How Best to "Replicate" Managed Futures Returns ft. Katy Kaminski

[00:00:02] Intro

You’re about to join Niels Kaastrup-Larsen on a raw and honest journey into the world of systematic investing and learn about the most dependable and consistent yet often overlooked investment strategy. Welcome to the Systematic Investor series.

[00:00:24] Niels

Welcome and welcome back to this week’s edition of the Systematic Investor series with Katy Kaminski and I, Niels Kaastrup-Larsen, where each week we take the pulse of the global market through the lens of a rules-based investor.

Great to be back with you this week, Katy. How are you doing? Where are you? Where do I find you today?

[00:00:40] Katy

Great. I’m in Arizona today, just been at a conference, so, just good to be in some warm weather.

[00:00:51] Niels

Indeed. We need a little bit of that over here in Europe.

But speaking of conferences, it was very nice to see you in Miami last month at the Global Alts Conference. But I have to say I think it was only like 10 seconds that I managed to speak with you. I hope you had a good time.

[00:01:06] Katy

We had a little walk by. Hey Niels, how are you doing?

[00:01:09] Niels

Yeah, yeah, exactly. But it was a good event for you?

[00:01:14] Katy

Yes, it was really good. I mean just, you know, meeting with lots of investors. I mean there’s definitely a good vibe for alts these days and so, you know, it’s just really impressive how many people. What a large event that Global Alternatives event is really. And a good time of year to go, right?

[00:01:34] Niels

It’s a wonderful time of year to be in Miami, I have to say. Now we’ve got some good questions that came in for you. You brought along a brand new paper that people should be excited about once we dig into that. So, we’ll go through that as well.

But as always, before we dive into any of that, I’m always curious about what’s been, kind of random things that have come by your radar. And I’ll share a couple of my random things. But anything in particular you’ve been noticing in the last few weeks?

[00:02:09] Katy

Well, I think. You know what’s interesting is I find it very interesting with the bank of Japan. I mean, I keep getting a lot of questions about that. I think, because you’re seeing such a divergence in policy and divergence in inflation numbers and we’re really in a strange, something we’re not used to. We used to talk about convergence. Now we have this very different monetary policy.

And so, I keep getting a lot of questions about what’s going to happen to the carry trade? What’s going to happen if the bank of Japan hikes rates? And I definitely think we’re not going to be bored this year.

[00:02:46] Niels

No, I think it’s a great point. We’ve been talking about it on the global macro series for a while and I do think people should pay attention. I’m not even sure people really, and that includes myself, fully understand what a difference it makes when someone like a Bank of Japan changes policy, which it hasn’t really done for like 20 years or so. It could have quite significant implications, but also be very exciting for trend followers.

And I think I mentioned last week, on the show, when I was on with Graham, that actually one of the trades that seems to be working out for trend followers at the moment is the widow trade, the short JGB trade, which finally, after decades, seems to be throwing off a little bit of a profit at least.

So anyways, now I’ve got something on my radar that has nothing to do with trend following or the BOJ or anything like that, just things that I sort of, when you read the headlines, you kind of wonder maybe, and then some of it could be a little bit political. But I’m not really trying to be too political anyways. But I couldn’t help kind of noticing the headlines about how much the share price of Tesla has dropped in the last couple of months.

And I’m kind of wondering, because there’s been a lot of publicity, I guess is the word, about shareholders, and Elon Musk, and what he wants, and what they want to approve. And I’m just thinking, well, maybe they, at some point, would be a little bit disappointed with the share value of the stock, given all the time he’s spending in the White House nowadays.

And I also notice, which I actually didn’t know, but looking into this, that the big sovereign wealth fund in Norway apparently had voted no to his quite generous pay package, I think, at the end of last year, which led to Elon Musk declining an invitation for dinner by the CEO Nikolai Tangen.

So, I think there’s a little bit of interesting thing that could come to the service later on. I’m not expecting you to comment on that. You may have no opinion, but I just thought it was fun and interesting.

[00:04:54] Katy

And my only opinion is, someone told me this the other day, just think of Tesla as an option. It’s going to trade like an options contract. So, there you go, lots of variability.

[00:05:05] Niels

Absolutely. Question is, is it a call? Is it a put? Who knows? Who knows? Anyways, the other thing I noticed in the headlines was two things about the crypto space.

One is that apparently the biggest one-day outflow from exchange traded funds happened this week on Tuesday. About a billion dollars was pulled out. We know, of course, from looking at the price, that bitcoin is and other cryptos have been under pressure. So that’s kind of interesting as well.

And also, the fact that there was the biggest, or maybe it’s the second biggest crypto hack ever also, recently in the last week or so, where they stole about $1.5 billion worth of Ether from an exchange called Bybit, I think from memory. So anyways…

No, actually I think it is the biggest hack. I think it’s the second biggest crypto exchange or currency which is Ether, but it’s the biggest hack. So, a couple of interesting things from the world of crypto that I came across.

Now back to some more familiar grounds which is trend following. And you know February hasn’t been, we are recording almost at the end of February. We have one more trading day tomorrow and it hasn’t been a great month for CTAs so, trend followers so far. Mainly thanks to, as far as I can tell, reversals in fixed income. Clearly there has been much more positive support coming for bond prices and shorter dated fixed income markets.

We’ve seen some changes in currency, so to speak. We’ve had some weakness in equity markets and then also a few kind of more idiosyncratic moves. Like Dutch nat gas has had a big move this month, a big drop actually in price, but it was in a pretty long term uptrend. And, of course, last year’s big winner cocoa has also come under some severe pressure in the last few weeks which will be affecting long term trend followers.

