Markets often appear logical in hindsight, yet behave unpredictably in real time. In this episode, Niels and Yoav explore how recent events, from geopolitical shocks to shifting narratives, reveal the limits of traditional explanations. They discuss why assets like gold and bonds no longer react as expected, how flows and positioning can dominate price action, and why dispersion across managers increases during periods of stress. The conversation also examines the role of machine learning, the impact of volatility on execution, and the challenges of managing risk in fast moving markets. At its core, this is about navigating uncertainty when structure and narrative begin to diverge.

-----

50 YEARS OF TREND FOLLOWING BOOK AND BEHIND-THE-SCENES VIDEO FOR ACCREDITED INVESTORS - CLICK HERE

-----

Follow Niels on Twitter, LinkedIn, YouTube or via the TTU website.

IT’s TRUE ? – most CIO’s read 50+ books each year – get your FREE copy of the Ultimate Guide to the Best Investment Books ever written here.

And you can get a free copy of my latest book “Ten Reasons to Add Trend Following to Your Portfolio” here.

Learn more about the Trend Barometer here.

Send your questions to info@toptradersunplugged.com

And please share this episode with a like-minded friend and leave an honest Rating & Review on iTunes or Spotify so more people can discover the podcast.

Follow Yoav on LinkedIn.

Episode TimeStamps:

00:00 - Introduction to the series

01:19 - What is on the radar: markets and recent developments

01:19 - The EM algorithm and its relevance to machine learning

04:12 - Election narratives and unpredictability in real world outcomes

06:51 - Gold, war, and why markets react unexpectedly

09:07 - Inflation narrative vs risk off reality

14:13 - Bond equity correlation breaking down

17:29 - Fixed income volatility and regime uncertainty

24:58 - Oil curve dynamics and supply constraints

27:39 - Risk management differences between discretionary and systematic

37:23 - Market flows, positioning, and forced liquidations

41:30 - Excess volatility and what really drives markets

49:36 - Trend vs value vs noise traders explained

56:34 - Diversification, capacity, and portfolio construction

Copyright © 2025 – CMC AG – All Rights Reserved

----

PLUS: Whenever you're ready... here are 3 ways I can help you in your investment Journey:

1. eBooks that cover key topics that you need to know about

In my eBooks, I put together some key discoveries and things I have learnt during the more than 3 decades I have worked in the Trend Following industry, which I hope you will find useful. Click Here

2. Daily Trend Barometer and Market Score

One of the things I’m really proud of, is the fact that I have managed to published the Trend Barometer and Market Score each day for more than a decade...as these tools are really good at describing the environment for trend following managers as well as giving insights into the general positioning of a trend following strategy! Click Here

3. Other Resources that can help you

And if you are hungry for more useful resources from the trend following world...check out some precious resources that I have found over the years to be really valuable. Click Here

Privacy Policy

Disclaimer

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.

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's great to have you back this week. I was just going to say I hope you're doing well, but I actually know you're battling a cold, so I really, really appreciate you persevering through this conversation today because I know how it feels when you're not 100%, so really appreciate that. But it's really great to have you. So, welcome back.

Thank you very much. It's always a pleasure to be back. Even when with a cold.

Even with a cold, absolutely.

Yeah, but it's a great time because the markets are really exciting. So, it's a great time. There's a lot to discuss. There's lots happening.

There is a lot happening, without a doubt. So, let's talk about some of the things that are happening that are catching our attention as we normally do. What's been on your radar recently?

Okay, so before we get to the markets, I'm going to talk about something called the EM algorithm. So, this is a mathematical algorithm which was really the granddaddy of all machine learning. So, that's the idea of that it's an iterative algorithm which allows us two stages; one of them calculating expectation, and one of them maximizing the likelihood.

And the reason why I want to talk about this algorithm is because the person who invented it is a guy called Arthur Dempster. He was a math professor at Harvard, and he came up with that in the late ‘70s, and he passed away since we last spoke last month. So, it's been on my radar. And he's not a very famous person, he's not very significant. And within the war and all things that are happening, we tend to forget. But actually, a lot of machine learning is predicated on his work. So, rest in peace, Arthur.

Yes. Well, you call it the EM algorithm. I'm definitely sort of, what do you say, gravitating towards something with efficient markets, but what does the EM stand for?

So, the E stands for expectation, and the M stands for maximization. So, it's not emergent market, it's not efficient markets. So, the idea behind it is, suppose you have a lot of observations, let's say, from two distributions; one of them is red, and one of them is blue. Then I ask you to estimate the mean and the variance of each one of them. You would use maximum likelihood estimate to estimate the mean and the variance of the blues and mean variance of the reds. But what if I didn't tell you which ones are blue and which ones are red? So, you have a problem of classification, which is exactly what you have in machine learning. You know, is this a cat or is this a dog?

And the way you do that is you have a second stage which says, given that you've estimated this distribution, you look at each point and say, which of these two distributions is it likely? Is it more likely to be a cat, or is it more likely to be a dog? Is it more likely to be red, or is it more likely to be blue?

And based on that, you recalibrate, you do a second round of saying, okay, now that I've classified each of the points according to assign them with probabilistically to each one of these distributions, now we can recalibrate the reds and the blues distributions, and then again, which point is more likely to be in each distribution.

