is when you’re positioned correctly and volatility is increasing. That helps drive the returns of trend following and momentum-based CTAs.

The algo that we put in place gives us a way to gradually get risk more in line with the current vol, take some profits off and cushion those periods where when trends are ending, we’re giving back some of the profits.

[00:17:24] Niels

Yeah, that’s certainly an interesting departure, which I’m sure we could have a whole podcast discussing but we’re not going to do that today, Brian.

Now, one of the questions that we’ve discussed with all our guests is a little bit the role, to some extent, that our strategy should play in a portfolio in the sense that we know that by including them, we or the investors achieve a higher risk-adjusted return. In lack of a better measure, we’re just identifying that as Sharpe.

However, we also know that a lot of managers have taken it on themselves to try and improve the Sharpe of their strategies, because as we know, it can be a hard strategy to hold on to. We’ve heard different perspectives of that and feel free to talk to us about that. I want to add one more thing to your particular question, and that is, if you want to stay true to trend following, as clearly you have, what are the things you can do in a sense to enhance the risk-adjusted return but without departing and adding other things just for the sake of improving the Sharpe? Feel free Brian to go wherever you want to with this.

[00:18:45] Brian

Okay. Well, the role that I would think CTAs have in an investment portfolio would be number one, just to add diversification, definitely different return streams. As you mentioned, that improves the overall portfolio returns and should lower risk. That’s always key. For us, we actually have one program that remains true to the higher risk, more trend following element. We also have developed other programs that looked at building a program with a smoother return stream. It incorporated long-only elements in equities, in bond holdings, in commodity holding such as always having a long gold holding in the portfolio.

So those programs were designed with a lower targeted risk-return profile, lower drawdowns, and just a more palatable return stream for the investor. So we gave the investor the choice what type of investment they wanted to make, what made sense for them, the amount of risk they’re comfortable with, and how either one of those types of programs fit into their investment objectives.

[00:20:14] Niels

I just want to stay on this for a little while longer. I can see if you’re adding some long-only stuff or some long bias stuff, I can see that makes a difference. Maybe I’m speculating here, but have you also had or tried to see how you could improve the risk-adjusted profile without departing from pure long-short diversified trend following, or is that just simply too difficult nowadays?

[00:20:45] Brian

Well, for a long time we’ve had a little bit of a long bias to the long trades in the portfolio. When we are initiating long positions as opposed to short positions, in our more traditional trend following program, that has helped improve returns. The biggest winners seem to always be on the long side, although you can point, certainly, to short trades that were just phenomenal. Shorting the energy markets a number of years ago when the pandemic started; things like that, being short the base metals or whatever.

There are times when the short trades definitely are the outperformers in the program, but historically, decade after decade, it’s been the long trades. You can always add a slight bias to the long positions, taking them a little bit bigger than the short positions. The other thing I think is when you’re doing research and you’re looking to stay abreast of how markets are changing and evolving from year to year and decade to decade, your process for re-optimizing your systems, using the new data inputs from year to year, is important.

Our systems all look at a combination of return metrics over varying timeframes. So we’re looking at the return over the entire historical price data that we have and then combining it with say, a 5-year or a 7-year Sortino number to improve that risk-adjusted return. Definitely, all of our systems have a short lookback window as part of the equation to try to adapt and evolve those systems as markets change over time.

[00:22:41] Niels

Yeah, that’s interesting. It’s rare that I’ve come across people who have actually done what you said, this thing about actually giving long trades a little bit more weight. Even though we all know as managers that when we look at our trade stats, by far the most profitable trades have been the long side of trades. I don’t know. Part of that is probably because things like equities, things like bonds (at least up until last year) were in some phenomenal uptrends. It is very interesting to hear. I appreciate that.

Alan, where do you want to go next?

[00:23:17] Alan

Yeah. Well, maybe just to move into some of the research and enhancements in the program over time. It does sound like that the program has evolved fairly materially from maybe the original program. You mentioned the change in the leverage and the risk protocols. It sounded like maybe that was driven as much by the institutionalization of managed futures and the investor requirements. That’s definitely useful to get a sense on what are the things that can be enhanced in a trend following program over time.
Obviously, a lot of times you hear about trend, it’s the importance about maintaining stability in the system. It would be useful to get your perspective on what things can be evolved over time in this type of program.