The other soft that has been more in focus this year is coffee but that seems to have, so far at least, only given back some earlier gains in February but is still kind of hovering around break even, in terms of price, for the month.

So as always, I’m curious to hear your observations maybe of the year so far or of the month, whatever you’re seeing when you look into the engine room.

[00:07:45] Katy

Yeah, I mean this year has been a lot of starts and stops, Niels. What I keep noticing is you have to kind of think about the longer term transition of these trends. I mean we started out really sort of on that Trump trade in November. So, that was really you know, long US, short fixed income coming out. And also you had long dollar. And you kind of saw that trade peter out and transition some, where the long dollar was the only sort of residual strength going into the end of the year.

Then we moved into January. And in January, this long dollar trade has started to unravel some especially more recently. And the US trade, although January kind of turned around, you’ve seen a lot of sort of hesitation around that sort of US versus non US.

And you’ve seen some adjustments to some of the positioning and some sort of views where things like Hong Kong have been sort of one that stuck out this month. And so, you know, it really feels a little bit like the market is trying to determine whether or not this US exceptionalism theme has wings or not.

And the last couple of days in particular, you’ve really seen some capitulation of investors kind of saying like, wait a minute, given valuations, given all of this turbulence regarding sort of trade policy, you know, I’m not sure what’s next. And you know, even though we might believe that the US has pro-business intentions, and that these things are going to be good, that that there’s a lot of things that could be stressful beforehand and there definitely already are things that are stressing out investors.

So, I’d say it has made trends a little bit more challenging, particularly if you look at the overall themes of trends that were long equities, short fixed income, long dollar, relatively US tilted, you know, there’s been more risk off, which is the complete opposite.

So, you have yields, even though everyone’s worried about inflation, yields have not gone up as much as some would guess. And then here you have, you know, Bessent in the US saying like, oh, we’re going to keep them low. And you and I both know that you can say that. But you know, if you push up inflation, bond yields are driven by pricing, by what people’s appetites are to purchase something. So, you can’t really force that if you add inflation. So, it is a tough time. I feel like there’s room for some disruption, but we don’t know what that is yet. I don’t know if you know what I mean.

[00:10:30] Niels

Yeah, I agree with you, and obviously it has been a disruptive time, in the last few months, in many regards. And, of course, this is completely normal within our world. We go through these times where the trend picture is not so clear as we would like it.

In fact, when I look at my trend barometer yesterday, it finished at 23. And for those who follow it on the website, they know that 23 is really almost as bad as it, it gets. I mean it might get into the teens, but rarely much lower than that. And that really just kind of confirms that we’re in a very weak trending environment right now.

Obviously, the indices are going to be negative for the month, I imagine. I don’t think we’re going to miraculous recovery in the last trading day of the month. February has actually, historically, I was looking at our own track record and I was just looking at February as a month and there’s been some big moves in February, performance wise, both up and down over the years. So, this is actually not an extreme in any shape or form.

But you know, the year has started off on a slightly softer note for the industry, for the indices. And so let me just sort of recap what’s going on right now.

So, these numbers are as of Tuesday the 25th. Yesterday, Wednesday, I would imagine people made a little bit of money. But as of the 25th, BTOP 50 was down 1.27% in Feb, pretty flat actually, down 8 basis points for the year so far. SocGen CTA index down 2.33%, down 1.72 for the year. The SG Trend index down 2.9%, and down 2.75 for the year. And the SocGen Short Term traders index down 76 basis points for the month, and down 71 basis points for the year so far.

In the traditional world, as we alluded to, equities have been a little bit under pressure. The MSCI World is down 41 basis points but still up 3% for the year. The 20-year S&P US Treasury Bond index is actually up very strong, up 5% and change so far in February up 5.4% so far this year. And that shows you that change in sentiment, that you were alluding to, from the short trades that we were all in at the end of last year. And then the S&P 500 Total Return down 1.29% for the month of February, but still up just shy of 1 1/2% so far this year.

Katy, we’ve got a couple of questions, that we appreciate, by the way, coming in.

The first one is from Rick, and Rick has read your January paper called Slow and Steady, and he says to thank you for that. So, that’s passed on right now. And then he says, “Katy mentioned a particular factor that you know will get me excited based on our previous conversations, positioning. I believe she quoted Nick, and this is Nick Baltas’s work on it, where he calls it co-momentum factor. I have been building and live trading off-side systems, fading extreme or off-side positioning.”

I’m not entirely sure what’s meant by that. Maybe you do, Katy, but he’s generally asking if this is something that you have looked at or done any work with.

[00:13:56] Katy

On co-momentum, co-movement, so yeah, we actually implemented the co-movement factor because we were huge fans of Nick’s work and, you know, one of the biggest questions we get from investors is a question about crowding, and position crowding, and sort of excess co-movement.

So, Nick in some of his work, had demonstrated how you can estimate a pseudo crowding measure by measuring co-movement. And co-movement is really the excess correlation relative to the overall benchmark of an asset class. So, for example, equities minus their overall sort of beta.

So, you can basically look at how, in excess of what you expect that particular asset class to move, how much do they move? And so, we use that factor to understand sort of more coordinated or co-moving assets that could potentially be a sort of indicator for crowding.

And the reason this is interesting is, I mean, you can look at it as a signal. You can also look at it as a, you know, just like anything in trend, you can think about these as either a shorter term signal, but you could also think of them as a filtering mechanism for positioning in a portfolio or a risk management opportunity as well. So that’s what’s interesting about a crowding metric of this type. And so, it’s something that we started actually tracking as a factor because you can look at which of the assets are more co-moving.