And it's an iterative process, a little bit like the back propagation process that we have in machine learning these days. And that's actually how you eventually sort of converge to a very accurate estimation of like classification without doing sort of any explicit determining of which points are the blues and which points are the reds.

Got it. Speaking of red and blue, actually what's been on my radar is something that happened yesterday. It happened in my birth country of Denmark. It was election day yesterday. And, of course, we're waking up this morning to one of those situations where the voters have basically given them a very, very difficult hand to find any agreement across parties, reds and blues. So, I'm sure I will be following that for the next few days to see how it all turns out.

What's been interesting about it, and why I find it a little bit relevant to even mention in this section, is that, you know, we talk about markets being very narrative driven from time to time and actually in the last few years, very narrative driven. And you would think something in like Denmark, given all the narratives around Greenland in the last six months, for sure that that would be a narrative that would be like front and center in an election campaign that started at the end of February.

Now we know, of course, that a few days after that election was called, a war started in the Middle East. That didn't even make it really through to the main topics of the whole election campaign. I'm not going to ask you this unfair question of what do you think became the main topics, because I'm pretty sure you wouldn't guess, unless you read the news of course.

It ended up being animal welfare and clean drinking water. Those were the main two narratives that actually pretty much defined the last two weeks of all the debates between the left and the right and even the candidates to be Prime Minister, those were the main ones.

And I just think it shows you how unpredictable things are, even at that level, that you think, oh yeah, definitely, global, geopolitics, and stuff like that will be front and center in a world where you’ve just been under siege, at least from a narrative point of view, about part of your kingdom.

No, it didn't really get mentioned whatsoever. So, I think that's a little bit of a lesson for us investors that it's not always that what we expect is what happens. Now, of course…

I think it's a credit to the Danish people that that was the center of the campaign and I wish all of us worried more about clean water and animal welfare and less about other stuff.

Yeah, I mean, they are important points. I'm certainly not dismissing them. Anyways, the other thing that kind of caught my attention, and maybe we'll talk about that now, a little bit more, is just again, the unpredictability of markets. Right? I think that just watching what's happened in the last few weeks with a market like gold, where it was so bullish for the first couple of months, or at least the first month of this year, but also to see how precious metals have reacted through the Iranian war. It seems like it's selling off when the war began, but it's almost like it's also selling off now that there might be a deescalation of the war.

I know this changes every day depending on what tweets we're reading, but it's really interesting to see how all of this is playing out and how unpredictable it really is. And even though, as we'll come to in a second, for sure our beloved industry and CTAs overall we haven't profited from it.

We've lost a little bit as an industry, nothing too much, but there will be a lot of dispersion between managers for sure. How did you navigate these last three weeks? There'll be some real winners, some real losers and lots of managers in between. And that's fine in the short-term.

We know that initial positioning is so important and, and you can be lucky, you can be unlucky with how you enter a situation like that. But it just keeps on giving us all of these, so to speak, examples of why or how difficult it must be to be anything other than a rules-based investor. In my biased opinion, of course, when you go through a situation like this.

So I’d love to get your sense of this and where you may think some (I know you focus on fixed income), where you may think that some, how should I say, determining factors have been in the last few weeks?

So, I think the points you make are exactly correct in terms of, first of all, the narrative in the market has been inflation. So, if oil rises in prices, then we're going to see inflation, and then we'll see something. But I think that's not entirely the story.

And the reason why you see that is you see, for example, as you said, gold behaving in a way that you wouldn't expect gold to behave in an inflation base. You wouldn't expect bonds, long-term bonds to be affected as much.

So, if you look at for example the oil curve, the spike in oil is at the front of the curve. Actually, if you look at Brent three years out, it hardly budged because it makes sense. We are going to have supply constraint at the front of the curve, at the back of the curve we're still in a structural situation where we're actually overproducing oil rather than consuming it. We are producing about 105 million barrels a day and consuming about 102. So, long-term was affected a lot less than the crisis.

But in bonds, you see 10-year bonds, in even like the gilts, or the treasuries, I mean, treasuries have moved from US$114, TYs has moved from US$114 all the way to US$110 - an enormous change. That's something that we haven't seen. That's something like more than a year's worth in a couple of weeks.

Similarly, in EM, we have seen in EM, again, not a sell off not at the front of the curve, we've seen, actually, a sell off at the back of the curve which is sort of indicative that what is happening is much more of a risk-off event than just a simple narrative of sort of inflationary impact on the country. And that really makes a big difference depending on which assets you're trading and which specific assets actually you're trading. So, which has more exposure to sort of the Iranian conflict - the Iran/Israel conflict.

And I think the dispersion that you're talking about is actually quite interesting because the dispersion is not going to be just the dispersion between us and different managers, but actually the way that you're looking at the model. So, you know, normally you would do a daily, maybe a daily or a little bit of intraday trading, and the intraday volatility is small enough that it actually doesn't matter what time of the day you are executing, for example. But in our case, if you look at what we've seen this month, oil is traded in a range of say, US$20, US$25 a day. So, at the beginning of this week, we started off with boots-on-the-ground, and Brent was spiking all the way to near US$120. And then, oh no, we've got an agreement with Iran. Oh yeah, we're going to trade all the way to US$90.

