My guest is Eric Crittenden, founder and Chief Investment Officer of Standpoint Funds.
Eric has spent his career with trend following strategies, first at BlackStar where he managed a fund-of-funds, then at Longboard, and now at Standpoint Funds. This background makes him not only a fountain of knowledge on trend following theory, but also the operational logistics and practical considerations.
In this episode our conversation ranges from the source of the trend-following premium to novel concepts for stress-testing managed futures programs. We discuss the struggles the space has faced, the evolution of CTAs, how to think about dispersion among managers, and how Eric thinks about solving for client behavior.
I hope you enjoy my conversation with Eric Crittenden.
Corey Hoffstein 00:00
321 Hello and welcome everyone. I’m Corey Hoffstein. And this is flirting with models the podcast that pulls back the curtain to discover the human factor behind the quantitative strategy.
Corey Hoffstein Is the co founder and chief investment officer of new found research due to industry regulations he will not discuss any of new found researches funds on this podcast all opinions expressed by podcast participants are solely their own opinion and do not reflect the opinion of new found research. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of newfound research may maintain positions in securities discussed in this podcast for more information is it think newfound.com.
Corey Hoffstein 00:50
My guest is Eric Cruttenden. Founder and Chief Investment Officer of standpoint funds. Eric has spent his career with trend following strategies, first at Blackstar, where he managed to fund to funds than at long board and now it standpoint funds. This background makes him not only a fountain of knowledge on trend following theory, but also the operational logistics and practical considerations. In this episode, our conversation ranges from the source of trend following premium to novel concepts for stress testing managed futures programs, we discuss the struggles, the spaces faced the evolution of CTAs, how to think about dispersion among managers, and how Eric thinks about solving for client behavior. I hope you enjoy my conversation with Eric Cruttenden. Eric, welcome to the show. excited to have you here to talk all things trend following which I know you are a long term practitioner and students are getting your perspective is something I am certainly looking forward to hearing. Thanks, Cory. Glad to be here. So on that front, where I really want to start knowing you have been a long term student of trend following Can we start with maybe what were some of your seminal and formative experiences that really got you into trend following?
Eric Crittenden 02:10
Yeah, there’s two that come to mind. And they took place in college. And they’re a little embarrassing, actually. I think it was my sophomore year. at Wichita State, one of the projects that I had to do was to build a mechanical trading system and run it for a semester. So being a novice and not understanding things like survivorship bias, or even dividends and corporate actions and whatnot. I, in my infinite wisdom, decided to build a system that bought 52 week lows, and sold 52 week highs, because like everyone else, and I looked at the charts and saw the inflection points and saw how much money you can make by buying low and selling high. And thought, wow, this is going to be easy. So I did that I built a system that essentially bought low and sold high and the back test was amazing. It was incredible. I was already pricing out my own private island and platinum covered helicopter. And I thought well, I’ve got this figured out, right. But an interesting thing happened when I finished the system and then went live not with money. It was a paper trading account that we had to run for the semester. The system actually lost money. I think this was 9097. I think it was in the fall of 97. So I took a pretty big ego hit at the time. And I thought what did I overlook? What did I do wrong? So I reran the back test. And amazingly, the back test was making money. So the back test is making money. But the system’s losing money, yet they’re highly correlated. So this was my first introduction to the concept of survivorship bias. So I pulled down my statements, and I started going through it line by line and looking at it and I realized that the back test was done on surviving instruments only. So every time he bought a 52 week low in a stock that was going to be alive, you know, 10 years later or whatnot, you’re sidestepping, and avoiding stocks that went bankrupt. So I learned two things from this, both equally important one is survivorship bias is a huge deal, at least for stock traders, it’s less of an issue for futures. And that was a painful lesson. But at least I got to do it with paper money rather than real money. And it took a while to figure it out. Because you really have to understand databases and how referential integrity is enforced in them. And it takes a while. I mean, I’m talking about it now as if it took a week, but it took a semester and a half to kind of unravel this and see survivorship bias for what it is. But the second thing that I learned that might be more important is how much nobody cares. Because when I started pointing this out to my peers, professors and other people, they just were not impressed. And they’re like, well, it doesn’t matter. They didn’t get it. It wasn’t important to them. And I thought wow, this is a that was a seminal moment for me, because I realized at the time that the entire industry is plagued with this practice of essentially curve fitting to a cherry picked version of history right rather than history as it actually unfolded. So that was one thing that at least got me to stop being counter trend at that frequency anyways, the second was the next semester, a project in a different class, I think this was a futures and derivatives class was to create an efficient portfolio using modern portfolio theory. So, our job was to write the code. I think it was Fortran, back then, collect the data, write the code, create the efficient frontier, the security market line and whatnot, and then select an optimal portfolio and submit it for that was for your project for your final. So I lived in Wichita, Kansas at the time, and you probably don’t know anything about Kansas that people they’re awesome, but the place is boring as heck. There’s nothing to do. After 8pm The weather’s atrocious, It’s scorching hot in the summer and windy and freezing cold in the winter. It’s a really boring place, but they do a lot of business there and agribusiness, Koch Industries is there Cargill and whatnot. So there is some quant finance talent, mostly working in hedging, trading derivatives, corporate farming, stuff like that. So at Wichita State, I started a club. That was there was no phrase for there was no computational finance degree back then. But the members of this club were focusing on kind of the intersection between computer science and finance. So I made a lot of friends that worked in these various hedging departments at big companies around Southeastern Kansas. And I got to use their Bloomberg terminals, their Reuters terminals to collect data. So for my project, I used a more comprehensive set of data than my peers, my fellow students. And I stumbled across something it wasn’t called Managed futures back then, I think it was systematic global macro, but included Salem, Abraham and Chesapeake in some of the old school, classic CTAs. And I didn’t know any better. So I had stocks, I had bonds, I had real estate, I had CTAs, I had long short equity hedge funds in my dataset. So when I created an efficient frontier, no matter which way I looked at it said put 30 to 50% in this in the CTA, or global macro category. So I did that. And I ended up with really nice Sharpe ratio, a really nice Sortino ratio, high compounded rate of return and very low vol compared to other portfolios. But an interesting thing happened when I turned this in. My professor said these results are unrealistic. What did you made a mistake somewhere? What did you do? And when we unpacked and looked at it, and he saw that there were these other asset classes in there, and he said, we’ll just get rid of those and resubmit it. Why are we getting rid of them, because when I kick them out, the returns deteriorate significantly, and the volatility goes way up, and the drawdowns get twice, maybe three times as large. And he said, no one invests in those things, those aren’t legitimate asset classes. It’s just stocks, bonds, and real estate. So I argued with him a bit and finally just gave up and moved on. And again, I went to my peers and other people and showed them the results. And it was total apathy, no one cared, which was interesting to me, you probably picking up on the fact that I’m an independent person, a little bit of a contrarian. And I have a saying that trend following is the most contouring thing you will ever do, at least in real time. So anyways, that’s how I got onto this trend following path. I didn’t like the idea of trend following the idea of just crowding into something because it’s going up or shorting something, because it’s going down, didn’t appeal to me at the time, until I realized that that’s about the most contrarian thing you’re ever going to do when you’re managing other people’s money. So that’s how I got started. And then in the early 2000s, I started allocating to CTAs as a part of my job, and kind of that became what it is today where I am a CTA. So natural progression,
Corey Hoffstein 08:47
and that is the bit of the progression I wanted to step towards. I know after finishing up your studies, you went and worked with the financial advisor for a bit, went and worked at a family office, but did eventually start your own firm Blackstar where you were running what was essentially a fund of funds allocating to firms like Winton and trance trend, Milburn, some hedged equity, which is a little bit different, obviously, than running your own managed futures program, which you have eventually come to move towards which we’ll get to but where I wanted to start was getting an idea from you as to what were some of the lessons learned managing a fund of funds, especially when it comes to due diligence on trend managers?
