Welcome to Top Traders Unplugged. In markets success doesn’t come from predicting what happens next, it comes from being prepared for what you can’t predict.

In each episode we go deep with some of the world’s most thoughtful minds in investing, economics, and beyond to understand how they think, how they prepare, and how they decide, and the experiences that shaped how they see the world. No noise, no short-cuts, just real conversations to help you think better and invest with confidence.

Welcome and welcome back to this week's edition of the Systematic Investor series, with Yoav Git and I, Niels Kaastrup-Larsen, where each week we take the pulse of the global market through the lens of a rules-based investor. Yoav, it is great to be back with you this week. Hope you're doing well. How is life in the UK?

Oh, Cambridge is sunny, but we are getting a little bit of water, a little bit of rain hopefully. And I'm sitting here sipping my Indonesian avocado coffee that my daughter, who's back from Indonesia, made for me. So, life is good.

Yeah, certainly sounds like it. I'm not sure I would pick the same coffee but I haven't tasted it. So, you never know.

We have to try new things.

Yes, try new things, that's absolutely true. So, now we've got a few new things on the agenda, some great topics that we're going to be talking about. We're going to announce the winner of the very important TTU merch competition that was running in August. So, I'm sure there'll be some exciting news there. And yeah, and then we got some papers to review as well, so all is good.

However, before we get into any of that, as usual, I'm always interested in finding out what kind of has hit your radar, your desk in the last few weeks that you found interesting.

Oh, hi Niels, it's always nice to be here. So actually, I want to talk a little bit about clients, not about current events or anything outside. Actually, we've had a few interactions with investors in the last month, and it really brought home about the nature of the relationship we have with different investors.

So, we had two very distinct meetings. The first one was in Zurich, and we were discussing allocation to trend, and the guy said, oh, you shouldn't worry about it because Claude wants us to have more trend in our portfolio. So, you know, we've been advocating TPA. It makes sense to be part of a portfolio.

And so, there's a little bit of like total portfolio approach which is coming along. And I think AI, when you give the data of allocators, they kind of realize Claude really wants trend. So, I think actually this is really exciting for the industry. I think that's being dispassionate about, you know, what is actually best for the portfolio. It's really interesting.

So, that was like the first interaction, you know, a very technical discussion. And then the second one was much more personal. So, we had a very long-term supporter of the ACoS strategy, the commodity strategy, and he’s retiring. He’s moving off to spend more time with his family, and I thought it was a really emotional meeting. And I wished him all the best, him and his family. And it really brought home to me how much the nature of the investor relations is really about trust and about the relationship. It's not just, you know… People allocate to you not just because you are running a CTA, but also because they trust you. So that was very… I didn't realize how emotional I would get.

I mean, it's interesting you say you bring that up because obviously that's the side of the business that I've been working in for more than three decades. And so, I completely agree. Of course, I think it's a very personal business or side of the business, very important. Often, of course, as you say, people tend to start by screening just for numbers. But I think once you get into the conversations, the due diligence, and ideally once you start having a business relationship, I think that it changes a lot.

And I also find that, I mean a lot of it is also about the journey, right? When you've been through a few difficult times with a client and they have felt supported along the way. You've given them the tools and the information they needed to essentially, you know, justify the investment to their investment committee or their colleagues. Once you build up that relationship, that trust, as you mentioned, and of course, as long as you don't screw it up completely on the performance side.

Yeah, absolutely.

I think it really solidifies the relationship. And it is often, I think, maybe part of a business that may not be as appreciated as it probably should be. Of course I'm biased when I say that because I think a lot of maybe (and this is a little bit unfair), I think a lot of people who are on the, say, the strategy side feels, oh, they're the most important because without the strategy there's no clients. But on the other hand, I think we could argue whether if there's no clients, there's no strategy. So, it’s a really…

It's a two-way process. It's really about conjoined response. I think the important thing is that you are actually working in partnership with business development in the investors in order to, as you say, they need to justify to themselves, and they need to justify to their IC. And you will go through periods where you are in a drawdown or performance isn't what is expected and they will come to you and you need to be able to actually explain to them what's going on and your thinking about when is the strategy working, when it's not working.

So, that responsiveness and that sort of direct relationship with clients is something that I haven't done until I came to Gresham and I really, really appreciate it. I have to say, it's really wonderful. It's an eye-opener about how the business is actually, as you say, running. It's not just strategy.

Well, somewhat related to this, and I have to be careful. I don't want to be… I don't want to say things that, you know, are not fair or reasonable. But I did notice over the weekend that suddenly there was a lot of chatter on X about one of the CTA ETFs out there. Actually, quite a big one - the Simplify ETF called CTA. I mean, it's got a great ticker. And I saw some people talking about it saying, what's going on here? All the positions are being wound down, there are only a few big positions now left in oil, but there was no real news.

Then when you looked around a little bit further, you know, some of the portfolio managers had left Simplify recently, Mike Green one of them, and some of the people who worked on the business development side. But again, no real news, no real announcement. But then someone shared with me, you know, the official announcement that then came out a few days later saying, oh, great news, we're changing the strategy. We found something that's better.

I bring it up because, partly because we just had the sub-advisor on the podcast, Altis. They're no longer the sub-advisors, as far as I understand. And as I said, they may well have found something that's much better. And you know, that's great for the investors, of course. However, two things, one, as you kind of suggested, yeah, communication with your investors is pretty important. So, that's one thing I would just mention. But the other thing is in the ETF land, we, I mean, obviously we have people who are great supporters of ETFs on the podcast. I have my biases, of course, as everybody knows. But one of the things that maybe one should be aware of in investing in these ETFs is that suddenly big changes can happen underneath the hood without the investors really knowing.