So, let me just give a simple example, for those that are not used to this type of technical terminology, is like imagine you look at something like the precious metals, right? So, some of the precious metals in a certain environment, let’s say it’s a tariff environment, I’m just making something up here, but you have a situation where the tariff theme is sort of affecting all of these metals in excess of just the typical moves in metals. So, people are kind of piling into similar positions.

So, you end up with excess co-movement in those assets. In that case, it could indicate that there’s more reversal risk for those types of assets. And what we’ll do with this particular factor is we’re just trying to understand return attribution. So, at the end of the year where there more crowded markets, better or worse traits? And that’s what the paper was really sort of documenting.

But in practice, anybody who’s interested in using this co-movement metric, it’s a cool one for both signal generation, it’s interesting for risk management, but we use it as a, in this case, as a way to attribute performance in a year. So, let’s say there was a year… And co-movement can be positive. So crowded trades aren’t necessarily always a bad thing. You know that, Niels. Sometimes those crowded trades are a macro theme that really actually extends even farther. So, feedback trading, which is part of why sometimes trend can actually work.

So, we like to use that factor to understand where crowded markets and crowded trends are kind of the profits center or not and in which asset class. So, that’s what that factor is about. I’m glad he asked because Baltas is the expert in that one.

[00:17:14] Niels

Yeah, absolutely. And this might be a little bit of a, but just to clarify, I mean, is this closely linked to some of the papers we’ve seen on principal component analysis and so on, and so forth?

[00:17:26] Katy

They’re related, but the construction is a little bit different. Principal component analysis is purely statistically based.

I think when you look at co-movement, you actually are specifically adjusting for an asset class’s beta in some sense. So, it’s constructed in a very different way. But they will be related. Because statistically PCA analysis is really just a purely statistical way of determining which are the key components that are driving returns. Whereas this methodology has a more explicit definition. But those two things could mean the same thing, depending on what is driving returns. But good question though, Niels.

[00:18:14] Niels

Yeah, good. Well, there’s another question from Rick and also relating to your work.

He writes, “Katy’s work shows us that the dispersion between slow and fast systems with an optimal window of 10 months. She mentioned this on the December podcast (that could have been with Alan, actually). How does this optimal window differ by year? Is it random over the past 10, 20 years or do certain windows get sticky? Meaning, for example, does a 10-month window, that just began in 2024, typically repeat for a couple of years at a time?”

So, I don’t know if you have any numbers on this, but otherwise I can share some thoughts.

[00:18:57] Katy

Yeah, I mean, I think it’s very idiosyncratic. Each year is very different. There are a couple of key themes. And this is why we don’t trade one window. Right? Because there’s just a lot of noise in trends and they have very different time series patterns.

So, the way that you want to reduce some of the noise around the variation of outcomes is to create signals that are diversified across different approaches, across different time horizons and methodologies, and that will smooth your sort of capture of trend over time. And so, what that means is that we like to look at this simple example that I put in the paper because it gives you an expost explanation of like, okay, which worked last year.

But you know, you could plot this (and I’m sure you probably have some interesting numbers on this, Niel), you could plot this for any year and maybe it happens to be that like three months was the worst and four months was the best. And it can happen like that just depending on how markets turn and twist. So, my general view is long.

There are a few key things to note though. Longer term has done better over longer time horizons. But longer term is a tradeoff for diversification in your signal, and your convexity, and your ability to react. So, a lot of the importance of the shorter term signals is really those pivot points. It’s also for periods like a Covid or others where, you know, if you use a 10 month horizon on an event like Covid, you’re just not going to move and you’re going to end up having big swings and too late to adjust to anything.

So, I think the key is that longer term has a slightly higher Sharpe ratio. But shorter term is really important for convexity and for crisis alpha. So, it depends on what you’re looking for. If you’re looking for a program that’s going to react and provide the most crisis alpha, you need those shorter term signals.

We didn’t really have crisis last year. We didn’t have drawdowns. So, if that’s the case, you know, guess what? Like longer term, more risk premia type strategies, buy and hold, actually did a lot better. But you never know, from one year to the next, if next year is going to be a year where actually you need that reactivity or not. So, I tend to kind of suggest diversification.

[00:21:27] Niels

Yeah, I mean I think you’re speaking for the whole industry when you say that we all trade different lookbacks and speeds. But of course, at some point you’re going to find that there will be some concentration in terms of these look backs.

What I can share with you and with Rick and everyone else listening is that we actually do an analysis at the end of each year just to look at what the lookback performance looked like between anything down to just a few weeks and then all the way up to about a year and a half, two years. And what’s interesting about it, because something I just noticed when reading the question and looking at the results. So, I have 25 observations all the way back to year 2000. The first observation I can share is that about 15 of those 25 years the absolute best lookback period was more than 200 days.

So, this doesn’t mean that you couldn’t have made money at 200 days or more in other years. For example, in 2008 the absolute best lookback period was something like 56 days. But again, if you had traded anything in trend in 2008 you would have made money. So just take it with a little bit of grain of salt.

I’m just looking at here the absolute best lookback period. So, the first thing it kind of confirms what you said, 10 months or more. And that’s what I see as well. The other thing I did not think about but actually Rick has an interesting point there and that is yes there is a bit of clustering because ’01, ’02, ’03, ’04, ’05, ’06, ’07, they all clustered at more than 200 days. Then you have about a, well, what is that, about a 10-year period where only 2 was about 10 months or 200 days or more as the best period. The rest were shorter term, like 100 days or so. But then you get another period of six years in a row, starting in 2017, where the best lookback period was more than 200 days.