So, depending on literally what time of day you're executing your trade will make a huge difference to your performance. So, your model, the ability for you to track your own model this month has been really, really difficult. So, that has been a real issue, actually. And I suspect we'll see dispersion.

I mean, the level of surprise volatility we have seen this month has been staggering. I mean, we run a scenario analysis of what historical data, all the way back to ’98, is actually more surprise volatility that we've seen. And the beginning of March immediately made it to the top chart in terms of effect that we have never seen before that is really, materially impacting volatility in a lot of asset classes. So, that has been a real pain point for us in terms of volatility.

I think one of the interesting things, if you talk about different asset classes, is the bond/equity correlation. I mean, I spoke the fact that in times of inflation we are going to see correlation between bonds and equity being positive rather than negative. And that's really important for an asset allocator.

And I think this month has been a really strong example because we've seen both equities and bonds selling off. So, if you thought one of them was actually acting as a defensive mechanism for you, then think again. I think bonds sold off just as much as equities, if not more, actually.

So, really interesting dynamic this month, and making different CTAs, depending on which asset class you're trading, depending which speed of model you're trading, and which time of the day you're executing, really difficult.

Okay, so let's unpack this a little bit, so, first of all, obviously I'm not entirely sure that I agree that it's difficult to find out when you should trade. You should of course do like all the people in oil that apparently traded 10 minutes before the announcement, Yoav, you should know that for sure. That's like 101. Trade before the news and you'll be definitely profitable. Right?

Yeah, I think we'll get into market manipulation later. But yes, there's been a few instances not just in the oil, but also in equities. On Friday we've seen the S&P selling off, and then selling off again with the boots-on-the-ground, and then beginning to rally ahead of Trump's… Like, it was really quite interesting to behold.

I think you're right. Being rules-based makes the decision to trade much easier. And we're going to talk a little bit about flows. So, I think it's actually really interesting to see.

We'll come to that. We'll definitely cover that as well. I just want to sort of stay with the trend following section and just give people also a little update on performance. But you're making some really interesting points, of course.

Now, just staying with this that you bring up about, you know, correlation and stocks/bonds correlation. Of course, some people will say, well, you know, CTAs, overall, didn't give us any protection either this month. I mean, on an index basis, okay, they're down a little bit less maybe than some of the bonds. But of course, when you have a “surprise” like this, yeah, I don't think CTAs would claim that we will be necessarily doing better than anything else.

So, I think what was interesting is the weakness in the two-year bond auction recently in the US government. So, that's not something that we have seen in a long time. From a personal point of view, I think that a lot of the Western countries are in a fiscal tough spot and inflation, actually, is a good way for them to get rid of that debt.

So, being like on a personal level, I mean believing that yields are actually on the march up, that's something that I think is almost inevitable to me. But timing, when is it going to happen is exactly the point where we cannot even make this determination.

What I can tell you is that the recent move has been, you know, you go into the month and you're thinking you're going to be long bonds. And very quickly you ended the month with actually being short because the moves have been very violent. The reversal has been very violent. But we've seen, in fixed income, actually, the inability for the markets to decide which way it's going.

I think what's interesting is, there are other pain points in the industry as a whole. Private credit is under attack. You've seen cracks. You've seen some funds beginning to gate because redemption. You've seen too many redemptions. So, some of the funds have started to gate. It certainly feels a little bit… the world is feeling a little bit on edge, Let's put it this way, fragile.

Fragile, yeah, that's definitely true. Now, another thing I was thinking of while you were talking, we obviously talk a lot about price, right? So, higher prices, for example, in energy will lead to higher inflation for sure. But I think this time around, again not being an expert in this, but I did notice that one European country now has kind of rationalized oil.

So for me it's not necessarily just price, it's actually supply. Meaning, can we actually get the oil we need. We need oil, maybe luckily for our part of the world, we are heading into warmer weather. You know, this is somewhat different from heading into a winter in terms of not being able to get oil and gas, and so on, and so forth.

And of course, in a much bigger picture, which we’ll leave to the experts to debate, it shows our continued vulnerability to a lot of things, and maybe in particular Europe, you know, we thought we were… Well, we were vulnerable when it came to Russian gas. And now, it turns out we obviously set ourselves up to be very vulnerable to Middle Eastern gas and oil, and so on, and so forth. So, yeah, it's an interesting development, no doubt.

Let me turn to the usual numbers that we talk about. So, my trend barometer, although numbers will show that it's not necessarily showing up in the performance, but my trend barometer is actually still in a relatively strong level at 52. It means that there is some breadth in the portfolio of the markets. It tracks 44 markets in terms of “trending behavior”.

But as you rightly said, the volatility in some of these markets is just incredible. So, even just something as time of execution may mean that you're not really doing great from a performance point of view. But you know, as it stands, that's where we are.

Now, because we're recording a day early, on Wednesday this week, the numbers that I have from the indices are as of Monday evening. But I think yesterday was a little bit of an up day for the CTA industry. But anyways, as of Monday evening, BTOP50 was down 2,25%, still up 6.25% for the year. SocGen CTA index down just shy of 2%, still up 6.25% for the year. And the SocGen Trend Index down 2.6%, still up almost 6% for the year. And the Short-Term Traders Index is, no surprise, doing a little bit better. It's about flat, up 10 basis points for the month, and up just shy 4% for the year.