Eric Crittenden 09:27
Yeah, I would say that the biggest takeaway from running a fund of funds that focused primarily on CTAs was that the surviving CTAs the ones that ended up doing well, all pretty much did the same thing. But they won’t admit to that. They don’t like this argument, but there’s some strong common denominators amongst the winners. And then there’s some relatively obvious common denominators amongst the losers. So I guess the lesson I learned is don’t listen to the marketing department. Look under the hood. And what they’re doing decompose their returns, it’s not that hard to figure out what they’re doing. And that’s valuable for in two ways. One is trying to figure out why this works and why it might be sustainable, and to looking for the kind of signal to skip on an allocation to someone. So let me elaborate on that a little bit. The common denominator that I’ve found amongst CTAs, where it didn’t work out was they used a lot of filters, they’re always trying to solve last cycles problem with a lot of filters and don’t have enough respect for the degrees of freedom issues that they’re creating. By using these filters, that’s a valuable lesson to learn. So and I know you have some strong opinions on filters as well,
Corey Hoffstein 10:42
I have just a few just fueled by filters, you mean adding more bells and whistles? So it’s not just buying when the signal turns off? It’s a confluence of different indicators that all have to sort of be conditionally on at the same time,
Eric Crittenden 10:54
yes, and why those filters are in there, cuz not all filters are bad, but most of them are. And the common denominator there is did they come up with a filter after experiencing some painful run of volatility or drawdown that would have worked to solve the problem at that moment in time, but it’s really just a spurious thing that they’re including in the system without understanding the unintended nonlinear consequences of having that filter in the system going forward? And this kind of lack of curiosity as to why that filter would give you a better chance of surviving long term that was the common denominator I saw is this lack of curiosity about why what we do works? And is this filter actually a creative to creating a more symbiotic relationship between us in the marketplace? Or is it just a hack that would to solve the problem that we just went through?
Corey Hoffstein 11:45
So sticking with this idea of trying to create a system that’s robust data sample, back in 2005, while at Blackstar, you published a paper that’s still pretty well known in the trend following space, which is actually all about trend following on individual stocks, which doesn’t really get as much interest or play, at least in the academic literature. It’s now been 15 years since you published that paper. And I wanted to ask you, what have we learned out a sample,
Eric Crittenden 12:12
I can’t believe it’s been 15 years, that’s a long time, I felt like I wrote that three or four years ago. So and for anyone who’s read the paper, you can tell I’m not much of an academic, but I feel like there was some rigor, a lot of effort went into being respectful of data integrity issues into that paper, so but I’m definitely not an academic. So there’s no Greeks in the paper, out of sample, I feel like that strategy did exactly what I expected it to out of sample. But I’ll share this with you. It was incredibly uncomfortable. It’s an incredibly uncomfortable strategy for clients, in my experience, doing the analysis, and identifying risk premium, that are sustainable, that can be done historically, you can ferret that out of the data historically. But it’s another thing entirely to convince other people to sign up for that put their money on the line and go through it step by step day by day. So what I found in practice, managing other people’s money, according to that strategy, is that they’re most uncomfortable with the strategy when they should be comfortable with it, and vice versa. So those risk premia that data scientists can tease out of the data look great in a back test. They look great on paper, but they exist for a reason, because they’re incredibly uncomfortable to harvest in real time. So those are the two things I learned one that the concepts I presented in that paper, I think are very valid. And I’ve been pleased with the out of sample performance of it. But to it doesn’t make you very popular with your clients.
Corey Hoffstein 13:52
And to that point, it’s actually isn’t an approach you run anymore. Right on individual equities. It was one you ran historically, but no longer maintain that strategy.
Eric Crittenden 14:02
Yeah, correct. That strategy belongs to my previous firm, and I believe they still run it successfully. I’ve moved on my current approach in business is around what I call all weather investing, which I think is better suited to my personality. And it’s just something that I’m very enthusiastic about doing for the next 2530 years, all weather investing. So let me break down what I mean by all weather investing, I’m talking about including as many global risk premia as I can, in under one hood, when you’re building a portfolio of different risk premia, how they interact together matters as much as how good they are on a standalone basis. So if you ask me, Eric, do you believe in some of these long short and or hedged equity strategies are superior to buy and hold? My answer would be yes, on a standalone basis. I think that they’re superior. I think they have higher returns, lower volatility, lower drawdown and are pretty durable. However, in the context of a diversified portfolio where you’re mixing managed futures, certain interest rates and equities together, things change. It turns out that the risk premiums that you are harvesting in a lot of these long short equity and hedged equity programs are somewhat redundant with the same risk premiums that trend followers are collecting in the futures markets. In other words, every strategy has a blind spot. And if you’re building a diversified portfolio in an all weather portfolio, you don’t want your different strategies to share the same blind spot. So what I found is that buy and hold global equity investing does not share the same blind spot as your typical kind of plain vanilla managed futures trend following program. And when you blend them together, in my opinion, that is a more durable, more tax efficient, more scalable, and more stable all weather approach than blending hedged equity or long short equity programs with managed futures. Sorry, as a long winded way of saying that buy and hold simply blends better with managed futures.