2023 it wasn’t, but then again 2024 it’s about 200 days. So, I had not thought about the clustering part of the question. I just think about it, as you say, in the long run long term trend following seems or slower trend following seems to be working better.

And frankly, just looking at the Short Term Traders index that we quote every week, if people go and look at the performance over time, there really isn’t any performance in that index to be frank. And so, it’s another way of kind of confirming what we see in our research data, that you need to be long term but of course you do pay a price for that and that’s just how it is.

[00:24:37] Katy

Yeah. and I think, Niels, in our book on trend, one of the key things we talked about there is that the real value, especially of shorter term, is that convexity and the ability to move for diversification. So, I think it’s a question of, for me, it’s a question of is a client looking for something that’s the most compatible with equities or are they looking for something that has the highest Sharpe? And we see that that varies quite a bit by the investor.

But we have certain clients where risk mitigation is what they care about and thus, they actually really care a lot about that reactivity because they care about how the conditional returns of trend are, compared to equities, as opposed to over 25 years. So, I think that’s another key component because as you go into longer trend as well, you’re capturing some of the risk premia which is good.

But if you’re thinking about this as an alternative, that’s going to be the only thing that can be a defense line in a really bad year or during a bad drawdown period, then having that allocation to shorter term, that reactivity is very important. So, I think that’s more of a long-term objectives question.

[00:25:53] Niels

Yeah, absolutely, of course. Now then we move on to a question from Brian.

Brian is also interested in the new ETF offerings that we will talk about later on today of course, in the CTA space. And it’s two very simple questions, but maybe not so simple to answer. One is, “Do you find that there are more individuals that tend to gravitate towards the ETF space for CTAs or is it actually institutions that tend to like these products?” That was the first question.

[00:26:28] Katy

That’s a good question.

I would say that it’s a little bit of both, but probably slightly more individual and more sort of, I think retail, especially for the US. And I think, Niels, you and I have talked about this quite a bit, is that the ETF space is exciting for sort of the disaggregated pension investor in the US. And I think what we’re seeing more and more is a lot of RAAs and other investors that are looking at sort of ETF-only model portfolios and other ways to access alts. So, I think that there are a lot of investors who like the vehicle of an ETF and find it easy to use within their constraints.

So, I think, in that sense, it’s an individual. And what’s exciting for the institution, and this has been interesting for me to see is that, you know, with an ETF as a wrapper, there are a lot of institutions that, you know, shy away from mutual funds, but they have sort of individual accounts or they do LP investment and there’s a lot of, you know, subscription redemption associated with those type of accounts. Whereas an ETF is something that they kind of think of something that they can just buy in a day or adjust their allocation without a lot of documentation.

And you know, that may be important in the sense that, so, they’re seeing it as something like a liquidity provider or something. You can park some assets into a CTA like return and that will give you something in addition to an allocation you already have. So, I think the use case is very different for an institution than it is for an individual.

What’s exciting to me, I mean, just to give an example is like the size of the ETF is very, you know, the amount that you need to allocate is quite small. So, you know, someone can allocate to an ETF in a health savings account or, you know, in an easy way. So, I think that’s why individuals are interested, and that’s not that much different from a mutual fund.

[00:28:43] Niels

That is true. That is true.

Okay, well, the follow up question from Brian is just, and these might, and I’ll explain what I mean by these might be difficult questions for us to answer. He’s asking whether ETF holdings tend to be long term. I mean, are investors long term in these or do they chase performance?

I’d love to hear your initial thought, but I’ll share with Brian why I think these are difficult questions for us to answer.

[00:29:11] Katy

Oh yeah, these are hard questions to answer. Obviously, I need to spend more time looking at the data. It depends on the use case. Right.

I mean, so here you can imagine it’s no different from a mutual fund. I think you see that in mutual funds as well. You can even see that in institutional accounts. But, you know, I think return chasing and sort of churning of portfolios I think is always a concern. I’m not sure it’s more of a concern there than in the mutual fund. I need to do more analysis to do that.

And you, you can explain why, I think, Niels, as well.

[00:29:50] Niels

So obviously, you know, our firm, we don’t have an offering in the ETF space we have in the mutual fund space, but in the ETF space we don’t. But one of the things that I remember from an early conversation with Jerry about this was he kind of liked the idea about the fact that now people could just redeem their investment. They didn’t have to call the company and say, oh, I’m thinking of redeeming or asking for paperwork, et cetera.

So, it makes it completely… Anonymity is there. They can just do what they want. And I’m thinking, first of all, that might make it very difficult really to answer these questions in terms of who are the actual investors, like the first one from Brian, and secondly, are they long or short term? We don’t know because we don’t, in a sense, know who the underlying investors are.

Which leads me to sort of also the fact that it’s more difficult, I guess, for us to do maybe a little bit of hand holding when needed because we’re not entirely sure who might be thinking about redeeming because they can just press the key on their phone and off they go and they’ve sold. So, I guess the transparency makes it harder for us to answer these questions, I would imagine.

[00:31:03] Katy

Yeah, I think so. But I think this is a different vehicle and I think there’s the same pitfalls with this vehicle that you have with mutual funds, I think, as well. And the biggest concern, and I think Morningstar has spent a lot of time documenting that, too, just this idea of actual returns versus investors, behavioral returns. And so, this is not a new concept.