[00:24:11] Brian

Well, the things that evolve in the trading systems, our systems stay pretty set right now. Unless we go for a number of years in a row where a trading system just is always at the bottom of the barrel, so to speak, in terms of return, absolute return, risk-adjusted return, it just doesn’t seem to be functioning properly anymore, then obviously, that system will have to be altered or removed from the portfolio. We haven’t done that in a long, long time.

The same core logic for our systems, both the trend following and momentum, have stayed in place. What changes over time as the parameter sets of the systems. Think of it as each system basically has a few parameters (the fewer the better) that tell you how to initiate a position. You have a few parameters on how to liquidate the position. Obviously, there is a max loss for each system every time you initiate a trade. More often, there’s one of the other system liquidation parameters that kicks in.

We very rarely (maybe less than 15% of the time) hit the hard stop when a system initiates. It’s some other liquidation criteria. Equally as important has been the evolution of what parameters you include in the systems that block you from taking trades in certain environments. We have a couple of parameters in each system that are looking at the historical volatility in each market that we’re trading and looking back over x amount of days. It could be 200 days or longer.

If the volatility in that particular market for that system reaches a threshold, say it’s in the top 15% of the volatility, looking back over x period of time, each system has a filter based on vol that’ll stop it from trading, because research shows that when you trade in a very highly volatile market environment, you’re more often to take a loser and you’re more often to hit that hard stop. You wait till the dust settles, so to speak. It may take a week or a couple of weeks for the volatility in that market to abate a little and the system to then allow trading to happen.

Each one of our systems has a vol filter. They vary from system to system because we don’t want to have such restrictions that we miss a trend. We always want at least one or two systems kicking in even in really highly volatile times so that we don’t miss putting a position on and having to tell our investor. He said, “Geez, why didn’t you have this market on?” We say, “Well, the vol was too high.” He said, “I’m paying you guys to take risk and manage risk, and not miss those trades.” Each system has a different vol filter.

Each system has what we call a range-dependent system where, in an open range filter, the market has to be moving in the direction for either the initiation or liquidation that has to occur. We look at the average daily vol in that market. Which for us, it’s a pretty short lookback window. We determine volatility looking back basically over only a couple of weeks. The way we look at vol in each market is very reactive to changes over the last couple of weeks in any given market.

That’s another thing I think that differentiates us from other CTAs is that very short lookback window in how we look at volatility in the market to size trades and to decide whether we should be putting positions on in different market environments.

[00:28:22] Alan

You mentioned the average hold periods being about 35 to 45 days, which as you say would put you probably a bit faster than maybe a lot of the medium to long-term trend followers. I think most of the research would indicate that being slower has been better over the last time, or at least being faster, has been more challenging. Have you guys looked at that research? I guess have you settled on 35 to 40 days from a philosophical perspective, or as just what you see as being the most effective looking at the data across all of the markets?

[00:28:58] Brian

Well, the truth is the longest-term trading systems have the highest absolute returns. Putting on a position and letting the trend go as far and as long as possible, when there are big trending markets, there’s no question that the longest-term trend following systems are going to have the best returns. They’re going to have the best returns with more deviation and bigger drawdowns associated with them. In our perspective, we like to have multiple systems that attack trends in a different way.

Our shorter-term systems will get in and out, and back in and out, over a time continuum. That helps improve the Sharpe and/or Sortino numbers of the overall. It’s really how you blend the systems together that determines the return stream. We could put all of our risk in the longest-term systems and we’d have much bigger drawdowns, much higher standard deviation, and probably have our program be less palatable to most investors. It’s really the blend of systems and the average duration will change over time.

Our shortest-term system in a 5-year period might, on average, have a holding period of 18 days, and for the next trading year or two, it might be as low as 12 to 15 days. That number, 35 to 50, will change – not substantially, but it’ll vary from period to period.