Now, that is in big contrast to what's going on in the traditional market. So, MSCI World is having a rough time, down 6% so far in March, down 3.23% so far this year. US Aggregate Bond index, as you pointed out, down 2% or so, and change, in in March, now down about 40 basis points for the year. And the S&P 500 Total Return is down 4.59% as of last night, and it's down now just shy of 4% so far this year.

So, interesting developments, interesting dynamics. And even though we haven't made money as an industry, we certainly lost a little bit less, it seems, as an industry, at least for the month of March, and are also doing significantly better, of course, for the year as a whole so far.

Now, before we jump any further, let me just mention to those of you who may not have noticed, but I did recently release the 8th edition of the Ultimate Guide. We've added another hundred books to that guide, so now you have about 600 plus book titles that you might want to go through and find something that will continue your educational journey in this industry of investing.

There are two ways you can receive it. You can either sign up to the Sunday emails, I'm sure there's a link somewhere on the website, or you can go to toptradersunplugged.com/ultimate and you should be able to get a copy, and it is, of course, free of charge.

Let's stay with the markets a little bit further. You mentioned the curve. And you said this thing that, well, if you traded oil three years out, it'll be very different to, you know, trading oil for the next two months. But as a CTA, we don't trade oil three years out, frankly, we trade oil a few months out. So, is there anything you want to add to that or was it just to showcase the fact that the action has been somewhat different depending on where you are in the curve?

So, actually some of us do trade oil…

Yeah, you're probably the exception to the rule, I'm sure.

Indeed. So, I'm not responsible for the commodities. My colleague Tom Babbedge is doing that. I think the idea is, the way he likes to talk about it is weather versus climate. The idea that one of the nice things that when you trade the back of the curve is that it is really about fundamental supply and demand. It's not really about the short-term shocks. And I think this month has been quite instrumental. So, you might think he's trading a lot of oil, but the volatility should be high. But actually, he hasn't seen that much volatility because he's almost away from the news cycle.

So, me, I look at the curve in terms of trying to understand inflation expectations. So, I mean, if you look at oil, oil features, energy features about 6% into the inflation directly as a component into inflation. And then of course there is a secondary effect on food prices and the rest of manufacturing. So there's a significant contribution.

And then you try to say, okay, how does the entire inflation curve sort of rebalances itself as a result of a protracted shortage in supply and demand? I think what is interesting, what you said earlier, is that in the case of commodities, it's not just about money. Like, as you say, for the love of money, you might not be able to get the oil that you need.

And you see countries like, okay, in Europe we see that a little bit because we are still very rich. But in countries like Egypt, they already instituted energy restrictions because they have issues. We rely on the supply chain, like Hormuz, so there is a point where it doesn't matter if you're willing to pay, it's just not going to be there.

So, that's a very interesting dynamic in terms of the supply shocks in the front of the curve, which you probably don't see that much in the back.

Yeah, well, let's stay a little bit with trend following then before we move on to the volatility paper we wanted to talk about. There are two other things, that I noticed in your notes, that you wanted to kind of flush out a little bit. One was flows, the other one was risk management. Let's go there.

So, let me tell you a little bit about my experience. So we've helped. Let me, like broadly, if you trade rates across globally, you would have entered the month probably long bonds, but you would have been short in different areas. So, like, in Japan, you would have been short. In Taiwanese dollar, you would have been short. Then you might be long in South Africa you might be long in other currencies.

Now the shock happens. Now, in normal CTAs, the way we would normally manage risk is say, well, the volatility has spiked and we would sell everything. Now, interestingly, what you'll find, certainly in emerging markets, is that spreads widen. So, if I look at the yield spreads that you were getting quoted its’s like 10 basis points. It's enormous. It became very expensive to trade.

And, in a way, the risk that you're experiencing is a directional risk. So, you may be long some DB1s ones, short some DB1s, and you're trying to reduce the first factor that we are seeing now, which is essentially a bond selloff. And normally we would manage it by selling each one of the assets, right? So we would shrink the entire book.

But that's actually very expensive at the time of crisis. Now, I actually say, I don't have a model to manage this properly. But it's interesting to me that, if I was a discretionary manager and I've seen how risk is managed. Then, what you would do is you would basically sell your receiver, you will close the receiver, you will sell your long bonds rather than just shrink both the longs and the shorts. The first point of action would be to take the very liquid assets, that are very cheap to trade, and just close your duration as much and as quickly as possible.

CTAs don't do that. And that's a very interesting dynamics that has just brought it to the fore, the way that we may be doing something which is lacking. If you think about what, discretionarily, you're doing, it actually makes a lot of sense because what we have is, when there is a spike in correlations, actually a lot of these assets becomes good substitutes for each other.

So hedging with the TY, with the US Treasuries, very quickly, is actually as effective as you would do when just shrinking your entire portfolios, both the longs and the shorts. It's not something that I do actively, but it's something that, when I look at my portfolio and I think, okay, this is not a behavior that I would have expected to do in terms of crisis. And certainly, as a discretionary trader, you wouldn't do that. But as a CTA, these are the rules that we have. We take each of the markets, we just scale it down.

So, what you end up with is scaling both the longer and the shorts at the same time, even though your short bonds might actually be hedging your risk as well. So, that was interesting to me.