What I’m hopeful for is that I’m a big fan of model portfolios and of sort of business where, you know, let’s say it’s an RIA Network and such where they have allocations to alternatives, and those are sort of blocked into groups, and sort of the idea is not that you kind of, Monday morning quarterback those.

And if that’s the case, you know, if you have those behavioral constraints in place, and you’re using investments in a long-term perspective, that’s great. But, you know, I do agree that anytime I always get nervous, as when people try to trend follow trend following, you know, and time what we’re doing, and sort of decide when a crisis is happening.

And, you know, and that has been the case, you know, for many years for us in this space. It just may be a little bit less easy to hold hands, I guess.

[00:32:25] Niels

Sure. Yeah, absolutely. All right.

Well, the third question is actually from me, Katy. It’s a question to you, from me, because it’s something, and it actually is a nice segue, I think, to the paper we’re going to be talking about very shortly. And it’s a question that came up twice in my travels this week. So, I’d love to hear your considerate thoughts about this.

So, when I meet people, especially people who are looking to invest for the first time into this space, they’ll have their discussions, they’ll have to put forward some kind of convincing argument in front of the investment committee. And often they have some kind of model, and I’m not entirely sure what kind of model it is, but they’ll have some kind of model they need to feed with, “expected returns” for the various assets that they have, or they plan to have in the portfolio. So, twice this week I was asked about, so, how should we think about future returns for something like, you know, trend following.

And I’m, I’m told by these people that you can actually put in your expected returns for equities and other asset classes. Even though of course I personally am very skeptical about anything that has to do with predicting the future.

But nevertheless, this is what they’re asking for. And this is what I wanted to ask you how you would approach a question like that because I think you have a slightly different approach to it than I thought of at the spur of the moment.

[00:34:01] Katy

Yeah, I’ve gotten that question so many times and that’s a very equity centric question. And whenever I get the question, I just clarify that in our space we allocate risk, we don’t target return. And what this means is that if you really wanted to think about targeting a return, you should think about estimating your estimated Sharpe ratio that you think that you could achieve in the strategy.

And depending on the strategy it’s much more, perhaps more believable to think about, okay, this could be a 0.5 Sharpe or, etc. So, then let’s say that you estimate a 0.4, 0.5 Sharpe and you just need to think about, okay, if I have a 12 vol or a 10 vol or 20 vol exposure, then I would think about, you know, multiplying that by the amount of vol exposure and think of that as a long term estimate because we don’t really think about return in our space. We think about allocating risk. And so, you need to think about what’s the relative return to the risk allocated as your estimate.

And that’s kind of a way to kind of turn what we do around into a more equity centric view with saying that, you know, returns are extremely difficult to estimate, but Sharpe ratios kind of make a little more sense in our space because we do allocate risk. So how much risk do you have? What kind of Sharpe do you expect? You can get an estimate over a long term horizon using that methodology.

[00:35:34] Niels

Yeah, and I think that’s a very elegant way of thinking about it, which of course I didn’t at the moment when I was asked. But I did say to them that I thought maybe when you take a firm like Dunn with a 50 year track record, maybe that’s a very good starting point to look at what that actually has done over the years and as you say, turn it into a Sharpe, and there you have maybe an even better way of thinking about these things also because I mean some of the benchmarks that often are used in our industry, frankly, I mean they started in the year 2000.

I just think that the year 2000 is kind of arbitrary and also maybe the world was somewhat different prior to 2000, and it maybe will be different in the future. Anyway, I’m skeptical about it. But I think your idea about a risk or Sharpe ratio in this case is a good way of thinking about that. So, appreciate your thoughts.

Which, which leads me into the topic of today, which really is thinking about maybe the beta of trend following and how you can access it these days. Because there has been a lot of stuff happening in the last couple of years in our industry. And you guys have been at the forefront of this and certainly when it also comes to discussing it publicly, so to speak, through your wonderful writing.

And so, I’d love for you to take us inside your latest paper tracking trend, a closer look at managed futures Index replication that you wrote together with your co-worker, maybe you should pronounce her name.

[00:37:12] Katy

Ying Shan Zhao.

[00:37:13] Niels

Ying Shan Zhao Awesome.

[00:37:14] Katy

Yeah, we’ve written a bunch of papers together. This is a really fun paper. It’s not a long one, but we decided to write an interesting paper about methodologies for accessing trend exposure in an ETF type of framework, when you’re looking for sort of a beta type of exposure. And I use the word beta, you know, not explicitly as like you’re getting an actual beta because one of the big challenges with the CTA space, it’s a very diverse space, a lot of the indices include investments that you can’t directly invest in that type of vehicle. So, you can’t just put the building blocks together and build up the exact portfolio of the index like you can in equities. Right.

So, you have to think about, I want CTA-like exposure, but I don’t necessarily access all the same markets or all the same investments. And so, we want to sort of track or follow the general views and exposures of the industry over time. If you want to do that, how can you do it?

And so, what this paper really is trying to do is sort of simplify the narrative of (we can talk about this later too, Niels), of just the different methodologies that you might want to use. And you know, yourself, we have different people on different sides of the fence.

And just a little bit of caveat on this is that Alpha Simplex is a firm that has been interested in and had been doing hedge fund replication and other things like that throughout its DNA and history, and sort of was one of the early people to actually do this. So, we do believe that these methodologies work for beta-like exposures.

But we also, you know, since we do run a CTA and sort of also really value the risk management and the sort of dedication to, you know, the art of running a managed futures best ideas program. And I think what’s interesting to us, looking at the ETF space, there are sort of two camps.
There’s sort of replication and then there’s direct implementation, or mechanical replication is what we call it. So, both of are, in some sense, replicating CTA like strategies.