And even if you get a letter a few days late saying, oh, we're changing, you know, that's… do you really know, then, what exactly you've invested in, and so on, and so forth. And I do highlight that a little bit, even though I may get some angry emails, because I think that's one difference when you go and you invest with an actual underlying manager where they're not suddenly going to completely change the strategy and certainly not without giving notice to the investors.

So, yeah, I'm just putting it out there. Not necessarily… I don't know if you want to comment on it or not, but I just feel it's an important difference to highlight that when you have these structures that are so convenient for everyone. Well, they're also very convenient if you want to change the strategy completely. So, you know, that's my biased comment, of course.

So, I think it is important that you know, if you are doing a CTA, you are actually a CTA. But I think your point about communication, I think my COO's favorite word is “sequencing”. If you want to implement a change, you speak to your clients first, you make sure that they're on board with it, you explain to them what you are doing, because changes happen. Right? You know, you raise volatility targets, you reduce volatility targets, you change, you start trading another strategy. All of these changes, you just basically need to keep them abreast of what you're doing, why you're doing them, and they will be on board.

But listen, let's be transparent about this or whatever the word is. I don't think, if I went to my clients and said, oh, listen, we're going to go and completely change the strategy, that they would be on board with that. I just don't think that's going to happen. So, I think there are nuances here where you need to be careful in terms of what you mean by changes - small tweaks, small improvement, for sure.

Actually, I don't know what's happening. It completely went under my radar. So. Yeah, I haven't heard of this so, I have no good comment about that. But yes, a CTA is normally a CTA.

Yeah, well, and anyways. All right, let's leave it at that. The other thing I noticed, and this was something someone mentioned, you know, just a comment, but I did find the source story, which was in Hedge Week. This Bobby Jain had been in the news about him setting up his own fund, etc., etc. And then later on, I think, he moved into the Millennium stable and is managing money for them.

But there was this little notice out in Hedge Week saying that the firm's gross gains since launch equate to roughly 34% of the capital raised, while investors received about 8.4% in net gains over that period. And I thought, hmmm, that's kind of interesting. And I know there's a lot of caveats. I'm sure there's a lot of information in detail, but still, those are the kind of headlines that don't make some of these strategies look great.

I think that was part of the rationale for the merger with Millennium. Right? So, the idea is, these days, setting up a multi strat is an expensive business, both in terms of data and in terms of infrastructure, let alone expenses of what you pay as a passthrough to your underlying pods. So, there's a lot of friction, let's put it this way. And I think it makes sense for him to have merged with the Millennium operation.

No, it certainly sounds like it.

And hopefully they will get a higher share of the gross profits as a result.

Yeah, let's hope for that. Anyways, it is time to announce the winner of the review competition, and who gets the TTU merch.

Should I do the drum roll?

I did not see any reviews from you, Yoav. But the good news is you already have a very prestigious TTU vest already. But before we get into the usual trend update and topics, let me just… Allow me a few minutes just to talk a little bit about that. So, for people who know nothing about what I'm talking about, it is because we had, in my opinion, one of the best conversations about kind of trend and why these strategies make sense, between Dave Dredge and Rich Brennan, towards the end of July, I think the episode came out.

It really is a must listen in my completely biased opinion, but we said, you know, this is great. For those who leave a really good review, which of course supports the podcast, let's give out one of these vests which, you know, right now is really only something that I created for the wonderful people who work with the podcasts and of course all the co-hosts.

But of course, once I start looking at these submissions and the reviews, it becomes very hard just to pick one because there were some really kind ones in between. So, I decided to pick a handful. So, the good news is that the winners are (and I will email you if you are among them and we'll figure things out), but the winners are Todd, Adam, Anton, Tim and Nikolai. So, those are the five winners. And yeah, first of all, thanks for submitting some really wonderful reviews.

I don't think I'm going to read them. I mean it would be, you know, a little bit too embarrassing for myself. But they are quite nice and positive so, no need for that. But I also want to say a huge thank you to all of those who left reviews and who did not win this time around because it really does mean a lot. And it's also a way to show (as I always end up episode saying), it's a way for you to show the appreciation for all the wonderful co-hosts that spend a lot of time in preparing for these conversations. So, there we are, some great news and we will have five more people around the world wearing the TTU vests.

Anyways, back to the regular programming as they say. So, performance was pretty broad based in August, which is important. Of course, contributions from equities and metals did the heavy lifting but actually there were quite a few sectors that participated which is pretty good. And I was actually curious from your perspective. You obviously trade some alternative fixed income markets, for example, that I have no exposure to and don't really follow. But you know, interest rates are certainly part of the conversation at the moment. And I was just curious, Yoav, if you could give us a sense of where are yields around the world, sort of broadly speaking, where are they? Because we always only really talk about the US but that's just one out of many countries.

I'm not going to do a full review, I think, but the stories the markets in our portfolio that we are watching… So, Japanese rates are the gift that keeps on giving, if you're sort short JGBs…

Where are 10-year JGB yields now, roughly?

Roughly 2.8% if you're in the swap book, the bond book is slightly different. But you know, it's the long dated which is actually at an all-time high. So, really quite impressive. Canada got affected by the tariffs so, we are seeing changes there as well, actually on the long side. And then Malaysia, there's a little bit of a story, and China. But really, I think Japan is the sort of the main worry at the moment because they're financing the debt. They've always been easy to finance the debt at 0% but now that the GDP… the debt to GDP ratio has always been high, it’s at 250% or thereabouts. Now that yields, the long end is sort of approaching the 4% level, that's really kind of painful.