Yes. But let me see if I can add a little dimension (I guess is the word I could use here) to that. So, when you talk about that might be an expensive way of doing it, I'm thinking here, ah, yes, but that's because you trade alternative markets. For someone who trades the core markets, where it's all futures and it's all super highly liquid, that's exactly what we can do.

And this is relates back to the paper that Nick and I, and Alan and I have been discussing a couple of times, and which we will be discussing with one of the co-authors next week, actually, when he joins Katy and me, and that is the difference between which market universe you put in your CTA portfolio. And so, I'm glad you bring it up in that way because these are some of the periods where, in a sense, it shows up, but it might show up below the radar. Meaning, when we see the final performance for the month of March, we may not be able to tell, in total, who's trading alternatives, who's trading call markets, etc., etc.

But then once you open the hood and you look down in the engine room, these are some of the things that start showing up. So, it's a good example to talk about, that in times of stress, liquidity is very important.

Absolutely, absolutely, and how to manage the risk is different, as you said. Right? If you're trading US treasuries, you just close the position. There's nothing to do about it. But if you're trading, you know, the cost of like, let's say, closing the book, the long/short book, is like 5% of performance. So, that's actually quite expensive. So, you have to think about how you do that. And I've got to tell you, I don't have the solution for that. It's part and parcel of the way that, you know, the struggles that we have.

But it was very interesting to me, you know, when I'm thinking post mortem, what can I learn or what can I do better about managing the portfolio on the fixed income side? That was one thing which was very, very clear to me in terms of…

But how would you do that from a rules-based point of view, in a sense? How could you do that? I can understand why you can do it from a discretionary point of view, but how would you even be able to program that without adding too much additional risk that we haven't thought about?

So, in a way it's about recognizing what is actually the way volatility comes about. So, in times of crisis you have the first factor which is correlated and its volatility is dominating everything else. And the way that you manage it is you manage it on a front based risk. So, you say, that's okay, I just want a lot less of that front risk. And by the way, this is just good. It's both good for diversification and good for risk management.

And he says, okay, so what you need to do is to have a set up the problem to be able to say I want to ensure that my risk factors are risk managed, and at the same time managing the amount of specific risks to each of the markets that you're doing.

Because one of the things that you probably don't want to do is take too much… Like, you can hedge everything with US treasuries, but then you're taking on quite a bit of risk in that market, right? So there's a specific risk.

So, the idea is can you balance a lot of the constraints which is can you primarily reduce the risk in the first component while the other components may be trading, the volatility on them hasn't increased? And secondly, making sure which instrument is used to hedge it in a way that you're actually underexposed. You don't expose yourself to too much specific risk in a specific country.

It's a mathematical problem. It's an optimization problem. It's not as easy as what we do normally, certainly. And certainly, in the case of liquid CTA, probably not worth the hassle. You would just basically down gear everything.

But, to me, when I'm thinking about what improvement I can do in my system, I'm thinking, oh, if I look at the cost of trading, this month has been very expensive.

And I get that, and I take that on board. The initial thought that I have, when I hear that, is that there's one thing though that you are doing, in doing that, whether you do discretionarily or whether you do it systematically, and that is you're taking on a huge amount of risk if this turns into a liquidity crisis.

Because in a liquidity crisis, yeah, correlations may go to one, but you can't get out. By having kept all the “less liquid”, more expensive stuff due to trade. I guess that would be my initial gut feel, that could turn into a whole different kettle of fish, so to speak.

So, you're right. Liquidity is something that we manage, by the way, as an industry, not just in the alternative markets, but also in the liquid markets in terms of making sure that our footprint is not as big as we think. It's not as big such that we can actually close in a reasonable manner. But when you see very violent moves, certainly, there's certainly a lot more stress in the alternative markets, 100%.

Yeah. Well, the other thing I think you mentioned in your notes was something about flow. So, I don't know if that's something we've already covered or not before we move on to the excess volatility paper, which I also agree with you is extremely relevant for what's going on in March, but was about flows you wanted to bring up.

So, I think it's interesting to see the dynamics of flows this month. And I think, let's describe what we have seen this month and then we can talk about the paper, the academic paper, and then we can maybe try to bring these two things together.

Perfect.

So, what we've seen is, first of all, there's a price gap because, you know, there's an attack, and there's an attack on Iran, and the market just gaps. There's not a lot of trading. It was happening over the weekend. It was over the weekend. There was no trading at all in oil. But when the market opens, the market opens a lot higher.

And then you see, a day later, or actually on the day itself, you start seeing CTA flow. So CTA flows coming in, of course, volatility (as we discussed), volatility is gapping. We will start taking risk off. That comes in.

And I think, as you say, the nice thing about being systematic is that we basically trade. The market does not say it, we just trade. So, that causes, again, an effect. And you're seeing equity being sold off. We are seeing bonds being sold off. And then, as the crisis progressed, you actually see other players beginning to close their positions.

So, suppose I'm a value trader and I believe in bonds, and I'm holding bonds, and I'm holding bonds, and bond keeps ticking down, keeps ticking down, keeps ticking down. At some point, you're just being closed out. There's not much you can do. Right?

So, you're starting to see flows which are driven by the value traders, who were hoping to hold on to their position. And they are just essentially, there's a long squeeze here. They just can't afford anymore to hold the positions.