Index replication is really looking at the returns of the industry and estimating them in different methods and sort of following sort of the overall industry footprint. Whereas mechanical replication is often sort of using a set of models, often a simplified set of models to replicate CTA-like returns in an individual strategy.

And so, if you look at this particular paper, what we examine is we look at sort of the pros and cons of the index replication and also the mechanical replication approaches. And these are some… Maybe we should just walk through these quickly, for a second, and give you a chance to ask a question if you want.

[00:40:26] Niels

Sure.

[00:40:27] Katy

So, we kind of, in the paper, kind of discuss the differences between these two. So, index replication is inferring positions of the index. Whereas mechanical replication is a method that is a simplified version of a typical strategy that would represent something similar to that index as a standalone. So, one individual approach.

Index replication is often regression based or ways of doing inference on sort of returns and other inputs. Whereas mechanical replication is an individual strategy. It can be trend following, it can be different signals, it could be moving average crossovers.

And so let me just discuss some of the pros and cons of each of those. So, if you think about index replication and we have sort of some vocal, you know, advocates of this area as well, we can talk about that in a second. So, in an index replication space, you know, you get an aggregated view of the industry exposure.

It’s often argued that there’s less manager parameter specific risk because the parameters will change and move with the index as the managers change themselves. And those are some of the pros of that approach, is that it’s not really sort of having one strategy and sticking to it. It’s really about following what the industry is doing over time.

Now on the other side it can be a little slower to react because it’s much more backward looking and it can also be misspecified, you know, especially as you do aggression, there’s a lot of colinearity. So, you know, you may sort of end up with more chunky type of exposures that may be less representative of the actual positions.

And, in general, risk management would be inherited from the process as opposed to specifically implemented in the approach. And so, I think there are some big pros where you know, you’re getting something that is following what the index and the industry is doing. The cons of course being that you know, it is a little slower and it may not adjust to certain environments.

Okay, quickly turning to mechanical replication. Mechanical replication is direct strategy implementation. It allows you to incorporate all the risk management techniques and some of the tricks and tools of CTAs, and it can directly react more directly with market movements similar to an individual strategy. The challenges of course of mechanical replication are it’s one set of parameters and it can be specific to that particular manager’s way of thinking about the CTA space. And as we know, return dispersion in the CTA space is real.

You have, you know, kind of one manager could be the best manager because they had that 11 month window this month, you know, as their focus and then they could be the bottom the next year. So, you have a little bit more return dispersion with one individual strategy.

So, kind of you can see the pros and cons of both of those approaches as you’re trying to think about beta exposure. The pro of the index replication is you know it’s going to follow the market, as moves, may not react as well. The other pro is you don’t have to pick a manager and their specific implementation per se. And then, you know, there are obviously advantages to both.

[00:44:02] Niels

Yes, absolutely.

I think that’s a great overview but I think what I really liked about the paper is that you took it a step further because you wanted to look at kind of the differences also, in a sense, the output, where it works well, where it doesn’t work, and actually what happens when you combine the two approaches. And I thought that was very enlightening because it’s often things that I’ve been thinking about but just haven’t had the time to go into this much detail myself. So, I’d love for you to talk a little bit about that part of the paper as well.

[00:44:35] Katy

So, this is great because is, you know, I mentioned the two different approaches and in the paper… And this is very much, I guess, maybe my view is I’m a very agnostic person. I think that there are pros and there are cons to everything. And I think diversification is great. So, what the paper does as well is it suggests that, you know, there’s not just index replication, there’s a mechanical replication that you can actually combine them.

So, we combine them in what we call in the paper informed regression, which is similar to a Bayesian type of approach where you incorporate sort of individual strategy insights combined with index replication to allow yourself to kind of lean into both of them at some times.

Let me explain, this is a little bit hard from a technical perspective, but maybe from an intuitive perspective, there are certain periods in markets where, if you’re looking at past returns or you’re looking at how CTAs might be positioned, as markets shift, they can shift quicker than you can kind of observe that using inference techniques. So, by using both methodologies, you have the ability to lean in or out of your positioning as the market changes.

So, for example, if you have a massive shift in positioning, an individual strategy will pick that up quicker than the index itself. And, at the other side of it, when you have one individual strategy, you may be overexposed more than the index itself and you’re kind of moving a little bit out of sync with the other managers. In that situation, you might sort of lean in to sort of the aggregate view instead of the individual view.

And so, I think for us, what was interesting about this is we then looked at some of the key metrics, and there’s an important point of this paper, we don’t care about returns in this paper. We care about tracking. We care about goodness of fit, we care about replicating similar to the index.

So, for the investor, there’s still going to be substantial tracking error to any index because of the nature of what CTAs do has a lot of variation, but you can still try and sort of reduce that tracking error by combining both of these techniques together. And I think that’s what we really wanted to show in this rather simple paper. But it does it at least in a technical and sort of more intuitive way.

[00:47:00] Niels

Yeah, I mean, what I noticed from the chart in the paper, which I guess I’m not surprised and given what we have talked about many times on the show, when you do index replication, you have to just react slower because you don’t pick up these changes in the performance until kind of after the event, so to speak.

And so, what you show in your paper is that the correlation between index replication and the index itself actually kind of breaks down often when there are some big dislocations happening in the space, which on one hand, again, it doesn’t really necessarily say anything about whether that’s good or bad in terms of performance. But if you were buying the product as something that you really wanted to replicate closely what the index is doing, I imagine that the correlation is relatively important.