Corey Hoffstein 16:10
To that point, going from sort of 30,000 feet into the weeds a little bit, one of the things I always think about and have a lot of conversations with folks about is knowing you will be combining certain strategies, let’s say we know, we’re going to be combining a Managed futures program with a buy and hold Equity Program. Does that change the nature of how we would think about designing the Managed futures program? For example, might we consider not having as much of an equity sector allocation in the Managed futures program, knowing we’re going to have that buy and hold exposure that it’s being complemented with?
Eric Crittenden 16:47
Yeah, I think that’s responsible thinking. It’s dangerous, because now you’re kind of messing with the original program, but it can be done safely. And I went through that process myself building my own program, and that, you know, it’s algebra, it’s an exposure issue, how much heat do you want on the upside? If you know, you have your long only equities? Obviously, that’s always turned on. And then the futures programs trying to go 200% long equities? Well, then you could be 300%. And that long, you can’t do that, not in my world. So I would say that, yeah, it’s your responsibility to evaluate the degree to which you want to dilute the equity exposure. In your futures program. It’s just the reciprocal to the other decision. Like if you’re going to add in, buy and hold equities, you have to consider what’s in your futures program. I mean, the end of the day, we’re chefs trying to make a meal, and you don’t want too much hung in in the meal. So
Corey Hoffstein 17:40
to take a step back, I want to go back to sort of just core philosophy here after 2008 Managed futures got a lot of interest, especially for this idea of being able to provide crisis alpha, as you see it, what is really the main benefit of managed futures?
Eric Crittenden 17:58
Well, for the record, I’m not a huge fan of the crisis Alpha concept, I recognize that managed futures has historically produced positive returns when you need them the most when equity markets are losing money and structurally challenged. But you know that the path traveled in the permutation risk can work differently, meaning that I can easily create a scenario where managed futures does not make a bunch of money when equities are going down. And I don’t want my co workers and clients to be caught off guard and misled in that thinking of managed futures is a very, very reliable hedge. There are situations where it will not be a very reliable hedge. That being said, fundamentally managed futures shouldn’t be expected to be meaningfully correlated with equities for an extended period of time. So from that perspective, when you you have two assets, let’s say they both have a 10% expected return, but their correlation is zero. When you blend them together, you can get a return that’s actually higher than 10. And you can get volatility that’s 1/3 to maybe a half lower, depending upon the correlation and much smaller drawdown. So I tell people, Look, if you want to build a diversified portfolio, bring your uncorrelated assets in but don’t expect them all to offset each other all the time, the ideas that win the marathon, we’re not going to win every sprint
Corey Hoffstein 19:27
after Blackstar. You went on to co found long board or you transitioned from a hedge fund structure to a 40x structure. And one of the arguments I often hear out there is running a Managed futures program is incredibly difficult under 40 Act law. I wanted to get your perspective operationally, get some ideas from me. Were there any surprising lessons learned and trying to take the Managed futures concept and really package it into a mutual fund vehicle?
Eric Crittenden 19:57
So that statement that it’s difficult ought to run a Managed futures program in a traditional 40 active mutual fund is both true and false. It’s true, in my opinion, if you’re doing it in a fund of funds format, where you’re trying to allocate out to see TAs and have them do all of the trading, it gets very complicated. And you know, they tried using swaps for a long time, they tried using owned LPS doing things offshore. It’s just not consistent with the rules and regulations enough to make it simple when you’re running it as a fund to fund so they really had to do a lot of dancing to make that work. That’s I believe what they mean when they say it’s very difficult to run a Managed futures program in a mutual fund. That is not true. If you are the CTA, yourself, and you launch a mutual fund, it is not difficult at all, to stay within the lines of the regulations, and trade these futures contracts and even forward contracts. It’s not difficult at all, in my opinion. I’ve been doing it for almost a decade. And I don’t think that it’s meaningfully more complex than just running a regular CTA. So I just want to distinguish between those two things. What was your other question, operational challenges,
Corey Hoffstein 21:11
other operational challenges are interesting lessons learned.
Eric Crittenden 21:14
So the biggest challenge, in my opinion is the people that work in the mutual fund industry are very, very comfortable with common stocks, you know, closed end funds, even things like warrants, and stuff like that. But when it comes to futures and forward contracts, it’s new to them. So no mutual fund has a board of directors, most of these are in shared trusts, you’ve got Chief Compliance Officer at the trust, you’ve got your own Chief Compliance Officer and a lot of these people, they have to deal with the SEC, you’re forcing them to deal with the NFA as well. So you really have to take a leadership role with respect to education, and making sure that they understand what you mean when you say risk. What you’re showing when you measure the risk. They watch Trading Places, they’ve heard the stories about semis showing up unloading cattle onto your lawn. None of that’s true, it’s not real, but still you have to. So there’s an educational headwind, you have to get through. And it’s easier to do if you can establish trust and credibility, but there’s a workload associated with it. And I can understand why some of my quant friends don’t want to go anywhere near that kind of workload. But I don’t mind it. Other than that, I would say that the operational challenges is no more complicated, at least for me than running an equity mutual fund
Corey Hoffstein 22:27
moving from the operational to the more theoretical trend following and managed futures more specifically really struggled over the last decade from a relative performance perspective. And in some cases, in an absolute performance sense. There have been some people who have argued that the space has become crowded and that the approach no longer works. Obviously, that’s something you don’t believe given that you just launched a Managed futures fund. So maybe we can take a step back and just ask, what is your thesis as to why CTA type strategies work in general and why do you believe that they’re going to continue to work big topics,
Eric Crittenden 23:06
great topics, though, why they work in general, I believe that medium and long term trend following is a way to consistently provide liquidity to hedgers in the global derivatives markets. And that hedgers are the only group of people that are both willing and able to lose money on their derivatives positions. It’s a zero sum game, if you think about it’s actually a negative sum if you factor in transaction costs and NFA fees and whatnot. So trend following is just kind of a naive, uninformed way of providing liquidity to the one group of people that aren’t profit seeking in the futures markets. And I love that concept. Because I think in any business, you need to try to understand where your profit margin comes from and why it’s sustainable through time.