[00:30:48] Alan

You mentioned that the very short lookback on vol, the two weeks, what influences your decision making there? Is that coming back to the risk management mindset that if there is a sudden shift in the vol landscape, you want to be cognizant of that? Obviously, I guess if you have just a one-day big move, it can pretty much skew the 10-day measure. Whereas if you had a longer-term measure, you might be less reactive. What drove that decision?

[00:31:25] Brian

That’s something that goes way back to the turtle trading days really. That is if vol is increasing in a market, it could be a one-day thing, but more often than not, if you get a big spike in price, something’s occurred to make that market react that way. If vol is increasing, if we have all systems with a position on and it’s going in our favor, great. If it’s bad vol, then we’re probably being pretty quick to liquidate at least some of those systems.

The purpose of making sure that we’re reactive to changes in vol is for new initiations, because we’re going to be sizing our position on the new volatility in that market rather than on, say, an 80-day lookback where it would smooth out and you’d pretty much be sizing today’s trade the same as you did weeks ago. Obviously, with more vol we’re taking less contracts, we normalize the risk in each system across the market.

If market vol is picking up and we’re initiating a trade, we’re not getting filtered out by the vol filter, we will be taking less contracts and just managing those contracts around the current volatility.

[00:32:45] Niels

Yeah, I want to stay a little bit on this topic because you mentioned early on that you focus on two core concepts in your systems, namely trend following and momentum. Can you talk a little bit about how you defined the difference between the two and how they conceptually are different?

[00:33:06] Brian

Yeah. Like I said, every system is looking to capture directional price movement. For us to diversify the program, we like to diversify in as many ways as possible. In the markets we choose, globally, commodities, financials, and trading systems. The trend following systems are based on looking at confirming the current price over a few discrete timeframes. We’re looking to confirm that over a long-term lookback it’s an up or down trend, and then a medium-term lookback and a short-term lookback.

We may be looking at the current price to 180 days ago, and then to 40 days ago, and then to 15 days ago. If we’re above a certain portion of the range, that confirms for us that a trend is in place and we will be initiating a trade if the filters don’t keep us out. We have a couple of different trend following ones, so it’s really the liquidation criteria then that differentiates the holding periods. One of the trend following systems is quicker to liquidate and then reinitiate the trade, and the other one is designed to hold a position maybe 35 to 45 days, whereas the shorter term one will be targeting more of a 15 to 25-day holding period.

The momentum systems are very different in the way they initiate. They are looking at a time-weighted analysis of the most recent daily data. We’re looking basically at maybe the last two or three weeks of daily data. We’re looking at two things, two different types of momentum systems. One, we’re looking at the closing prices in a market to identify basically that a trend is emerging. The more positive closes you have looking back over various time slices, the more likely you are that a trend is beginning to emerge. That’s one type of momentum system.

The other one is just looking at the speed and the magnitude of price action from day to day. If volatility is picking up and prices are accelerating in one direction or another at a rapid rate, that system tends to kick in. What it does is diversifies for us the points where we’re initiating and liquidating trades in the markets we’re trading. There may be periods where we have only one system that’s either long or short, we have periods where two or three systems may kick in and have a trade on.

If a market is really trending, typically we’ll have all four systems on. But it really diversifies where you get in and where you get out of trades, and each system has an equal risk weighting. There’s no heavier risk weighting to the longer-term system because data shows that you never know from what timeframe to the next which system is going to generate the best returns. It’s pretty random. Overall, having an equal risk weighting ensures that we’re not overdoing it in the system that’s going to hold a position for 100 and 150 days, something like that.

[00:37:21] Niels

Yeah. There were a couple of things I was really intrigued by when I read through some of your stuff in preparing for this. Of course, I don’t want you to share any of the secret sauce, which I’m sure you wouldn’t anyway. You talk about something you’ve developed, like a super value that you’re looking at. Also, you talk about two things that you are looking to improve. One of them stood out to me or something. I said, “Well, what does that really mean?” One of them was the risk-adjusted returns. That’s perfectly fine. I think we understand that. The other one you mentioned is the technical logic of the system. What does that mean and how is that important to your process?

[00:38:09] Brian

Well, each system has a fitness metric or a super value that it’s optimized to. We can change the various performance metrics that go into the algorithms that build the parameter sets for each of those systems. If you change the core logic, say you’re blending a return element with a risk-adjusted element like Sortino, that’ll give you a system that has a certain roadmap in terms of return generation.