And therefore at least you should be aware of the fact that there seems to be this tendency for the correlation to break down when there is a big, big event or big movement in our space. And then compare that to the more mechanical way of doing it, which has less or fewer events where it breaks down.

And then as you say, it seems like when you combine the two, you kind of offset some of this weakness and you get a closer tracking from a correlation point of view to the index. Is that how it should be read, your paper?

[00:48:31] Katy

Exactly. And I think there’s some intuitive sense to that is that both of the methodologies can have breakdowns in different environments. And I think that’s what we were interested in is this idea of that if you do a little bit of both of these things, you combine the insights of CTAs and design your own strategy and risk management. With index replication type methodologies, you get something that is going to lean into the industry when the industry is doing something different from you. And then that may lean into your sort of CTA strategies when things are changing quicker than the index is able to adjust.

And so, I think it’s not surprising that the combination of both methodologies gives you something that has lower tracking error. And I think for us, even though the tracking error is still high, I mean, you see a substantial reduction in tracking error over longer time horizons. And, in this case, if you’re looking for something that’s CTA beta, having a little bit of both of those makes sense.

And you notice this approach is a little different than sort of slapping these two things together, because you might do that as well, in a simple way, where you buy one replicator and one index itself. We didn’t compare those two, but, wow, that could be a fun paper. So maybe we should compare that in the future.

But, you know, what we were realizing is just that if people really want to reduce tracking error, it seems to help to do a little bit of both.

[00:50:09] Niels

Yeah, I mean, I think it’s fair to say people, probably, who listen on the show to many of my conversations, certainly with Andrew, I’m always the one that’s kind of a little bit of a skeptic and on the fence. And maybe it’s also just because I’m trying to “defend the history of trend following”.

But, I will say that when I notice some of your numbers here and the tracking error that index replication has, according to your data, I guess the whole idea of replication, I think the word for me is a little bit loaded. It makes it sound like you’re going to get “the index”, we’re replicating the index, but with a tracking area of this size, then people should be aware that that’s probably not what they’re going to get. They’re going to get something that might be better, but it could also be worse, so to speak.

[00:50:59] Katy

Yeah, I think. I mean, I’ve had a lot of fun conversations. I had some fun conversations in Miami about this too. But I mean, my view is more sort of that I don’t see it as you’re never going to replicate in the way that you think about equities. And maybe we should just not use the word replication. That’s fine. Let’s just say that we want to use the word beta-like, okay. If you’re looking for something beta-like, I mean, there are pitfalls to that. But, you know, you can, if you’re not aware of what the industry is doing in your direct implementation, you can do something very different.

And I think that’s where the pitfalls are. And you were actually already asking me questions about this earlier in that, you know, if you go in and you randomly pick an ETF, you could get something very different from the industry. And so, I think if you are coming to the ETF space with a beta hat on at all, then, you know, you do need to actually think about like what methodologies can give you the most beta-like.

I’m not saying you can get a beta in the classic sense, but I think if you buy one individual strategy, you could get something very, very different. And I think that’s why, you know, sort of a beta exposure is still relatively an important goal. And there are investors that come in with that goal.

And that’s why we cared so much about tracking error instead of looking at returns. Because as you know, you can always pick the strategy historically that has the best return. What’s interesting to me is which methodology is going to give someone the most beta-like if that’s what they have their mind on, if that’s what they’re looking for.

[00:52:34] Niels

Actually, it leads me to a question that I wanted to follow up with as we have a few minutes left, Katy, and that is, you know, we have this rush and by the way, let me just say, so I don’t forget, you should definitely go to the website of Alpha Simplex and download the paper. I’ll put a link in the show notes so you can do it easily from here because it’s definitely worth your while to read.

But let me then go on and say that there is this rush into the CTA, so called, sort of ETF space (Not so called, but it’s CTA ETF space) at the moment. And you hear the word CTA or managed futures being used a lot in all of this. And maybe it’s just because I’ve been doing this for a very long time that I have a certain expectation. I guess, when I hear that. And maybe I should think more about innovation and all of that stuff in a space, and I accept that. So, just take it for what it is.

But I have noticed, and maybe this is a paper for you to write, Katy. I have noticed that there are some very different types of products now out there, but they all seem to be using the same kind of terminology around them. And I wonder if some of them really, you know, resemble what we normally define as a CTA product.

Now, as far as I can tell, most of the products out there, they all trade futures. Okay, that’s fine. That’s kind of how our industry started. And most of them, if not all of them, I don’t know, I haven’t done full due diligence on this, but I imagine most of them are using rules and algorithms to be implemented, in a systematic way. So, I think that ticks the box as well. That’s fine. That’s kind of how we define our industry.

But then comes diversification. This is where it gets a little bit interesting to me because I think of our industry as an industry that really offers people access to a wide range of markets. That’s just how I think about it when I look back. Of course, in the 70s we were limited by markets, but as financial futures became available, we added them to our commodity futures portfolios and it became a fully diversified portfolio.

What I now see, in some of the products that has come out and become very popular, actually, is that it’s not diversified at all. It’s actually some of them are trading one or two, at least two sectors. And so, in my opinion this is where it becomes a little bit difficult. Unless you are kind of an expert and you look inside and you say, well hang on, it’s not exactly what I expected from a “normal” CTA managed futures portfolio. You’re only trading a couple of sectors that mean yeah, performance could be better, it could also be a lot worse, but most likely it’s going to be very different in any event.