Corey Hoffstein 23:54
Can you explain why trend following would be the liquidity provider in this case?
Eric Crittenden 23:59
Sure. So in my experience, hedgers, let me define hedgers. hedger is anyone that is participating in a way where they’re not seeking to profit, so it’s someone who’s taking profits. So you can have hedgers in Intel stock where it’s going up and they’re up 40%. And they take profits while marginally they’re not seeking a profit on those shares anymore. In the futures world, a corn farmer would sell corn futures to lock in that price, which would give him or her the certainty that they need in order to do their own budgeting. So, you have to think about when are hedgers incentivized to sell or buy? Well, in the two examples, I just used both of those the corn farmer in the person who owns Intel are incentivized to sell just because the price has gone up. That’s the reason when the price goes up, it creates a situation ation on their balance sheet where they look at it and say, I’m not profit seeking beyond this point, the benefits that I get from locking in the price right now exceed the potential marginal benefits of holding on further. We call those short hedgers there are people that sell to hedge, but they need someone else to buy from them. Now who in their right mind just sits on their couch and waits for the market to be up a lot and then say, Okay, now I’ll buy. It’s not comfortable. It’s the concept we were talking about earlier, where what works to harvest these risk premiums is psychologically and socially uncomfortable, at least in real time, the back test, no problem, but in real time problem. So that’s a short hedge or on long hedge or someone who buys a declining market. So that would be someone who’s a consumer of commodities and airline that’s buying jet fuel, a baker that’s buying wheat or flour or whatnot, anyone who looks at it and says, Wow, my input costs have gone way down, I can expand production, I can hire more people, I can dust off some of the equipment, but that’s expensive and time consuming. And that can all go away if the price goes back up. So my profit margin is improved because the price went down but only if I lock it in. So they start buying as the price goes down, the more the price goes down, the more they buy and then their peers start to buy. So anyways, my point here is hedgers fade trends. Not always, but on $1. Weighted basis by and large over time, if you trade in the direction of the trend, you will be trading opposite hedgers at least on $1 weighted basis, and I believe that is the source of the risk premium that trend followers collect.
Corey Hoffstein 26:40
So as the current risk that trend followers have come now to outsize the producers and the consumers operating in this market and crowded out the trade. And there’s so much ample liquidity, that it’s not really a trade that can necessarily be profitable going forward.
Eric Crittenden 26:56
It’s possible. That’s something I think about all the time. But I don’t see trend following as being meaningfully different from any kind of business profit margins fluctuate. And as an industry matures, and scales and becomes more liquid. Those really high profit margins of the past become Fiction and Fantasy. So I always expected trend following margins to come down to some sustainable level and not just sit there. But that’s going to be the mean going forward. And I think I sincerely believe that that is what has happened. But it’s also corresponded with the risk free rate of return going from 6% in the late 90s, down to almost zero. And what a lot of people don’t realize is that when you trade futures, you get to collect the risk free rate of return on 90 in some cases 100% of your money. So getting 5% A year is great. When that goes away, your compounded return is quite a bit lower. So those two things together have conspired to make trend following returns look lower than what I think their long term average is likely to be.
Corey Hoffstein 28:04
With some of the performance struggles, there’s been a real trend among CTAs. To de emphasize price trend in their models. Some are starting to incorporate signals like short term mean reversion, carry sentiment, sing signals, others are starting to introduce ideas like machine learning as risk overlays, or even to discover different types of signals. And then some of just sort of abandoned trend following and started moving towards a more multi strategy macro approach. What’s your take on the evolution of CTAs?
Eric Crittenden 28:36
I think it’s a natural phenomenon. And the jury’s not out on a lot of these things. I have some really strong opinions about those topics you just brought up. But overall, I don’t pass judgment on people that want to evaluate and improve. I went to school for machine learning. I’ve studied a high, I chose not to use it. And I’m happy to tell you why. I just want to make clear that those tools are powerful, and they’re valuable, but they’re only as good as your use of them. You can use them to destroy yourself or you can use them to tease out valuable information is yet to be determined how they’re actually being used. I think it’s exaggerated. You know, my personal opinion is that the risk premia in the market, they are what they are, you can’t create new ones. There is no alchemy here. A bunch of data scientist can’t create a bunch of new models and push them into the market and everyone starts making more money. The hedgers can only lose what they’re going to lose. You can’t it’s not alchemy, you can’t create returns out of nowhere. Can you do a better job than the next guy? Sure. And is satellite imagery and neural networks and these things? Are they going to give someone an edge? Maybe I’m not convinced though. So it’s kind of the filter thing we talked about earlier where there’s a good chance that 80% of these new tools that people are using are really just cover to solve last cycles problem, and that there’s going to be a letdown and a rude awakening going forward for some of them. That being said, I do see value in them. But it’s as tools, you know, just like a wheelbarrow, or a pickaxe or whatnot, computers, algorithms, these things are tools, how you use the matters more.
Corey Hoffstein 30:23
So let’s talk a little bit about the machine learning side. Because I do know that prior to the launch of your most recent fund that you did a very long deep dive into a deep research project into building your own managed futures model. As you mentioned, you spent a long time studying machine learning and have elected to not use it. A lot of people see machine learning as sort of the future evolution of quant almost becoming table stakes over the next couple of years. What made you steer away from introducing machine learning in your process?