Yeah, it certainly changes the appetite for foreign bonds, like US bonds, I would imagine, in that sense. And the other thing I noticed, for the month of August, is that I think, relatively speaking, even though there was a little bit of, how should I say, news that affected the fixed income markets in the US with Scott Bessent's announcements. I mean, generally I think fixed income exposure in the CTA land was really relatively stable. Maybe equities exposure reduced a little bit because of you know, a slightly weaker momentum, if I can call it that. But, on the other hand, given the new conflicts, you know, coming back again in the Middle East we have seen some firmer, I think, exposures in the energy sector. At least that's kind of how I see it at the moment. But as you say, there probably were a few single markets that…

Well actually, in commodities we're seeing, across the board, very strong performance. The book is relatively long elsewhere, everywhere, and we're seeing things in anything from agriculturals, we are seeing it in electricity, we're seeing good performance in freight. It's been a very strong month. And the nice thing is about the cross sectional, the fact that pretty much every sector seems to be sort of earning its keep so to speak. So yeah, very good.

Fantastic, good stuff. So anyways, as of yesterday my trend barometer finished at 52. That's a pretty strong reading and that is very much in line with what we see in terms of performance. I'm not going to talk about the first couple of days of September, but the August numbers were as follows.

BTOP 50 up 2.95% for the month, up 11.14% for the year. SocGen CT Index up 2.49% for the month, up 10.79% for the year. SocGen Trend up 2.77% for the month and up 10.9% for the year. And even the Short-Term Traders Index, very strong month 2.44%, and up 5.7% so far this year.

Now in the traditional world, MSCI World up 2.48% (that is actually as of last night), and up 12.13% year to date. The… (where are we?)… the US Aggregate Bond index is down 20 basis points and down 40 basis points for the year. And the S&P 500 so far in September down 25 basis points, and up 11.99% so far this year.

Now, we actually had three questions come in for you, Yoav, so let me try and do that. The first one is from Jose and a good supporter of the podcast, so appreciate that. Now he did boil the question down to just a few lines. So, I'm only going to read out the few lines because there was a longer explanation to it. Anyways, the question is as follows.

Does trend following pay too much attention to correlation and not enough to variance and covariance at the instrument level? Given that correlation tells us how stocks spread, while variance might show where stress is becoming most acute?

Oh, that's a great question. I've got to tell you, as a mathematician, I love variance. I love correlation. We love them both. You know, we have no privileged children in our book, so I just want to highlight where each one of them is fitted. At the end of the day, correlation is important in terms of how it feeds into the covariance, actually.

So, a lot of the CTAs, most of our trading strategies are univariate, i.e. You look at the price of the asset to predict that asset. Right? So, it's not necessarily cross sectionally. So, correlation features in several places.

Let's take the easy one which is volatility. So, we look at a single market, we would basically like to trade in units of risk. We obstruct the asset and we say okay, given our position, what is the number of units of risk? And we like to work in sort of units of risk. So, we would generally both scale the signal in terms of Z scoring the trend. So, we will take the trend but then divide it by the volatility of the asset to get a sort of a Z score. But we would also, more importantly, we would actually divide our position by volatility. Not just volatility, by the way, FX, or the size of each contract, just to make sure that we get a constant unit of risk. So, if you're holding a bond, in my case, it's the duration of the bond, which is important and it's just sort of the measure of risk. So, that's what volatility is usually used for.

Correlation in your portfolio is used in multiple places. I mean, the first place is really about the overall portfolio risk. Right? So, if I've got all the individual assets and I've normalized them so that each of them is a unit risk, what is the overall portfolio volatility?

And that's really an important question because most CTAs would manage their book to have a certain level of volatility per day or per annum. So, say, 15% per annum, maybe 1% per day. And you have to look at the correlation right now between the markets to size that position appropriately. So, that's why we kind of care about correlation.

But the other aspect of it is that also gives us the diversification. So, if we think of the portfolio, and we have 100 assets, and the correlation takes the volatility down from, you know, instead of being 100 units of risk to 10 units of risk, then what we have done is we have reduced by 10x our volatility, but our expectation has changed, has stayed the same. So, the square root of the variance, the volatility, that portfolio diversification is what actually gives us the performance of CTAs.

So that's why we kind of care about it. It's not so much that we use it as a predictor between two assets, it's really thinking about where the risk is of the overall portfolio and how do we manage that to give the investor the return that he wants, the volatility that he wants.

All right, the next question is from Vladimir. Vladimir writes, when a systematic process permits a human risk veto, what evidence should be preserved so the override can be judged separately from the original signal? For example, would you keep distinct records for a model forecast that authorized a position, the order that reached the broker, and the later outcome? Combining them may hide whether the model failed, the control level layer intervened, or execution changed the result. Separating them may also invite too much post trade interpretation.

I would value your view on where the record keeping becomes useful governance rather than an unnecessary complexity. That's a good question.

So, one thing is that we don't really intervene that much. So, it really is important to understand that we're not being paid to be risk managers… not to be risk managers, to make a macro call, to make judgment. That's not what we are paid for. We are paid to monitor the risk and make sure that the risk is correct.

And most of risk intervention that we do is to avoid left tail and actually cost money to the investor. So, we do it very rarely. But actually, for your question, it's important to understand that even the risk intervention is systemized.

And what I mean by that is we would never then tell the traders, oh, don't trade this. What we would do is, there are levels that we already set within the simulation that will tell the system, don't increase this position, right. So, the system can load from a database an intervention, and then it will trade the way it trades, it's incorporated into the code.