So, we've seen very interesting flows. On top of it, of course, there's a lot of noise traders. Right? So, we're talking about the Tweets. Okay. We're talking that there's a lot of market volatility. But I think of the dynamic of the liquidity, and who's driving the P&L. Because what you saw, we saw, in the beginning, CTA's selling. You saw real money beginning to come in when things were beginning to stabilize. So, the value traders were actually buying.

Then you saw an escalation with Israel bombing the South Pars gas field, and then, again, another bout of selling. And then you saw a beginning of a squeeze, actually. And you saw that in the gilts market. You saw that even in the treasury markets, to an extent, where people were selling because I think they were just being closed out in the value side. And that's really interesting dynamics in a crisis.

And that's not just in the EM markets. We saw that in the gilts. Although, whether the gilts is an EM market these days is open to debate. But, you know, it's certainly a market which is traditionally very liquid.

But we see very interesting dynamics for flows.

Okay, cool. All right, so the next thing we want to talk about is an excess volatility paper. I'm going to let you completely do this. I opened the paper, I saw all the formulas, and I completely tuned out. I got, you know, stone cold when I saw all that math.

But I do recognize some of the names, one of them being previous guests, Jean-Philippe Bouchaud from CFM. And there are some other people, Adam Majewski, and Jutta Kurth.

Anyways, before I also manage to butcher their names, I'm going to hand it over to you, Yoav, and talk us through this paper, which is called Revisiting the Excess Volatility Puzzle Through the Lens of the... And how do you pronounce that…?

I think it is Chiarella Model. It's an Italian name. So, I think it's Chiarella.

Must be.

But, again, don't blame me. I can hardly speak English, let alone Italian.

So, I think it's a really interesting branch of finance which is understanding flows and understanding how flows affect prices. And it's interesting, for example, that almost all discretionary traders will pay a lot of attention to flows and CTAs in general pay almost next to no attention to flows. And it's quite funny because we feature quite prominently in any such model.

Okay, so I'm going to describe like the generic model, and why this model has come to being, and a little bit about the efficient market hypothesis. And one of the reasons that we have chosen this paper is actually about that it uses the EM algorithm in order to calibrate.

So here's a nice example about the way dynamical systems are being calibrated using this very important algorithm. So, here's the efficient markets hypothesis. It says, well, you can't take money out of the market, basically. So, you know, the efficient market hypothesis is actually much stronger. It basically says…

Shiller came along and says, actually, the reason why you can't take money out of the market is because the market knows everything. All the information about the fundamental value has been incorporated and therefore the price is the right price.

Okay, if you like, in Shiller’s universe, we are all value traders. If we think the price is $US100, and the price is actually US$90, we would go in and we would buy until there is an auction mechanism and everything balances it out with everybody's expectations of where the true value lives. And that's how we trade.

The problem with this worldview is that markets are a lot noisier than what we would get if everybody would be like that. Right? And the reason we did is because we would essentially actually auction, at a certain price, at which all the supply and demand actually balances it out, and then there would be no more trading until we expect the value to change again. So, really there's just not enough… You would move very quickly into the equilibrium point and you would not see the volatility that we actually see in the market.

Okay, now let me tell you the other reason why you can't take money out of the market, which is the completely inefficient market hypothesis, which is, like, I've just invented it. So, the way that like nobody knows anything. Suppose like it's completely noise. Everybody's buying, selling like completely randomly. We are truly like a Martingale. We are truly just a Brownian motion.

Of course you can't make any money because the expectation, given of tomorrow, given today, is zero, is the same. There's no information at all in the market. Okay? So, that actually would be also like, to me, an efficient market hypothesis. It's basically you still can't take money out of the market not because everybody knows the true price, but because there is no true price. It's like complete, complete noise.

By the way, it’s too long - don't read, the outcome of this paper (and actually I can tell you in advance) is that most of the market is noise. So, there is a reason why option pricing and everything works very nicely is because the vast majority of volatility is sort of noise. And this paper said, okay… A few papers said okay, let's abandon this idea. And the way that we will examine price action in the market is rather we will think about it in terms of flows.

We will think about who is buying, who is selling and how the dynamics of those people, who are buying and selling, will sort of play out in practice. And these models traditionally have three players, okay?

The player number one is the Shiller player, the fundamental player. If the price is below your target price, you would buy. If the price is above your target price, you sell. And you kind of tend to converge, right? You can think of it as like it's gravity pulling you to the center.

And then you have the noise traders, which we just described. They just come in and like just introduce the volatility to the market. They don't really know what they're doing.

And the third guys are us, right? The CTAs. And we are the evil people that violate the efficient market hypothesis. We shouldn't exist. We shouldn't be making money because like we trade based on like price went up, we will buy up. It's nothing to do with the value. We don't think there is any... there's an uncertainty. We just, we just follow that process.

And the interesting thing about it is really what drives the flow from each one of the players. So, the value trade, value traders, it's about the position is the one that determines whether you are going to flow, to create a flow when you're going to create a trade.

The Trend followers look at the first derivative, we look at the returns essentially. Okay, so, we do things in the return space.

And the noise traders are just there to excite the system to create some action. Okay.

So, this paper is really nice. And he says, okay, so given that we have these three players, let's introduce some reasonable assumptions about the way that the response functions to position, response functions to trend, and then let's see if we can calibrate the model and, at the same time, sort of introduce being able to handle genuine drift in the underlying value of the market. So, there is a recognition that there is some the value, that the value of things does change.