Eric Crittenden 30:57
It’s a great question. So I guess that’s my seminal moment because clearly I’ve taken a stand on this issue. And I’m willing to put my name behind it and manage money for the next 25 years. So here’s what I did, I took two years off, I went back to school, and refreshed my memory and education. The concepts that they’re teaching today are the same ones that that I learned in the 90s. They have different names, a couple of them are slightly new, but 80 to 90% of the coursework, it’s the same fundamental topics in data science and AI. The difference is the computing resources exist today and the data is out there, there’s actually data that you can use. So now it’s become a fad. And it’s become very popular. So I took two years off, got clarity of mind, and I sat down and gave myself a mandate, Eric build something that you would be willing to put 100% of your own investable money in and just leave it alone for the next 30 years. That was my mandate. I hit that with everything that I had, I talked to people that I respect and trust, I got their opinion, talk to my mentors. And I spent almost a year building the most elaborate, sophisticated managed futures program that I possibly could. And I loved this thing. And I thought that the results looked good. They looked good to me. Not fantastic, but really good. And then I thought well, the next thing I’m going to do is I’m going to build the simplest, easiest to understand no moving parts, durable benchmark, just a simple thing as simple as the s&p 500. And just use that as my benchmark, I wanted to see what that looked like. So I did that, then I compared the two. And there was almost no difference. If I put the charts up and you stood 10 feet away, you wouldn’t be able to tell me which one was which. So I had to ask myself the question, well, what do I trust going forward, and the thing that’s got three moving parts, that’s worked for 50 years, that you just can’t I can’t kill it. No matter how many Monte Carlo simulations I do, no matter how much I stress, test the parameters, the variables. And I’ll tell you in a bit about all the stuff that I do to try to break a system. This thing’s just a cockroach is a survivor. And then I compare it to this thing that’s got 16 moving parts and all these levers, and it’s elegant, and it’s sophisticated, and it feeds my ego, which 1am, I going to trust going forward, I’m going to trust the 1940s Jeep, you can submerge that thing in water, you can set it on fire, you can shoot it with a 50 Cal, it’s still gonna start and drive off. So I’ve learned that check your ego, keep it simple, keep it durable, keep it robust. So I made my decision that we’re going to see who ends up being right going forward. But I’m very comfortable with the decision I made. And that’s my answer. Does that make sense to you? To that
Corey Hoffstein 33:51
point? It sort of leads me to another question. There is this ability, there’s so many degrees of freedom in the Managed futures space that you can actually find among managed futures, you know, same category, there’s massive performance dispersion, you’ve got the different trend model specifications, the different contracts, people are trading, how you wait, your exposure, the different volatility targets, and all of these things sort of work together to make it such that you can have probably more dispersion in the category than almost any other category out there. With that in mind, how do we how does an investor really create confidence in the asset class, given this dispersion?
Eric Crittenden 34:34
Yeah, you’re right about that. It’s the dispersion in the Manage features category has been the source of tremendous anxiety in my world, and I’ve seen people and they agonize over this, picking a manager and having the category be up 10 But this manager is down 15 and having to defend that and vice versa. I would push back a little bit on the one statement you said though, about this category being the worst offender when it comes to dispersion. If we look at like say small cap stocks, you could have stocks that are up 80% for the year and a whole bunch that are down 50 or 60. How do they deal with it? Well, they diversify. That’s one approach is you can diversify, so that you’re not going to be so unlucky, that you get the bottom decile performers, you can do the same thing in the Managed futures space, you can pick multiple managers. That’s one way. And I’ve had a lot of clients over the years telling me that that’s why they have three managed futures managers in their portfolio. But the managers themselves can do something about it if they want. So, when I look at the Managed futures industry, if you look at my favorite indexes, the SG trend index, I think that’s the most intellectually honest one that’s primarily trend following and it’s just the big guys, right? That’s a form of diversification. And I think that that accurately communicates the beta of trend following managed futures. So when you’re building a CTA, there’s nothing stopping you from using multiple models, such that you’re getting rid of this dispersion risk to a large degree. So that was important to me. And that’s why I use three different models, short term, medium term and long term, because I know that any one set of models can be terribly out of favor, long term trend following probably outperforms overall history, but it can be terribly out of favor for three to five years. Medium term trend following can be out of favor, and so can short term. But if you use all three internally, I think you’ve gone a long way towards mitigating this dispersion risk.
Corey Hoffstein 36:30
With all these degrees of freedom. When you’re thinking about designing and building a Managed futures program. Are there ways in which you can create confidence for yourself that there isn’t maybe one of these degrees of freedom that’s leading to sort of outsize performance and not realizing it? So for example, maybe you have just more weighted exposure to the metals complex? Or maybe there’s a certain model that you’re running that ends up just in sample had a very lucky run, from a designing perspective back testing, are there certain things you can do to try to create more robustness?
Eric Crittenden 37:09
Yeah, so that goes back to the concept that I brought up a few moments ago about trying to break a system. And I know you’ve talked about this in the past, what I like to do is get a feel for how many moving parts are in your system, start turning those features off and paying attention to what happens to your system. So for example, I get this argument from people a lot that CTA is probably not going to make a lot of money going forward. Because bonds prices can’t go any higher. They say the CTAs have benefited from this kind of generational decline in interest rates from 16% and 1983, to almost zero today. And you can’t really repeat that going forward. And they look at the depth and the liquidity of the bond market. And they look at the size of the CTAs. And they say, Yeah, you guys disproportionally participated in bonds, and you’re not gonna be able to do that going forward. To which some CTAs respond well, in a rising interest rate environment, we can go short bonds, but savvy market participants understand the term structure and the coupons and whatnot. They say, Well, yeah, you can make some money, but not the same degree. It’s not the reciprocal of being long bonds. So there’s element of truth to all this stuff. And the way I dealt with it is I would sit down with someone and say, Look, if you think I’m overly dependent upon pawns, why don’t I just turn bonds off and show you what it would have looked like if I never did a bond trade in my life. And you can see, you can determine for yourself if you think I’m overly dependent upon bonds. So a bad system, a non robust system would fall apart when you kick out an asset class. So if you have a system and it’s compounding it, let’s just use round numbers here 10% a year. And you simply kick the bonds out of your portfolio reran and all of a sudden, you’re making 3% a year yeah, you’ve got a problem, you’ve got a bond problem. But if your 10 becomes nine, and your volatility goes up a tiny bit and your drawdowns basically unchanged, you don’t really have a bond problem, that return contribution, you should see some deterioration from any asset class that you kick out. But as long as it’s marginal, you probably don’t have a serious problem with that asset class not working going forward. Now, that’s a simplistic way of looking at it, I like to take it a few steps further and redo Monte Carlo simulations, use bootstraps, create different variants have history with the bonds missing, to try to get a feel for the full spectrum of the impact. But if you can’t do that, if you can’t kick an asset class out and still survive, then you’ve got a problem and you need to go back to the drawing board.