So, it's never a case that we manually override something. We program into the system, into the simulation interventions, the interventions that we are likely to do: reduce the risk of the portfolio, reduce the risk of a certain asset, stop trading, stop increasing, all of these. You pre-program it, so, you're always able to play back because the process is, I think there is the Hormuz crisis in Iran, so I think maybe we shouldn't increase the positions in oil. It will go through the CRO, we will get an approval, and then we would basically add a new line into the database which is this intervention. That would be loaded by the simulation, and the simulation would know not to increase positions in oil. That's kind of how it works.

It is really important, as you say, in terms of governance and later auditing what's happening. We definitely also monitor, by the way, the execution. So, sometimes if you get a partial fill, we need to know about it. Any drift between the simulation and the actual position we have to monitor. And we need to understand how much risk that amount is, and what happened to that, and why that is happening.

So, all of these are really important. But I think it's important to understand that the simulation is a tool that does many things and it's important to separate what they do. And for each question that you ask, it will be a different answer. So, like we ask, we look at the simulation as one way to look at the simulation is to generate what should be my position today? Okay, that's really important part of the simulation.

But we also use a statistical question which is, you know, does this code change actually makes sense? And there you have to answer separate questions. There is a difference between, is there gross alpha in this signal? And, is there difference between is correlation calculation correct? And there's a difference between what is the trading cost? Can we actually bear the costs of a particular, say, fast trend predictor with this market? All of these questions are separate.

The simulation is not just one tool. It is designed so that you can tweak it and you set up the parameters of the simulation to answer the specific question that you want to answer. And that's something that people don't really appreciate.

Yeah. And of course, all firms do it differently. I think for a long time, I think generally there was a school of thought among established CTAs that it was very rare that you would intervene, however, and I mean discretionarily change things. I know for sure we certainly don't do it over at DUNN. But I do know that some managers came out after, say, COVID and were very open about it, said that yeah, we did intervene, and we did reduce the risk, and so on, and so forth. But I think as long as people are transparent about it and, as you say, probably a good idea to, you know, pay attention to it, and log it, and all of that stuff, it makes sense.

All right, final question from Bandassian. You talked about different market participants leaving different footprints along the curve. Where today can you most clearly see hedges, CTAs and discretionary macro players fighting each other?

Oh, I love this question. And the reason is because this is follow up for the question last time that I was on TTU. And I think it was Marcus who asked exactly this question about different points on the curve. And I think we also have a paper exactly addressing this question. But I'll just answer it generally. It really depends on the commodity market and the liquidity in the market.

In general, where you will find most of the macros, and the CTAs, and what I would call the financial players, you will find them in the front contract. This is the one which is most liquid. You will see physical players playing in the spot market. So actually, cash, cotton, or cash wheat, or cash whatever. And then there will be producers and consumers that will be hedging out or further out the opposition, further out on the curve. So, if I'm an oil producer, I'm likely to sort of forward sell my oil sort of towards three years, five year points. So, that's kind of the broad, a broad brush stroke picture of who's playing what.

Okay, very good. All right, now we are back into the topics that we usually bring up and there are a couple of main papers I think we want to talk about. And I, as I mentioned to you, to be completely transparent about it, it has month end, we're recording very early in September, so it's been pretty busy on my side. So, I'm going to let you lead the way through both papers here. But the first one is of course a classic, if you have come from the trend following world, and let me just point out there was another Dane involved.

Absolutely. I love him.

So that makes it even more special. Maybe we should ask him to join the podcast one day if he's interested. Anyways, we are of course talking about a century of evidence on trend following investing, a very famous paper. I know Lasse Pedersen was involved. Who are the other authors? Do you remember?

Oh, I can't remember. You know, people from AQR.

Yes.

Which is really embarrassing because I really should know who the people are, but I don't remember.

I think Hurst was another one.

There's Brian Hurst and there's a Yao Hua Ooi.

Yes, I think those are the three. Yes. Anyways, over to you my friend.

Oh yes. So, this is a classic paper, and it came out, and it's really like one of those papers that come out and you think, damn, I really wish I wrote it myself because they've done a great amount of work. They've basically taken, created a 137-year data set.

And the reason I'm putting out this paper is because it's been essentially 10 years since they've done it. So, we've got another decade in the mix. And I thought, oh, let's have a look. By the way, there have been papers afterwards that have criticized it, which we can address as well. But I think one of the important aspects is the sheer body of evidence supporting trend. Right?

So, I think my colleague Tom Babbage, he wrote a blog recently (he's an astronomer), and he's looking at the way that you examine distant galaxies, and you have an image which is very noisy and it's very difficult to detect the actual star. But what you do is you take many, many, many photos from different places and over different times. And over time the noise disappears and cancels out and the phenomenon appears.

And that's exactly how trend works. We've got an individual asset any individual year. The Sharpe is not great. In fact, the noise level, the signal to noise ratio is very similar for distant star and trend following is a signal. But as you diversify, in terms of the asset classes and in terms of the years, a clear picture emerges. And I think that's a really beautiful analogy that Tom uses. And it held up the good news, it held up another positive performance this decade. So, we've seen another 10 years have passed and trend has held up.

2017 was an outstanding year. But there have been good years also during the COVID time. There have been good years during the Ukraine crisis, during inflation crisis, pretty much in all regimes. Trend has actually done decently, and it's really delivered on what it was supposed to do. But I think it's, we can go deeper, not just in terms of…

Now, I know that we interchange trend following with CTA and of course that's not exactly the same. The paper obviously talks about trend following, so that is different from “the CTA indices” that some people follow. And I don't know if this is true or not, but even within “trend following programs”, maybe there's a little bit of non-trend that has snuck in during the periods where trend didn't work so well.