Whereas if you decide you want to add maybe another system to the program and reallocate the risk; instead of four systems with 25%, five systems with 20%. Then you take another system and you look at the return element with the concept of efficiency. Meaning if a market trends from point A to point B, how much of that trend do you capture? That gives you a different set of performance data and a definitely different target for that particular system.

That’s what we mean by changing the core logic. It would be what goes into the fitness metrics and how does that change the return profile of that system, number one. Then even if you change the return profile of that system, you still have to blend it with the other systems to see the effect on the portfolio. That’s the change in the core logic, whereas the other changes from period to period are really just how the data affects the system that’s already in place and change the parameter sets. Those are the differences.

[00:40:06] Niels

That was actually just my follow-up, before I hand it back to Alan, in the sense that you have this adaptive process, which frankly, the firm I work for, we also are adaptive in all parameter selection, so to speak. Of course, a lot of people don’t believe in that and stick with the same parameters, and that seems to work as well. How do you see the importance of your methodology? I imagine it’s something that has come later in life, so to speak, this adaptive nature. Correct me if I’m wrong here. Maybe you can describe it a little bit. Also, how does it help in your goal to be reliable, predictive, robust, some of the things that I noticed.

[00:41:00] Brian

In the re-optimization process that all of our systems undergo on a yearly basis; by that, I mean in Q1, we’ll re-optimize a short-term trend following system, in Q2, we’ll look to re-optimize the parameters of the long-term momentum system, and so forth. We stagger the reactivation process of our systems from one quarter to the next so we’re not making wholesale changes to all the systems at one time, say, on January 1st of every year. When we’re looking to re-optimize the systems, we could do it on a shorter time basis than that, but really, there’s not enough data points.

Can you state your question again? I kind of lost track.

[00:41:51] Niels

No, that’s perfectly fine. I was more curious about how you feel about the process. You use a walk forward process, how that helps you achieve your goal, in a sense, if you could just talk about that, again, without getting into the secret sauce. I definitely think parameter picks are important, without a doubt. I also find that the methodology of how people go about changing them, for example, the fact that you mentioned this, that you don’t do it all in one go. I’ve certainly come across people who do it all in January. Those kinds of things seem to be relevant.

[00:42:37] Brian

Yeah. To get back on track, the forward walk out-of-sample testing that we do is very important because when we’re looking at creating systems that are reliable, reproducible, so that our investors can have confidence that we’re going to perform the way they expect during different market environments, that forward walk methodology is very important. We’ll reoptimize a trading system and then we’ll compare at the end of that period, we can compare the actual trading results to the hypothetical out-of-sample.

As long as a trading system is very close to performing in real-time what the out-of-sample track record is showing you’re doing, then you know or at least you have a better degree of confidence that what you’re doing is reliable and reproducible. If you start to see, in real trading, something that’s diverging from the expected out-of-sample returns, then that’s something for the investment committee to discuss.

[00:43:56] Niels

Sure. Absolutely. Alan, where are we going next?

[00:44:00] Alan

Maybe to pick up on one of the points here, just so I’m clear. When you talk about the core logic, is that the overall investment objective? You talked about whether you want absolute return, or risk-adjusted returns, or I guess Sharpe would be in that. Is that the idea and what might prompt you to change that at some point?

[00:44:23] Brian

I guess. When you say you’re a systematic quantitative CTA manager, there’s always those discretionary things that you’re doing in research that determine what makes your systematic quantitative approach maybe a little bit different than other CTAs. Our goal is to have trading systems much like a multi-CTA fund of fund. You hire certain CTAs with certain return expectations and you blend them with other CTAs to build that multi-manager program. The same is true for EMC with multiple systems.

We want to make sure that the roadmap that we have for each system is going to generate returns with positive expectation, but they’re going to differ from system to system so that the overall effect is a certain rate of return based with a risk element and a certain drawdown expectation. It’s really the discretionary things that go on in the background. It’s like what types of systems do we want to blend together to get where we want to go?