Corey Hoffstein 39:42
So as someone who spent a lot of time in the lab working on building your own trend following systems and models, what’s the best way you’ve come up with to stress test these systems? Well,
Eric Crittenden 39:55
I’m not sure it’s the best way. But the most unique way that I haven’t seen a lot Other people do is to change the currency denomination of the futures contracts. So let’s say you build a system and you’ve got 100 futures contracts in there, and you’ve got them all structured the right way, meaning some are denominated in British pounds, others in Singapore dollar, so on and so forth. Well, when you’re denominated in your native home currency, you’re going to have a certain path traveled, that’s going to match up with what we all saw, you know, over the last 10 or 20 years of that market. But if you read denominate your futures contracts into something obscure, like the Turkish Lira, or the South African rand, you’re basically going to be introducing, I don’t want to say randomness, but a completely different path traveled, but everything else is held static, the risk premiums are there, the trends are there. But you’ve changed the denominator, right? So if you rerun your system from the perspective of an investor who has a different currency, and you get terrible results, that probably means that your system was highly curved fit to just the path traveled that you saw. But it wasn’t durable and resilient enough to still collect those risk premiums if the path traveled is modestly different. So I haven’t seen other people do this very often. And it’s very illuminating to change the currency. And if you can use the yen and the Rand and the ruble and all these other currencies, the ruble and the lira are probably bad examples, because their own trends will dominate during periods of time, but it’s still eliminating to see that. So again, I don’t know if this is the best way to do it. But it’s the most unique way, in my opinion, to just change one thing, which changes the path traveled of every instrument in your analysis without actually affecting the risk premiums. Does that make sense?
Corey Hoffstein 41:51
Absolutely. You could almost look at it, I guess, as if your system works in one currency, but you were to change the underlying currency, then it’s more likely that your models fit to the currency and not the trends of the underlying markets. You’re trading.
Eric Crittenden 42:05
That’s exactly right. And I’ll share one other observation with you. When I did this. It told me one thing, but when I shared this observation with other people, it told them something else. So I’m curious as to what you think. So I took all the futures data, and I basically randomized it. I just randomized all the data. And I used a bootstrap methodology to reconstruct a different history. And then I ran the same profitable trend following models on it, and it did not work. Meaning your compounded annual return was not meaningfully different from zero going forward. Your drawdown control methods still work, because you’re risk budgeting along the way, and you’ve got your stop losses and whatnot, but you just didn’t make any money. And then if you turn it off, and you go back to what actually happened, you do make money. Right? So, and I tried this many, many different ways. And my conclusion was that well, that’s telling me that the markets themselves are not completely random. Because trend following does not work. When you use random data. It’s not profitable, doesn’t blow you up, you don’t make or lose any money. You just have some volatility. Sometimes you’re up a little bit, sometimes you’re down a little bit and net net, you don’t make any money, right? What’s your take on that if you just use completely random data, and you don’t make any money and my assertion that that means that while the markets themselves then are not completely random?
Corey Hoffstein 43:27
My take on trend following is that there’s really two components to trend following. There’s the sort of mechanical convexity that trend following has just if you follow the systems, there’s a tie to options theory that gives you this mechanical convexity, but that’s different than the premium. And that premium emerges from either being a liquidity provider, as you said to hedgers, or more naturally through mathematically the autocorrelation that’s going to exist in the market. So if you knowingly take away the autocorrelation, then the only thing that exists that’s leftover is, well, I shouldn’t make money on average, on a premium perspective. But I would still expect hopefully, at the mechanical convexity aspect would work in that trend following is going to have a lot of small losses and a couple of big gains. And by that nature, not quote unquote, blow you up. So I think that is what you’d expect to see, interestingly enough, is the conversation I have with a lot of people about can you Monte Carlo simulate, to test these types of models. And my argument has long been, you don’t need to simulate it. Because either your simulation is going to include the anomaly in the data generating process, in which case I know beforehand, the system is going to work, or it doesn’t include it, in which case, I know it’s not going to work. So it’s sort of defeats the whole purpose. But that said I do know you’ve done some Monte Carlo simulation with your approach and have found it to be somewhat useful, informative test,
Eric Crittenden 45:02
I agree with the point you’re making. The benefit I got from Monte Carlo simulations was getting a feel for the dispersion of potential future outcomes, how bad or good could it get at the 95th, percentile 99th, so on and so forth. It wasn’t to prove that trend following works on random data. But you just brought up an interesting point. And people talk about skew and the convexity, I spent probably three months of my life trying to build a system that traded on observable skew in the markets. And I thought, well, there’s a chance that if we can find right skewed assets, and participate in those will get right skewed results. What I found, though, is that the trends, they start off with right skew, that’s what gets you in, right, but after you’re in it, your experience during the trend is actually quite left skewed, after you. So let’s say oil breaks out when it goes from that 50 or whatever, and you go along, and then ultimately, you stop out at 75. But the path after from 50 to 75, the daily price moves, I expect them to show more left skewed and right skew. So and I think that might be what you mean, when you say that’s the risk premium, that skew risk, the skew flip after you get the signal is evidence that you’re collecting a risk premium. And it’s the stop loss that allows you to not fully participate in the big left skew when it does come. Is that consistent with your findings?
Corey Hoffstein 46:34
I will say intra market, I have seen research and done some research myself that if you try to measure skew and create a long short trade, you can actually if you go long the assets that have larger negative skew that theoretically have a fatter left tail, and short the assets that have less you are in theory capturing a risk premium there. And it does seem to be positive over time. So those two things do seem to actually sound like they line up. I actually I haven’t done the test. That’s an interesting take where you look at how the skew flips before and after taking the trend.
Eric Crittenden 47:11
Yeah, it was very interesting and surprising. Because I at the time, I was thinking, Well, if you’re on the comfortable side of skew, you’re probably not a premium collector, you’re a premium payer. That was my mentality at the time. And I thought, well, that kind of flies in the face of what all my CTA friends are telling me that no, no, we’re always right skewed, right, skewed, we’re buying stuff. And then our results are right skewed with a trade result looks right skewed to the client. But after you put the trade on, if you look at the daily or hourly data, after that, you’re eating a lot of left skew risk until you finally stop out which is consistent with my kind of bias in that you don’t get paid to be comfortable, you get paid to be uncomfortable. And that means taking the ugly side of skew risk. That doesn’t mean that if you always take the ugly side of skew risks that you’ll make money will probably blow up at some point. But along the path traveled, I guess my point is you don’t get paid to make yourself comfortable.