But anyways, I mean, I'm rambling a little bit here, but are there any thoughts, now that we sit 10 years later, towards some of that “criticism” that we've received?

So, I have to say for what you're receiving, you would normally pay for. So, you think about trend as actually buying protection. It is a defensive strategy. There are very few strategies out there which have both a defensive nature… They have positive convexity, in fact, if I plot all the HFR hedge fund indices, so normally you would kind of pay for that. You would expect to pay. And in CTA land we have very strong evidence that you've made money over the years.

So, in weird way I think CTAs are pretty much the only… not the only, but one of the very few phenomena that have been super consistent over many years. It's very understandable what's happening and actually provides you positive convexity versus sort of just tapping into sort of liquidity premia.

The funny thing is here, Yoav (sorry to interrupt), the funny thing is that I agree with you. Normally you should pay for this. Right? But in our industry it's almost become like, oh, you don't have to pay for trend. It's so easy to do. You don't have to pay for it. That's kind of the funny thing because it's actually…

No, absolutely. I think when I joined the CTA industry, the trend industry, then it was considered like an alpha. And of course, as the secret sauce has been, well, people know it and we are actually educating investors about it, then yeah, there are more entrants to the field. But I think, what really begs the question is that the AUM of the industry hasn't really gone up.

So, you know, it might be that it's gone over to the pods inside the multi strat, and so forth, but you know, QIS products that have been ETFs. Absolutely, and it shows you that it's about supply and demand. The reason why people are not willing to pay for it is not because they don't think it doesn't work. It's because they think, oh, I can, you know, I can get it cheaper. Okay, which is fair. Which is a fair point.

So, it's not just how good your product is. It's also the fact that there is a lot of competition for it. Right? So that's why, you know, the price of consumer electronics has been dropping. It's not because consumer electronics have been getting worse, it's just that. But you know, we're just becoming very good at, you know, competition. You can buy a Samsung, you can buy an Apple, you can buy… There are plenty of providers of, you know, a TV, a consumer TV.

There is a genuine phenomenon that we've seen in the last 10 years which actually the paper sort of hints at. So, if you look at, historically, a decay of performance, there has been a decay of performance. The markets were a lot more liquid, the markets were a lot less efficient as mentioned in the good old days. So, one would expect to see, you know, one would expect to see some deterioration over the years.

By the way, I'm pretty sure that if you were to run, if you were able to run, you know, relative value equity, some sort of long/short equity fund historically, which is very difficult to do, you would still see the same phenomenon exactly with any hedge fund strategy.

But actually (sorry to interrupt again), but I just want to challenge that a little bit because I think that's what a lot of people will say. Oh yeah, just look at the returns. They're much lower today. Yeah, but so is the volatility. Right? I mean I wonder, if you vol adjust returns for the drop in volatilities… I mean, when I started in this industry, I mean back in the ‘80s we had the CTA. Even the Barclay CTA index had a vol of more than 20%. Right? Now it's down less than 5%. So…

Well, personally, I mean, I measure things in Sharpe because we can provide pretty much any volatility that the client would want. Right? That's the whole point of what we do in terms of leverage. But what I would say is that 10 years is actually a very short time in terms of making a statistical decision, in a weird way. So, you will find if you're a new quant, if you're a young quant just going into the industry, they don't believe in trend because they haven't… The attention span is like two years. Okay.

But the point being is that actually in order to see data clearly, in order to see a certain phenomenon clearly, you first of all need to understand the fundamentals, secondly why this is happening. But also, the amount of data involved is substantial, which I think is why this paper came about.

So, I would say variation in performance in the last 10 years versus the historical (you can put it over the historical performance) is a little bit lower. Yeah, that's fine. Still positive, definitely. You know, faster… faster seems to have degraded in the last 10 years more than slower. Yeah, but you know, what can I say?

I think the one thing I actually do want to go back to is the correlation story and the diversification.

Okay.

Which is the question from Jose.

Yeah.

One of the findings of the paper is that the system performs well, trend following performs well when the assets are uncorrelated. And that's exactly what we're talking about. The more diversified your portfolio is, you can lower the volatility while keeping the expectations and then you basically level things up to get a higher return. And over 130 years, the more uncorrelated your assets, the better your performance is.

And that's something that people don't normally think about in CTAs. People think about us as crisis alpha. And in fact, we do things, we do well when there is a crisis alpha. And I think it's important to understand that the correlation of the trend P&Ls and the assets are not exactly the same thing. Right? So, we do like it, just like we had in August, when all sectors are firing for different reasons potentially. Right?

That sort of correlation we actually like in terms of the performance P&L. Market correlation we don't like, we like low correlation because then we are actually are able to leverage up the portfolio and get a higher return for the investor.

Absolutely. I think if you look at the variability in volatility in equity markets, it's enormous. Right? You can go through very quiet periods and then volatility spikes up. We get paid for that. And I think it's quite funny because one of the biggest criticisms… there have been papers criticizing this work and there's a famous paper which basically criticizes trend following. It says it's all about volatility scaling. Okay, the risk management is what the performance is coming from. And to me it's like precisely what we do.

The reason why we're able to essentially obstruct each of the markets and actually put them together in a coherent fashion is because we manage the risk and because we volatility scale to get a proper diversification of the portfolio. And it feels odd that people criticize, oh, you know, it's like saying the cake was only good because you put eggs in it, of course, but that's part of the recipe. Right? You can't criticize that. It's completely integral to what we do. So, you know, that harks back to the first question that we asked.