[00:45:47] Alan

Good stuff. I guess maybe it’s a good segue into something that we’ve talked with everybody about and had a good bit of debate about, which is the number of markets traded. We’ve heard different views on the value of adding more and more markets. Some managers are now trading hundreds of markets, others trading more than 80 to 100 types of traditional macro markets. What’s your perspective on the merits, and the costs, and benefits of trading lots and lots of markets?

[00:46:18] Brian

Well certainly, there really is no right number of markets in my opinion. We’re currently trading somewhere in the 70 some odd markets. We definitely have quite a few markets that we’re actually looking to add to the portfolio. We’re going to look at a number of different factors. Obviously, first and foremost, are they liquid enough for the AUM that we’re trading? If they are, great. Then they can get a certain risk weighting.

Every market in our portfolio gets its own risk weighting. Anywhere from a fully weighted risk market, which is a 1, down to less liquid markets might have a 0.1 risk weighting in them, and that’s primarily because of their liquidity. We’re still trading markets like lumber, and orange juice, and oats, and things like that where the liquidity in those markets is smaller so we can’t take full position sizes, but we want to keep them in the portfolio because quite honestly, you never know from year to year, what the best market in terms of profitability is going to be.

Over the last several years, some of the biggest trenders have been things like palladium, and orange juice, and lumber, and hogs, and markets that we’re still able to trade nimbly. Those are the ones that add meaningful profit attributions by the end. Liquidity is definitely a consideration. The correlation of the market that you’re going to add to the sector that you’re looking at, so one of the markets we’re looking at right now would be iron ore.

We would put that in the base metal category or sector of our portfolio. We’d have to look at how much risk we want to keep allocated in the base metals and maybe reduce a little bit of the risk in the other markets that we’re trading, nickel, copper, zinc, aluminum, to add that market. You’d have to adjust the market weightings a little bit. If it really does have non-correlation and is a market that’s dissimilar to everything else that you’re trading, then it’s a good addition to the portfolio.

Another thing that we look at when we’re determining how to add markets or how to weight markets in the portfolio is that particular market’s return stream when we apply the trading systems that we use in that market. Historical performance, there’s some markets that are more difficult to trade successfully than other markets. They move from point A to point B in a much choppier fashion. That’s another reason why we have a blend of systems, because if there’s more choppy markets, we want to be getting risk in and out, and back in and out of those markets as they chop along.

We look at the return expectation for that market. Some markets just trend better than others. They have more linear type trends, less hiccups along the way. A market that has that kind of return expectation, we can give a little higher weighting to. We’re always looking to add markets that would improve the diversification of the portfolio and the return expectations of each program.

[00:50:12] Alan

Niels, I know it’s a favorite topic of yours.

[00:50:17] Niels

Well, since you took one of mine, which was the number of markets, I’m going to take risk management, Alan. No, I’m just curious because I think it is an important topic that we want to talk to you about as well, and that is just, how do you think of risk and what are some of the key ways of you managing the risk inside the strategies?

[00:50:42] Brian

Yeah. Well, as I mentioned, at the outset, risk is probably the most important thing that a CTA does. You can have a marginally good blend of trading systems or one trading system and if you manage risk properly, you can generate positive returns. For us, we manage risk really at every level of the investment program. I guess at the start with our trading systems, as I mentioned, each system has a max loss for each trade in every market that’s automatically generated once we initiate the trade. Different systems have different max loss thresholds.

Some systems, as I mentioned, are designed to accept a little more volatility, a little bit bigger drawdowns. At the end of the day, every system has a max loss associated with it. Every market has a limited amount of risk that it can take at initiation. The more aggressive, higher risk-weighted markets in, say, our traditional trend following program, willl risk maybe half of 1% per trade.

Every system is going to have a little bit of a different risk associated with it. Obviously, then we’re looking at volatility of the markets with what I said was a short lookback window, the open high, low close and the true range over the last two weeks, and volatility then gets built into building new initiation. As vol is increasing, smaller position sizes in more consolidated periods will take a bigger number of contracts when we’re initiating a trade.