Corey Hoffstein 48:04
One of the other areas of which there’s a large degree of freedom is in the asset classes themselves. So one of the great things about managed futures arguably is you have all these different markets that you can tap into. And even markets that sort of fall into the same sector can behave totally independently. So it’s this opportunity to really create diversification among a large number of independent bets. But it does bring one of the difficult aspects of managed futures and constructing one of these models into question, which is the whole nature of risk budgeting needing to look at sort of the covariance between markets. And this diversification is often really a moving target. How do you in practice, try to deal with this problem,
Eric Crittenden 48:47
I think I wasted about 12 years of my life chasing my tail in this area, it is very complex. It’s very challenging, very nonlinear. So I’ll share with you the observations because I understand what you’re saying. And this is where the marketing departments oftentimes will run wild with things that I don’t believe are helpful. So two observations that I’d like to share with you one, when I tried to do the intuitive thing to take more risk when portfolio diversification was high, meaning we had a lot of independent bets in the portfolio when I tried to take more risk. This is in simulation, mind you not real money, things did not go well, which is very counterintuitive. You would expect that when the diversification quality is really high, that you should take more risk and get rewarded for it. And then when everything becomes highly correlated, you should back away and become more risk averse and you’ll be rewarded for that. It turns out in practice, the exact opposite is true, which was difficult for me to digest to the point where I had two other people go independently do this work and come back and show me that their results were almost identical to my on. So that was fascinating to me. And it turns out that you see a lot of independence between markets when markets are really choppy and not trending. So if and I’m not a big fan of predictors or things of that nature, but that one seems to hold pretty consistently is if the independence between assets is really high, it’s probably a pretty crappy environment for trend following at least in the near future. Alternatively, when the correlations are rising, and all your risk bells are going off, and you’ve got no diversification, and all these trends are just taking off at the same time, that lines up really tightly with wildly profitable periods of time for trend following for months and months after that. So again, cognitively and emotionally very difficult to take risk, to increase risk when diversification quality is going down. But that’s what the data keeps saying, and it didn’t matter what decade it looked at either 70s 80s 90s, it was very consistent through time. So
Corey Hoffstein 51:01
to that point, a lot of managed futures programs will operate by looking at the total volatility of the portfolio. If what you’re saying is true, isn’t there sort of a counter cyclical de risking that’s occurring going into the environments where perhaps trend falling might be most profitable, that as the correlations are converging, poor total portfolio volatility is going up, but that’s forcing the Managed futures manager to de risk?
Eric Crittenden 51:27
Yeah, your observations true, but I don’t look at it in a linear fashion, I think of the portfolio, the risk that you want to take in the portfolio operates kind of like a thermostat that is calibrated to the heat index, rather than just the temperature. So in the heat index is a function of a few different inputs, you know, the raw temperature, the humidity. So what you say is true, where these vol targeting CTAs, if you look at what they’re doing, you think, well, you’re shying away from taking risk at the moment when you should be hitting the gas. And then vice versa, vice versa, when your volatility is too low, you’re amping it up, right? So that’s dealing with one component of the heat index. But there are other things typically in the system’s meaning, if I only see trends in two markets out of the 75, I track, well, I’m not going to take any risk just a little bit, I’m not going to take you know, 18% portfolio heat onto the market. So there’s kind of a push pull in that thermostat that considers both the breadth of opportunity, but also the implied volatility of participating in that opportunity. So and I think that’s just responsible, you can’t just I mean, some people do some people with these tail risk programs will just kind of bleed, bleed, bleed, bleed, and then go for it with huge leverage when it when the stars are lining up. I don’t have a problem with that. But if you want to survive, and your job is to compound wealth at a reasonable rate with reasonable risk over the long term, I think you need that thermostat to have multiple inputs going into it.
Corey Hoffstein 52:54
One of the topics you and I have spent a lot of time chatting about in the past is the problems of investor behavior and behavioral finance. I know that, in parallel to doing all this research of how to build a Managed futures program. One of the things that was very top of mind for you, before launching standpoint was how do you package your managed futures program into something that is more sustainable in an investor’s allocation? And you actually ran something that you’ve sort of come to call the test when we’ve talked about it in the past? Can you elaborate on what this test was? And what were some of the results that you found? Yeah,
Eric Crittenden 53:35
I call it the experiment. Now I like that word better.
Corey Hoffstein 53:38
That’s better. I like that.
Eric Crittenden 53:40
So before I go into that, let’s just say that my goal here is to figure out a way to get people to use enough managed futures to make a difference without driving them to the brink of insanity. And without diluting or sacrificing product quality, meaning that I don’t have to create a bunch of filters to make manage futures more palatable to people. So tall task. So in order to figure out if this was even possible, I started doing a series of experiments, not very scientific, but they were very useful to me. And one of them was, you know, I would show people the calendar year returns of a Managed futures index, and let them compare those calendar year returns to the US stock market and ask them how would you feel about making a 5% allocation to this managed futures index? And the vast majority of them enthusiastically declined to make an allocation to manage futures? And I said, Are you sure just not even 5% And there’s no way my clients are gonna fire me it underperforms? This thing sucks. Why would I do that? Okay, all right, no problem. Let’s kick the Manage features out and then I would show them this thing that I call weather. And I didn’t tell them what it was. But between you and me, it’s just a 5050 split of managed futures and US stocks but they don’t know. So they do the same thing. they scroll down, they look at the years and they say, All right now we’re talking. This is something I can get behind. I like the up capture the down capture is great. What is this? I say, Well, I’m not going to tell you yet. But tell me more about why you like this. And the only way to go through it and like, well, the Vols low the drawdowns, great, the returns are great. I really liked this, if it’s not something crazy, sure, I can see myself using it. And then I revealed to him, it’s just a 5050 split of the thing you kicked out and the thing you were, and they’re baffled, absolutely baffled to the point where they don’t believe me. So I have to give them a spreadsheet and let them look at it and see the math. And it’s just rebalanced annually. 5050, split. So the point here, and so this is where I think people go wrong, people in my position, they then point their finger at the adviser or the client and say, See how crazy you are? See how you get it wrong. I think that’s a waste of time. Because it’s so pervasive, I would estimate that 99 out of 100 people are very predictable when you do stuff like this, and they’re gonna say no to this and yes to that, right. So the change the difference in my life right now is I look at it and say that’s my fault. They want me to provide something to them, that solves problems, not something that creates problems. So I’m aware that managed futures is the best diversifier out there, at least in my opinion. Instead of trying to cram that down their throat, why don’t I blend it together into something that’s not just something that they need, but also something that they want. So figure out a way, if I can take pure managed futures, mix it up with something else that I happen to believe works, and deliver something where there’s an overlap between what they want what they need, maybe that is what we should be doing, as asset managers, instead of constantly telling people decade after decade, you don’t get it, you need to read all this stuff, you need to read these papers and trying to educate, educate. I’m not against education. I’m all for it. But I don’t think that works. I think the Managed futures industry has missed their calling. They’ve missed the point. And they haven’t been empathetic enough or taking the time to understand the opportunity with the client. And I think that one, that’s one of three experiments that I’ve done, but I think that one really drives home the point that yes, they’re being irrational. But what’s more likely that you’re going to get them to start becoming rational, or you can create something that makes everyone happy.