And they do a really nice… Well, they pick some sort of drift turn over time, which basically centralizes the time series to be oscillating around that value. And then using the EM algorithm, they basically calibrate the model to say this is the proportion of noise traders, this is the proportion of value traders, and so forth. This is a proportion of trend following traders.

And interestingly enough, by the way, CTAs we do not contribute that much to the volatility. So, one of the outcomes of this paper, which is not surprising and, actually, I think very true, and I could have told them straight away, is that we are a very small part of the market.

So, if they look at the contribution of volatility, the first order of magnitude (by a long way) is noise, the second order of magnitude are the value traders, and the third order we just don't feature in the equation at all. And I think it's actually very… The reason why you can tell this is in advance is because there's just not enough auto correlation in the markets.

CTAs would basically create a positive auto correlation in returns, but we just don't see that in the market. And I think it's part of the way that the industry operates. We talked about trying to be small in the market. We try to make sure that we are not actually… the footprint is not that great.

So, that's the paper, really exciting. I really like the mathematics the way that they do them. They use dynamical system in a very, very intelligent way. They use the EM algorithm, the calibration is interesting. There are a few interesting observations there as well. So, what is very interesting is that they say, oh, the market either spends a lot of the time being underpriced or overpriced rather than in the center.

Right.

Okay, now let me be a little bit critical of the paper. But before we do that, I'm just going to say, let's forget about the maths because all these papers, all these models will share a certain feature. And I think the way to think about it is actually to leave aside the mathematics and think about a pendulum swinging.

So as long as there is a real value, you can think about the dynamics in terms of the distance from the center of the pendulum, and you can think about the way that the pendulum swings, and then a lot of the results, actually, in the paper becomes very intuitive.

So, what does it mean? As we go through the center, through the value, the pendulum is swinging fast and the trend followers are the dominant ones giving an impulse. As the pendulum swings upwards, then the value traders begin to dominate. Okay, because the trend followers are kind of running out of steam, and the value traders are saying, well, you're very far away from where true value is, so, I'm going to start pushing you back.

So, you might get a dynamical system, but the way that it looks, the phase diagram would look a little bit like a pendulum swinging. And then it becomes actually quite obvious why you would spend time either being overpriced or underpriced. There are two equilibrium points where trend followers and value traders balance each other, on the top of the pendulum, and on the bottom of the pendulum – so, on the right and on the left, in the pendulum case. A lot of the results that you see are actually what you kind of expect from that sort of a dynamic.

So, I kind of like the paper, but there are a few things which I don't like about the paper. Let's do the ideological one. The first one is an ideological one.

I think that, in some sense, the paper sets out to sort of disprove the efficient market hypothesis. It says, here it is, we can calibrate this, we can make this make sense. And I don't like going all out to disprove something.

I think in reality, and what we see in the market, is a mixture of flows which are driven by flows, and prices which are driven by, genuinely, a change in the underlying valuation.

Like, the dynamics, like this month, what we have seen is the market was fine, everything is happy, and then suddenly the fundamental value of oil has changed. Okay. And we don't really know. And I think that's a dynamic which is really interesting. And we can clearly see that, in the market, where there is a jump, suddenly the value traders will really need to scramble to start doing some trading. And I think it's something that we see in a lot of market events.

If you look at non-farm payroll, for example, ahead of non-farm payrolls you're going to see liquidity actually shrinking, like half an hour before non-farm payroll nobody wants to trade. Market makers don't really want… they don't know what the economic announcement is going to be. They don't want to take on the risk. You'll see, actually, a lot less trade trading.

And then you would see non-farm payroll coming out and the market would gap and the market gaps because instantly the value has changed. And I think that not recognizing that in your model I think is a problem. You're trying to explain all volatility through trading I think is a mistake. In fact, with markets, a lot of the price change can happen without any trading whatsoever.

There's a difference why close price on day one, and open price on day two, are very different, because things have moved. Right? There has been news. The weekend has exposed ourselves to an Iran war.

So, that's dynamics that I think would be very valuable for the paper to recognize and to recognize that's where a lot of the volatility is coming from. And it's not just noise traders. Noise traders are very different to like genuine jumps in the underlying value of the process.

Yeah.

The other issue that I have there is those, again, dynamics in terms of volatility and being taken out, the value traders being called out. Again, this is a dynamic which is really not recognized in the paper. So, I think that you want your models to be simple, as simple as possible as long as they actually describe reality.

And I think the dynamics that we see in the market, in terms of, I think the assumption that value traders have got a response function which is a cubic, which basically goes off to infinity, is completely unrealistic. Value traders get called out just as much as trend followers, in terms of risk management. So, that response function, to me, is completely unrealistic. Inability to understand how volatility and risk management really affects it, again, not something I would really love to see.

To be fair to the paper, the paper deals with very long-time horizons. So, it’s using sort of monthly data all the way back rather than the dynamics that we kind of see in our trading, which is the very short-term sort of daily cycles that we do.

I think for the industry, it would be really useful and it's something which is really important in terms of understanding the way supply and demand affects what we think is the real value of the commodity, say, or the asset, and then the dynamics around that. I think that would be really useful because, at the moment, these models are not useful for providing any prediction or any way for actually for us to improve the model.