You know, other people criticize trend following because of costs. You know, they look at the simulation that they did and they said, oh, you know, the costs were higher that day. That may be the case. But again, simulation is about answering specific questions. And the question is, is the gross alpha there?

Okay, we can discuss later whether it's relevant today in terms of being able to extract at today's costs level, whether I'll be able to extract it. So, it's a different question. It's a different question to ask.

I think this paper has been a wonderful paper and I think investors, you're right, investors who actually care for long-term and they care about allocation for the long-term growth of capital, that CTA and maybe TPA and Claude will help it to do, it will actually allocate quite significantly to trend following.

I'm excited about Claude, I have to say, because I think Claude has some friends in the OpenAI world and any other AI because, as you say, when you take an unbiased look at these types of things and you want to build the “best portfolio”, you definitely need, as probably everyone realizes on this podcast, that you definitely need a meaningful allocation to this wonderful strategy.

Okay, anything else you want to mention here before we move on? By the way, I think our friends over at CFM, they did a 200-year paper. Maybe that's the one you need to bring next time.

Yeah, it's absolutely right. There was a period when people stretched back as long as possible.

I mean, I think did Katy and Alex in their book do 800 years or something like that? I can't remember. Maybe they did.

I think that what is really interesting is that I did it for other asset classes like, you know, option trading and stuff like that. Yeah, I mean you don't necessarily publish it, but it works. And the reason it works is because we understand why it works. It is the two reasons, the two effects that drive trend. The first one is behavioral in terms of people buy winners. And we understand that. It's exactly why people don't believe in trend because it has had a few sort of weak years. That's fine, they're literally trend following and demonstrating why trend following works.

The other reason, of course, is supply and demand, which is the mismatch between supply and demand is actually what drives prices. And the fact that price adjustment takes time, that sort of auto correlation in the price really drives why trend following works. It just makes sense.

Let's move to something different, but also about supply and demand as far as I can tell. Again, I'm going to leave you to lead our way through this one.

And Ilia Bouchouev is very illustrious, both a quant and a manager of sort of energy portfolios. He's published a really good book called Virtual Barrels, about the oil industry. And this paper is a short paper that he wrote which is amazing. I mean, if you're an oil trader or anything like that, an oil quant, you really want to read that paper and absolutely…

What was the name of the paper?

So, the name of the paper is A Stylized Model of The Oil Squeeze. And what is this paper about? So, it has like three distinct parts and is getting to the COVID crisis. So, the first one is reviewing the theory of storage. That's the first section.

So normally we have supply and demand in economic theory and what you have is inventories which make up for imbalances, and they can do that until you run out of storage. And at that point you switch to a regime where supply and demand really have to do all the lifting themselves. And then instead of price stops being sort of relatively small, movement tends to be very volatile, and that kind of predicts very strong positive skew.

Because the point where you run out of storage is the point where the commodity is in high demand, and basically the price spikes up because people run out of things. And we can see that in oil, for example, recently, when, you know, in Asia, people would pay US$200 to just have a barrel of oil because demand and supply in the oil industry is very inelastic.

Can I just, sorry, I want to make sure I heard this correctly. You say the time when things run out of storage is when it's in high demand, but actually with COVID it was the opposite, right?

Exactly, so, in traditional storage theory, when you run out of any inventories, stocks are going down. What happened in oil is potentially, it's not quite a story, but potentially on the other side, which is what happens when there's too much supply and you cannot store it anymore.

Yeah, okay.

That's actually one of the things he points out. That's actually not what happened, but it's the common story, people were saying. Okay, so that's interesting. But what he's pointing out is that, in this model, there is a right skew for the commodity. And we see that in agriculturals.

So, for example, we saw coffee spiking up, cocoa spiking up when there is actually a shortage of the commodity. But he says, well, actually, if you look at oil, it's the other way around. Oil is essentially, the volatility of oil is constant, whether it's 20%. If it's trading at US$20, the volatility will be around US$20 annually.

And if it's trading at UUS$100, the volatility will be around US$20 annually. In a percentage term, it's actually inverted. So, it's got a strong negative skew in many ways. And he's saying this doesn't make sense to us. What we really want to do is to skip the demand function altogether and we're going to model spreads as a direct function of the capacity utilization.

Exactly this point about storage, we look at how much of the storage is available. And on the one side, when it goes to zero, when capacity is not utilized, we have the spike going off to infinity. But on the other side, when capacity is fully utilized, then we should go to minus infinity. Right? The price should go to minus infinity.

And then we kind of use a bit of a hokey pokey magic sort of expectation, forward expectation, by saying actually we can think of the capacity utilization as a random variable which mean reverts. Okay. Which it does. Which is correct, and we basically think of integrating that, plus infinity at zero, and minus infinity at one, to the current levels. And you get a nice shape which is, you know, a nice spike near zero, to the upside, and the negative spike where capacity nears one. Okay. So, it's a really nice model.

It's a heuristic model and you can calibrate it, and it calibrates the model to actually fit the oil spreads that we see in the market. And then we can think about, he talks about COVID because COVID, what's actually happening there, we didn't actually run out of capacity. And he's talking about a specific incident about the Chinese retail investors and about the way they've been unable to essentially hold the roll during the roll. So, it gets complicated. If you want to read about TAS, you know, trade at settlement prices, that's…

But I think it's one of those papers where you know that this person is both a quant. So, he's able to come up with a model and he really understands the dynamics of the oil market. And it's a beauty to read. And I really love this paper and I'm going to qualify it because I've got a disagreement with it as well.