So, I’m kind of curious as to, if you were to draw some distinction between what is a managed future CTAs product like that, some general definitions that we can kind of put our hands around and say, okay, this is what we mean? What does it mean to you?

[00:55:55] Katy

Oh, this is a fantastic question. Because it’s something… You know, we’ve actually already started working on a paper on this because, really, the ETF space is in the nascent phase or the innovative phase of the vehicle. And you know, that’s what we wrote about when we were talking about it in our other managed futures paper. At the beginning of something there’s always all these interesting ideas and they’re very diverse and can be very different from where we end up.

And so, I think what’s interesting right now is a lot of things that people have wanted to do. Things like portable alpha methodologies and managed futures and asset class specific approaches that don’t trade equities and other things. They’re all showing up there. So, you’re ending up with beta-like products versus direct implementation or mechanical replication products.

And so, what’s going to happen is, I think with time we need to sort of have better nomenclature to compare these because they’re like apples and oranges. It’s kind of like a fruit salad right now. I guess that’s a good analogy.

And so, I think it is actually quite difficult for an investor universe to understand the sort of differences between having a portable alpha strategy with managed features versus having, you know, a sort of an index replicator in the same bucket. Right. And so, I think one of the things we’re hoping to do is maybe write another article to explain some of the distinctions. But I think it’s going to take a while for the industry to kind of get to that point where people can understand the differences between those products. And I think time and sort of use cases will evolve those into different buckets as well.

I mean, if you think about our space, we have buckets that we like and we understand. So that’s, you know, you have your trend, you have your multi diversified trend, you have your short-term traders approaches, and then you have commodities. So, we have those distinctions because we’ve been around for that space has been around for a long time.

I think the jury’s still out and I think it’s exciting because, you know, there are a lot of things that we can try. That doesn’t mean they’ll all work out. I mean, take a look at what I loved about that paper, which was so interesting to me, about the managed futures ecosystem. Look at the correlation to the index of the top mutual funds. It was very low for the first few years of the 40 ACT space.

That’s because there was a lot of stuff that wasn’t necessarily, it was widely ranged and people weren’t necessarily possibly knowing what it was. And eventually it evolved to having a very high correlation with the industry itself.

So, I think that’s what you’re probably going to see is happening. It’s just right now we’re kind of in a phase where we might end up with different buckets. So, maybe a portable alpha, maybe you have beta, maybe you have… The key distinctions are the goals, right? So, is it a portable alpha strategy or not? Is it a beta strategy or not? Does it have a different asset class exposure or not, are the key three that I see. I watch the space closely and, on a weekly basis, I’m examining how different managers in the space are doing and those are some of the themes that we’re seeing evolving. So, I think it’s exciting. There’s going to be a lot of stuff to talk about.

[00:59:16] Niels

Yeah, absolutely.

I have one maybe final question and that is, I’ve heard you mention, when talking about your paper, that the mechanical replication is often simplified versions of a full trend following system, so to speak. I’m curious, does it have to be a simplified version, so to speak? And why would people generally use a ‘not the best ideas’ type strategy in the ETF space?

[00:59:51] Katy

Well, I think one of the things is it’s a strategy where, in the ETF vehicle, with the daily liquidity, you might choose, for example, less turnover in signals and you might also lean to markets that are easily hedged by market makers. So, there is that additional layer of daily liquidity, intraday liquidity for ETFs that does distinguish that from a classic SMA program or a mutual fund. Because think about it, with a mutual fund you just have a daily mark to market really.

And so, you know, if you’re trading a wide range of markets, you don’t have to think about that interday hedging necessity. And I mean, I wish Jerry Parker was here right now because, you know, he is obviously would be the one that knows more about things like trading orange juice and ETFs.

But, you know, I think right now, sort of in that wrapper, it naturally fits well with a very highly liquid set of markets and you know, not too much turnover as well. So that it can be easily… You know, you don’t want too much slippage on the pricing as well, for those that purchased the ETF intraday.

[01:01:07] Niels

No, no, that’s fair. That’s fair. Katy, why don’t we wrap it up?

It’s very early morning where you are in Arizona, so I really appreciate you getting up so early for this as I’m sure all of our listeners do. And I’ll save the active/passive topic that I have also prepared for another time.

And if anyone listening out there also wants to show the appreciation as much as I do for not just getting up early this morning from Katy, but also all the work that goes into producing these wonderful papers that we can review and talk about and helps us all learn more and understand better what we’re dealing with in our industry. Then feel free to go to your favorite podcast platform and leave a five star review for Katy for this conversation. We certainly appreciate it, and we do read all of them, by the way.

And next week I will be joined by Cem. So, another exciting conversation, I’m sure in a slightly different space. But if you have any questions for Cem about his thoughts on the current environment, then by all means you can email them, as usual, to info@toptradersunplugged.com.

From Katy and me. Thanks ever so much for listening. We look forward to being back with you next week and until next time. As always, take care of yourself and take care of each other.

[01:02:34] Ending

Thanks for listening to the Systematic Investor podcast series. If you enjoy this series, go on over to iTunes and leave an honest rating and review and be sure to listen to all the other episodes from Top Traders Unplugged. If you have questions about systematic investing, send us an email with the word question in the subject line to info@toptradersunplugged.com and we’ll try to get it on the show.

And remember, all the discussion that we have about investment performance is about the past, and past performance does not guarantee or even infer anything about future performance. Also, understand that there’s a significant risk of financial loss with all investment strategies and you need to request and understand the specific risks from the investment manager about their products before you make investment decisions. Thanks for spending some of your valuable time with us and we’ll see you on the next episode of the Systematic Investor.