Sure. I appreciate that. And since we gave the title of the paper, people, of course, are free to go and visit that themselves.

Anyways, let's talk, before we wrap up today and before you run out of any voice with your cold, let's talk a little bit about your diversification paper or blog post that you did on LinkedIn, I think together with Tom Babbedge.

Yes.

If I'm not mistaken.

Yes. So, Tom Babbedge did all of the work. So, I'm taking a little bit of credit for his work as well. So, I contributed a little bit. But it's a very interesting question.

It's a very interesting question about capacity, which is how to manage capacity, and is capacity, managing capacity is it actually important? So, that's a question that we struggle with in the alternative markets because capacity is very real, and the question is, is it really actually important?

So, the amount of risk that we like to put into each market is limited precisely for the same discussion that we had, that we don't want to trade against ourselves. We don't want to have too much of a footprint and we do want to be able to unwind our portfolio when there is a crisis like we've seen this month. So, the amount of risk that we can put into each market is limited.

We can put more in the treasuries, we can put less in South African bonds. Now, at all, do we need to have those small markets in our portfolio? Do we want to have them at all? And maybe we just should just trade all these big capacity, high-capacity markets.

You know, if you've done index replication, you know that some of our peers might be trading 12 or a dozen or so assets in their portfolio. And, you know, this discussion, you've had it with Rob, Rob believes in trading 300 futures. I'm sure you guys might be trading 50. Okay. And the question is like why? Where's the choice here?

And the observation that we make is that it really depends on the underlying correlation of your market. So, if you have a portfolio where the underlying correlation is high to start with, then, actually, if you have more markets you don't necessarily gain that much of diversity because there's a plateau which you reach very quickly.

So, at that point, the fact that you may be putting a lot of your money in only the very big markets is not that damaging to the overall diversity of the portfolio. I mean the way that we think about CTAs, is the way CTAs make money is the diversification times the average market Sharpe. So, the more diversified the portfolio, in principle, the better quality your portfolio is, the more resilient it is, and so forth.

However, if you start with things which are correlated to start with, then the fact that you're taking a lot more money and you're pushing a lot of the risk into the big market or just start by trading the big markets in the first place, that's not too damaging to your portfolio.

Conversely, if you have a portfolio, you've selected it so that it's very low correlation, which you're trying to select local relation market, now, at this point things are becoming important because a large component of your performance comes from that diversity. And if you start taking a lot of money in, and you start managing it, because you can only push that much risk into your smaller markets, you end up pushing more and more risk into your bigger market and therefore, effectively, you're reducing the number of assets that you're trading. And that is actually a lot more damaging to you.

So, there's a lot of discussion in the industry about whether 200 futures is good, 50 futures is good, or maybe there's 12 futures is good. And I'm saying there's a choice here. If you're looking just for the first factor, the CTA factor, highly correlated market, highly financial, that's great. I mean, in fact, you might think that this means that you are very highly correlated to everything. But that's okay because you're doing index replication. High correlation is actually what you're after.

Conversely, if you're just saying I want a high quality portfolio, then you are handpicking, you want very low correlated market. And then you start having to think about your capacity.

So, we are Just trying to be, trying to sort of frame the debate that is going on in the industry between them.

Sorry to interrupt you here, but the funny part is, and I completely agree with you that saying that, yeah, you theoretically think that the fewer markets you trade should be fine if you want to try to correlate highly with the index you're trying to replicate. But actually, when you look at the numbers, that correlation isn't that great. I mean, the tracking error is substantial. But that's another conversation. I just want to throw it in there for the record.

So, absolutely. I would say that trying to replicate the S&P, I mean, that's the funny thing. If you're trying to replicate the S&P, if you get less than 99% replication, people say that you're not tracking very well. Whereas in the CTA universe, replicating the CTA, you have an 80% correlation. And people think, oh, that's great, you're correlating enough. Right?

So, I think replicating CTAs, where there is some opacity to be in terms of what's actually underneath, it's a much more difficult question.

Yeah. Anything else you want to talk about from your paper?

No, I really shouldn't be talking about myself too much. So, I think I'm good on that one.

Okay. Well, anyways, people should go and read it on your blog, on your LinkedIn, just to get the full flavor of it.

Yoav, I really appreciate you working through your cough and your cold today. Yeah, that means a lot, and you bring up some interesting and important topics every time you come on. So, thank you so much for doing that.

Always fun.

Always fun. Now, I mentioned that next week Katy and I will be joined by an extra person. We'll keep it… Well, it's not really a secret. It is one of the authors of one of the best papers I've read recently, which is the one about how to decide or what is the effect of trading different market universes. So, it's Harry Moore, from AHL. He'll be joining us and I'm sure this will be a really fun, interesting, insightful conversation.

The topics will, of course, be much broader and if anyone has a question that they want to bring up with Katy and Harry next week to make the conversation even broader, they should definitely email me at info@toptradersunplugged.com, the sooner the better, and I'll try and do my best to bring it up.

If you want to show the appreciation to Yoav and all of the co-hosts for preparing and sometimes having to work through a cold, please do so by going to your favorite podcast platform. Leave a rating and review because it really does help more people to discover the podcast and the the content that we produce each and every week.

With that, from Yoav and me, thank you 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 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.