Okay, let's hear it.

So, let’s do the good things. So, he really understands the market. He understands the participants, the physical players. He really explains the carry trades. One of the things that I really have a bugbear with lots of quants in the industry is, and by the way, with his last paper in which he wrote about carry.

Okay.

It's, I think carry in commodities and carry in fixed income are very different beasts. Okay. Because money is fungible, you can move it easily from side to side. In commodities, the way that carry interacts with the inventories and with the capacity with the storage mechanism is really important. Like, if there is no oil, you cannot borrow it. Right? It's a fixed commodity. And really most quants really simply fail to recognize how carry trade is tied to inventories.

And in fact, you know, if you plot capacity utilization against the spread, against the contango versus backwardation, it's a very clear line. It's a complete sort of spread, it is a very good function of inventories. And he really understands the dynamics of inventories about the way the traders actually look at inventories now. But also look at what it's likely to be because supply and demand in the oil market is very inelastic.

So, if we have an imbalance this week, we are likely to have an imbalance next week. So, the fact that you recognize that, that's a really beautiful paper. And I think we'll come back to the COVID explanation is 100% correct.

We talked about GIV, about calibration of supply and demand when we talked about equity markets, actually, a few months back. And they've done this exercise and they've done it for commodities on an annual basis. They've looked at anything from bananas, to oil markets, to copper.

And what comes up is that oil is indeed quite inelastic. It has about a multiplication factor of 0.15. And what it means is that if we see a 10% rise in price, we will see production increasing about 1.5% and we will see demand, the number for demand, actually it's quite symmetric. We will see demand following by about 1.5% for a 10%.

And so now let's look at the oil industry as a whole. We started a run crisis with about somewhere between, depending on how you quote it, somewhere between 86 and 100 million of oil being produced every day. Okay. And we started the crisis with WTI roughly at US$60, a little bit above it. And what we know is that there are 20 million barrels of oil going every daily through Hormuz. And if you lose 10 million barrels, which actually is what we got, then what we see is, in order to balance that, we would need to see a 33% rise, which will give us a 5% decrease in demand and a 5% increase in supply. And that will make up roughly for the 10 million barrels of oil we lost. So, like the new stable equilibrium would be at US$80.

If you lose the whole 20 million barrels per day, the price would have to go to US$100 precisely to get the supply and demand sort of balanced. And I think what's really interesting is that this study, which was done by a bunch of academics from the IMF, based on annual data and you might think it's not accurate, actually came out pretty spot on. Because we've seen WTI trading in the US$80 to US$100 range. And I think it really shows you how understanding price elasticity is really important for our purposes.

Can I just, before I forget. I mean, yes, but… I mean, keeping oil… and I guess that is a little bit of a surprise that we haven't seen more than US$100ish, you know, in oil. I mean it did require a lot of reserves to being released.

Absolutely.

And you know, there are still, I remember our conversation a few weeks ago with Adam Rozencwajg. I mean it's like a billion dollars that's gone out of the market. And in his opinion, you know, it takes a few months before it shows up as missing. I know the Chinese stopped doing exporting of products, so that obviously helped a lot, but still…

So, what you're describing is exactly the events that bring… so, that price change is precisely… So, the Chinese stopped consuming oil precisely because it was becoming more expensive to them. So, the economic theory is actually being played out in the market, but reserves are actually a great point because what the US has done is essentially bringing more oil to the market, increasing the supply. They reduced inventories by about 120 million barrels, which is not insubstantial but also elsewhere.

So, I think that the discussion about inventories and about how these are managing that process, because what we are seeing is that the front of the curve has gone up. But if you go five years out, you remember we're discussing five years, the price is still at US$60. Okay. It hasn't budged. So, you know, I think it was pretty much spot on from what it was then.

If you look at the shape of the curve, inventories are allowing us to essentially to delay that shock. And that plays out in the shape of the curve, which used to be, you know, the front spread when the crisis started was something like US$0.40. It's now sitting at sort of US$3.00, US$3.50.

So, it's exactly that, the way that the crisis has played is through a much steeper backwardation of the curve in the front. So, we are seeing front sitting at sort of US$90 and the back sitting at the back of the curve five years out, still hasn't budged, still at around US$60. So, inventories are a very strong part of that story.

Yeah.

But there's one issue which I have with this paper which is there's a little bit of a magical switch in the paper because it starts talking about the oil inelasticity, which I completely agree with. If we just looked at it, it's a very small number, 0.15. But then the second part of the model, where he fits it, he fits the spread. It looks at capacity utilization and it is fitting the spread.

And the spread is a different beast. And in fact, the spread is actually a lot more elastic because then you have different players in the market. And I think it's really important to understand supply and demand not in oil, but in oil spreads. And that's actually where the action is because the players out there are a lot more elastic but they still obey economic theory. Because what you have is you have the tank storage, tank farms in WTI. There are, I think, about 70 million barrels of capacity there. They are a lot more dynamic to the shape of the curve.

For them, the price is exactly the roll yield, the difference between the front curve, the front price, the front contract and the next contract. And they are the suppliers of inventories. And then on the other side, you have the misbalance between producers and consumers, like refineries on the demand side. And the dynamic between them we can actually model very much like economic theory with supply and demand. And it makes perfect sense because they are very sensitive. The physical players are very sensitive to the level of inventories. Because I can only offer storage if I have storage left in my tank.