Then obviously, by limiting the amount of risk that can be taken in any one market or one sector. When you look at max risk in the portfolio and you take the max risk in the markets within the energy sector; so we’re currently trading six global energy markets. When you look at the risk weighting for each of those markets, and you add them together, and you divide it by the total risk that could be allocated across the whole portfolio, you’ll get a percentage of risk that can be allocated to that particular sector. In the energy sector, it happens to be like 10% or 11%, in our trend following. That’s the max risk at initiation.

Then again, as I mentioned, you might get a period where, say, when you’re long the energy markets in anticipation of conflict or war breaking out, and then vol jumps 25% or 30%, well, the risk in all of your current positions has increased by that much too. The overall risk in the program at that point might be 50% more than it was before, in terms of the total risk from day to day. That’s good vol, that’s driving good returns.

If we reach the portfolio threshold that I mentioned where open trade equity is growing, profitability is high, then the scaling mechanism will start kicking in and getting those positions back in line to mirror what the current risk in that market is. I guess that’s one of the big departures is this concept of utility and marginal utility. Utility would be the satisfaction that an investor gets from going from up 5% to up 10%.

Well, utility is going up for the investor every step of the way. Say going from 15% to 20%, utility is still going up but the marginal utility or the amount of satisfaction he’s getting from going from 15% to 20% is decreasing each percentage along the way. He’s much more interested in preserving those gains and not experiencing the big drawdown that could occur with a trend reversal and locking in that profitability. That’s really the concept that overlays the risk management at the portfolio level is how far along the way do we want trends to go before we start scaling back.

It can be gradual, it can be maybe 2%, or 3% or 4% of all the positions in the portfolio on a day-to-day basis. If we get a big spike where, say, a program is up 3% or 4% on the day, we could get a significantly bigger scaling effect. In the most aggressive scenario, I guess, if you’re up 10%, 15%, 20%, 25%, we can cut back the positions in our more traditional trend following program up to 60%, 70%, 75% of the original position. It can be meaningful. It’s been very helpful. We started doing this risk management scaling in the late 2000s, in 2008 really. It definitely has its pros and cons.

By implementing it, you’re sacrificing some upside potential. In real explosive periods, we might underperform some of our peers. We’ll still be putting up returns that are comparable, but we might be underperforming. Conversely then, when the reversal periods come, we might be outperforming because of this mechanism. That’s, I guess, one of the things that you asked earlier, what differentiates us. I would say that’s definitely one of them.

[00:57:00] Niels

Sure. It’s very interesting, because as you know, we often talk about how trend following encapsulates human behavior in general. But here, you’ve taken sort of one specific bias, human bias, into a little bit more of the forefront. Very interesting indeed, Brian. Thanks for sharing that.

Alan, do you want to do one more round of topics and then we’ll slowly wrap up?

[00:57:27] Alan

It’s useful to get your perspective on markets generally, liquidity in particular, but even more broadly than that. Obviously, you’ve been involved in markets since the 1980s, been on the floor. Now everything’s gone electronic. Obviously, some things don’t change. In some senses, trend following is still working. Obviously, other things are changing. What do you think is different now, and specifically, how do you think the liquidity of markets is maybe relative to what it was going back over time?

[00:58:01] Brian

Well, I think markets are more liquid than they used to be. I think the real difference is the flow of information is certainly much quicker than it certainly was in the ’80s and ’90s, where unless you were on the floor, the information flows that you would see, just from the various trading desks and the hand gestures into the pits, and what big firms were doing what, that was your inside information when you were on the floor.

In today’s environment, obviously with everything electronic, there’s a lot more liquidity. Markets tend to react to news much quicker than they used to, for better, for worse, sometimes for worse. An unexpected news story can come out and a lot of CTAs and hedge funds are positioned the wrong way, you’re going to have a cascade of liquidation orders pushing the trade against you.

I guess that’s another reason why we prefer to have a blend of systems, some that are shorter term in nature so that when those market moves occur where we have nice trends in place and then there’s something that causes a steep reversal in the trend, we want to be pretty nimble in liquidating at least part of our position, because if we waited for longer timeframes the drawdown would be bigger. We’ve learned, over time, that if a lot of market participants are positioned the wrong way. Those liquidations tend to continue over longer and longer periods of time and it just gets more painful.