Corey Hoffstein 57:17
So as you emerged from the lab, doing all your research, building a Managed futures product as you came back from the field, doing all these studies with advisors, and allocators? How does this come together? What’s the final picture look like?
Eric Crittenden 57:32
Well, for me, I had to satisfy my original mandate, which was build something that I would be willing to put 100% of my investable money into and leave it alone for 30 years, and also had to be something that satisfies an actual real need in the marketplace. And then optimally, it’s got to be something that can also satisfy what people want. So after all the research and the effort, and we’re skipping months and months and months here stuff, but in the end, I felt like a 5050 blend of global market cap weighted equities, and my own really simple, durable managed futures program blended together rebalanced in a thoughtful, methodical way where taxes are minimized as much as I can possibly minimize them. And fees are kept as low as I can keep them and still justify having a profitable, sustainable business. That that combination of stuff delivered in an all weather format would be my best bet at satisfying my mandate, and giving people what they both want and need. So that’s what we’ve done, the jury, we’ll see how it works out, I think that you’re gonna see more of this. And not because I’m doing I see other firms starting to do it. I think some people have seen the light and said, you know, this is probably an acceptable way to introduce managed futures into the marketplace, just don’t call it manage futures. Just call it all weather simple.
Corey Hoffstein 58:59
You’ve now over the years, brought a number of managed futures products to market first in sort of hedge fund to funds that long board and mutual fund structure and then just recently launching another 40x vehicle. How has this sort of operational structure of managing funds evolved over time? What’s sort of the modern structure look like today? Well, in terms of
Eric Crittenden 59:23
cost, it’s a lot lower, I can tell you that. So we use Python and open source products now and direct API’s for collecting data. So it’s pretty efficient, clean. It’s not that complicated. So just 10 years ago, it was a lot more complex. So now, you know, we’ve been doing it long enough where collecting the data is effortless cleaning it databasing it processing it, creating blotters reconciling trades. All these things can be done with open source software. Everything we do is in the cloud. Everyone is remote at this point and has been for months. So I would say that technology has gotten To the point where you can build a lean, efficient, secure CTA and do everything in the cloud remotely. And if anything goes wrong, you just use your backup plan because to your actual hardware. So I’m very happy with the workload associated with running. I mean, just 10 years ago, or even 15 years ago, you needed a decent sized staff, a 24 hour desk. And I just don’t feel like that’s true anymore. Last question
Corey Hoffstein 1:00:26
for you. It’s a question I’m asking everyone, given sort of the recent turmoil we’ve had 2020 has been an odd year, and we’re all trying to find ways to be hopeful and excited for the future. So what are you most excited about going forward? And it can be either related to business markets or just life in general?
Eric Crittenden 1:00:44
Wow, that’s a tough question. Because I think you’re gonna see hard times going forward. The current environment, if I have to use an analogue from the past, I would say it’s a lot like the late 60s, I don’t know how much of a student of history you are. But in the late 60s, it was all about what’s deflation, that’s what we’ve got. And you had social unrest. And, you know, the central banks were very active, very similar to what we have right now. And nobody respected the idea of inflation, inflation was a joke, not something to be feared. The idea that we were going to have a stock market go down 40%, and then 50%. And the interest rates would go from one and a half up to 16. And inflation would be rampant. And a 6040 portfolio would make no money in real terms for 14 years. If you had said that stuff. In the late 60s, you probably would have gotten punched, you definitely would have lost all your credibility, but that’s exactly what happened. And then when I look at the current demographics, and who’s retiring, and the entitlement programs, the tax structure, like how it’s all set up, and I think, well, the path of least resistance is exactly what they’re doing stealth monetization of debt, I’d be doing the same thing, there’s really no other choice. And everyone’s super confident that we’ll never have inflation. People are very confident with the stock market, you just had the fastest 30% decline in history, and it’s almost at new highs right now. and the NASDAQ isn’t new highs. And so I guess my point here is I’m not saying that the past is going to repeat what I’m saying is, when you’re most confident, and you just don’t think something can happen, oftentimes it does. And if you’re unprepared, you really have only yourself to blame, and it’s gonna hurt. So the idea is, be prepared, as best you can and be happy with that realize that you’re going to underperform the people that are unprepared when they’re getting lucky. And don’t view that as you did something wrong. So what am I optimistic about? That’s a tough question. I’m optimistic about all weather investing. I’m very optimistic about CTAs. And trend following going forward. I think that if you ever wanted to buy something that was out of favor, I’d be looking at that category right now. I think that they’ve been beaten up to the point where their profit margins have gone so low and stayed low for so long. But with no fundamental reason for them to not work in the future that I’m happy to. I can’t guarantee it’ll work. But it’s a great bet, in my opinion. So there’s that I’m optimistic about people that are realistic, prepared and disciplined, like discipline and preparedness are going to pay off in the future. And there’s a lot of people, a lot of good people I know that have severely underperformed the stock market or a 6040 portfolio that have been getting beat up by themselves and their clients for a number of years now. And ask the question, Why does doing the right thing hurts so bad? Well, in my experience, and I’m 48 Now, usually when you’re asking that question, doing the right things about to pay off, but the closer you get to don the darker it feels. So I’m optimistic that old school, fundamentally sound business decisions are going to start to work again in the future. And I can’t tell you it’s going to be this year or even next year, but that kind of old school pragmatism will pay off going forward and it’s long overdue.
Corey Hoffstein 1:04:02
Eric, the insights have been great. Thank you for joining me today.
Eric Crittenden 1:04:06
You bet. My pleasure. It was great being on