So, if the amount of storage I have in my tank sort of goes down, so, say in this crisis, I become more price inelastic. I will require a higher spread to release oil from my tanks. And similarly, on the other side of the equation, when we reach capacity, like in the COVID sample, I will require much, much, much stronger contango in order to store it in my tank. So, the supply and demand works perfectly if you think about it not as an oil market, but as a spread market.

But I think the spread market is really also very exciting in its own right. Because there's a bunch of people that trade spreads that don't even know they're trading spreads. And I'm talking about the CTA markets, the CTA people. So, I mean, if I ask you does DUNN trade spreads, and if the company clients would come to you, do you trade spreads, you will say no, of course not, we just trade outright, whatever.

I mean there are CTAs that specialize in this. There's Moritz and Mauritz who are very much into spreads. But broadly we just trade the outright oil. If I ask you if I'm a long only investor in oil funds in the USO, okay, are you a trader in spreads? You would say of course not. This is just long only oil. But actually you do trade spreads. And you trade the spreads, but only when you roll your position.

So, what you have is you have physical players who play the spread market, and they do it during the month, and everything is fine. And then we're getting to the point where we need to roll the contract and suddenly you have an influx of financial players and they are truly price insensitive. They don't even realize how price insensitive they are because they just don't want to get delivered the oil. The CTA will never hold their position, a long only position. And at that point what you have is you have a bunch of financial players coming in with excess.

If you think of it as a demand shock into the spread market and the physical players have to make up for it. And in normal operation, this is great. But when they are not, like in COVID, you have the financial players who are 100% must roll the position, okay? And that's exactly the incident that we saw in COVID. So, it wasn't the case that the physical traders were running out of space to store, but it was the case that the amount of demand shock that we saw in the roll wasn't matched by the physical player. And suddenly you saw the big spike in the contango and the price went negative.

So, I think you can really explain this as supply and demand, but you just have to be aware of this is not the oil market per se, this is the spread market.

Good stuff. I think we have to wrap it up, Yoav, because we are past the hour by some margin.

Can I just say something about trend following just for this because you know, we talk about the physical markets and like if you're a CTA, you're thinking, okay, this is a really nice story, but it's not relevant. But I think it is important for people to realize that, which is the first one, the first lesson is that for CTAs and for any investors, the roll is an issue. In most liquid commodities it’s not. But if you're trading alternative commodities or slightly less liquid markets, it's not just the daily volume, it's also the liquidity during the roll because you don't want to be short squeezed or long squeezed in the position. So, roll is really important.

And the second question is really when you understand the dynamics of the market, trend following in commodities is not necessarily a symmetric process. Right? There are some assets that will have negative skew, some assets will have positive skew. And you really want to think about how you want to handle it. And good luck with that. And I think Moritz and Mauritz probably know all there is to know about trend following in spreads. But I think it's really interesting in terms of understanding the dynamics of who are playing there in terms of understanding how to harvest trend in spreads.

Sure, sure, I think the most important thing, actually, my takeaway is that when people think about oh, trend following is easy because all you need to do is develop an algo. No, I mean I think we have to appreciate the fact that the experience of the people who implement the trades, especially in the commodities, is hugely important. And that is, as we talked about very early in our conversation today, there are so many parts of a CTA business that need to work together in order to make the whole…

Absolutely.

… the best it can be. And that also goes in this case. So great stuff, very interesting. I've certainly learned a few things here, Yoav, so very much appreciate that.

And by the way, even though the competition for leaving a great review is over for now, I will encourage you to do it anyways because it's a way to show your appreciation for Yoav and all the other co-hosts for the work they do and all the preparation. So, please go to your favorite podcast platform and leave a rating and review.

Now, I think that's pretty much it. You can, of course, always go and check out the website for the updates we do every week, blog post, and the trend barometer to follow along. In addition to tuning in each week. Next week I will be joined by Rob, who comes back from his very extended holiday that he always takes in the summer. No doubt he'll be full of energy and it'll be a very insightful conversation, I'm sure.

If you have any questions for Rob, please send them to info@toptradersunplugged.com and I'll do my very best to get it in front of him. That's it for now from Yoav and me. Thanks ever so much for listening. We look forward to being back with you next week. And in the meantime, as always, take care of yourself and take care of each other.

Thanks for listening to Top Traders Unplugged.

If you feel you learned something of value from today's episode, the best way to stay updated is to go on over to your favorite podcast platform and follow the show so that you'll be sure to get all the new episodes as they're released. We have some amazing guests lined up for you and to ensure our show continues to grow, please leave us an honest rating and review.

It only takes a minute and it's the best way to show us you love the podcast. We'll see you next time on Top Traders Unplugged.

This podcast expresses the views of its hosts and the guests appearing on the podcast as of the date of its recording, and such views are subject to change without notice. Top Traders Unplugged do not have any duty or obligation to update the information contained herein.

Furthermore, Top Traders Unplugged make no representation to its accuracy and it shall not be assumed that past investment performance is an indication of future results. Moreover, wherever there is a potential for profit, there is also the possibility of loss.

This content is made available for educational purposes only and should not be used for any other purpose.

The information contained in this podcast does not constitute and should not be construed as investment advice or an offer to sell or a solicitation to buy any securities or related financial instruments in any jurisdiction.

Certain information contained herein concerning economic trends and performance is based on or derived from information provided by independent third-party sources. Top Traders Unplugged may believe that the sources from which such information are obtained are reliable.

However, Top Traders Unplugged cannot guarantee the accuracy of such information and has not independently verified the accuracy or completeness of such information or the assumptions on which such information is based. This podcast, including the information contained herein, may not be reproduced, copied, republished or posted in whole or in part in any form, without the prior written consent of Top Traders Unplugged.