[00:59:58] Niels

Yeah, absolutely. I completely acknowledge that pain. Now, another thing a little bit maybe related to pain here, Brian, and that is we always want to hear from our guests, what’s the one thing that they hear about trend following that they really disagree with? I’m going to ask you the same, if there’s one particular thing that you hear people say about trend following that you think, and maybe for a while, that you completely disagree with.

[01:00:34] Brian

Well, I think the more astute the investor is, the better. There’s times when I’ve heard people say, “All you guys are doing the same thing. I could just throw a dart and pick one of you, and it wouldn’t make any difference.” I think that’s not correct. Trend following, I’ve heard it multiple times throughout the last 30-some-odd years, that trend following is dead, trend following really doesn’t work. If I want to do that, I can just flip on a moving average and be either long or short based on whether it’s above or below moving average.

All those kinds of things have maybe a grain of truth in them, but for the most part it’s frustrating, especially when you know that long-term positive returns definitely add diversification and benefits to a portfolio. People that just say you’re too volatile for me or too volatile for our portfolio, I don’t know, I don’t agree with that.

[01:01:56] Niels

That’s fair enough. It is quite interesting actually, Brian, when I hear you talk. Obviously, you’ve explained a lot of things today and so have all the other guests in this series. It is quite interesting when you hear how much work, how much testing, how many ideas get tried before we find the things we believe work. Of course, they’re all a little bit different as well. I hope that makes it pretty clear to people that what we do is certainly not simple or easy in any way, shape, or form.

[01:02:36] Brian

I was going to say, it’s funny, I was just thinking of all the testing and all the things that we tried to do over the years to devise and develop a counter-trend system that would counteract say if we had a lot of long positions, a lot of risk exposure in the portfolio, come up with a system that would be a counter-trend system that would help dampen the risk and capture those reversals. It’s very, very difficult to find a trend reversal system at least, for us, that worked.

What that eventually evolved into was this algorithm that scales back positions rather than allocate more of the capital and margin money to putting on short positions. It is a process that the light bulb goes on, and you say, “Instead of this, let’s try this.” It is definitely an ongoing process and you can’t… (I shouldn’t say you can’t rest on your laurels). Some traders don’t change what they’ve done and have remained successful, but we like to look at how to improve things from period to period.

[01:03:54] Niels

Yeah, absolutely. The last thing I just wanted to ask you, Brian, was really, obviously, as I stated in my introduction, in 2022 trend following was really the one strategy that stood out. Looking now, we’re kind of almost halfway through 2023. What are you getting excited about, so to speak? It could be whether you’ve had more productive conversations in the first few months of this year after last year, or whatever it might be. What are you excited about as you look into the rest of the year and are there any concerns on your radar for us as an industry?

[01:04:36] Brian

Yeah, it’s interesting if I’m excited or a little nervous. It’s always interesting because obviously nobody knows what’s coming next, whether it’s a resurgence of inflation over the next 6 to 12 months to 18 months or whether it’s some issue with debt around the world. It’s definitely an environment where you want to have something in your portfolio that’s trading some of these unique markets, trading in commodity markets, trading interest rates, the ability to get short equities.

As I look into 2023, there’s a lot of uncertainty and that’s why I’m glad that we’re not discretionary traders because that would really be difficult. I like the quantitative approach that we take. It helps us sleep at night. We know, looking at stress testing, what the potential outlier risk is. When we look at our daily internal VaR that we’re calculating for each market and each portfolio on a day-to-day basis, actually, right now, it’s fairly high historically. We do have a lot of positions on in various sectors.

We are seeing trends and we’re definitely seeing increases in volatility across a lot of different markets and sectors. We’re just glad that for the investor and the investor base that we have, that people are starting to look at systematic commodity trading strategies a little bit more to add to the portfolio. We’re optimistic, but we’ve always been cautiously optimistic from period to period.

[01:06:29] Niels

Yeah, that makes perfect sense. Brian, this was really a wonderful conversation. Great to have you back on the podcast. Thank you so much for sharing your thoughts and your insights with us. We hope of course that we can do that some time again in the future. To all of you listening in 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 the website. Of course, not least, take care of yourself and take care of each other.

[01:07:16] 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.