Trend Systems Under Strain ft. Nick Baltas

[00:00:01] Intro

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

[00:00:23] Niels

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

And let me also say, a very warm welcome if today is the first time you’re joining, and if someone who cares about you and your portfolio recommended that you tune into the podcast, I would like to say a very big thank you for sharing this episode with your friends and colleagues. It really does mean a lot to us.

Nick, it is wonderful to be back with you this week. It’s been a little while. It feels like that at least. How have you been? How are you doing?

[00:00:59] Nick

Yeah, I’m glad I’m back. It’s been maybe like a couple of months almost.

[00:01:02] Niels

Yeah.

[00:01:03] Nick

I think the last time we caught up was post this whole liberation situation in April. I’m doing well. I’m doing well. We’re getting closer to, I guess, the end of the first half, which is quite insane, to be honest with you. I was kind of thinking that through. It’s already June. I’m doing well.

I’ve done my fair share of my global trips, being in all regions so far. I am spending a bit more time with the team now, thinking about new products, new ideas and researching, but also looking forward to a break. And it’s going to come in, I guess, five, six weeks or so.

[00:01:34] Niels

Yeah.

[00:01:35] Nick

How about you?

[00:01:36] Niels

Yes, not too bad actually. Like you, not a lot of business travel at the moment, so it’s more relaxing. And, as you say, the summer is here, although it’s been a bit wet in Switzerland. But anyways, we’ll see how the summer turns out.

As usual, there’s actually something that’s always interesting to ask you about when we don’t speak that often, under this what’s been on your radar? But I’m going to let you go first because I’m always interested to hear kind of what’s come across your desk that we’re not going to be talking about but that you find interesting at the moment.

[00:02:11] Nick

I was kind of thinking through what should be the one. And maybe the one that is catching a bit more of my attention is this whole proliferation of those new machine learning techniques in the AI space and how it has been evolving in terms of tools we can use, but also in terms of harvesting and extracting data.

These days, every other day I get an email about a new data vendor that simply came up with a new algorithm that somehow tries to extract sentiment, or extract alpha, or extract something that fundamental data or more conventional data cannot allow us to get access to. So, I’m kind of witnessing that.

I’m not saying that I’m in favor or against, by any measure. So, I don’t want to be misread in this regard, but there’s so much data that, in itself, is giving rise to so much data.

And then being in a business whereby we build systematic strategies, I’m always kind of questioning, what’s the value add, but also, I guess, how perpetual that value can be through time, and what are also the best ways of assessing this type of data and assessing the value that it can bring, or the alpha that they can be used to extract and how that can decay through time, and what are the mechanisms that it can decay through time? So, all this part is one of the things that I guess keeps me up, not at nights, but certainly keep me up in daytime.

And maybe related to that part, there’s a good amount of debate right now (and I’m not even claiming that I’m an AI expert in this regard), there’s a lot of debate as to how plausible and truthful the reaction of those LLM models is to the prompts that we give them, which also can go the other way around and make you kind of doubt every answer you’re getting. And maybe that’s more like a philosophical discussion, or maybe that’s more of a mental aversion that, you know, I start building, maybe for the wrong reason.

But every time now go and kind of ChatGPT, okay, let me ask one thing, or the other thing, or the other thing. I’m like, is that actually true, or should I kind of start doubting about it? And the minute you end up doubting so much, you’re almost kind of becoming averse to the use of it. But I cannot argue that it’s like, it’s extremely powerful.

So, I guess, in managing text, and summarizing, and doing all that stuff that, you know, many people have actually discussed (I think even Rich mentioned that last week), it’s extremely powerful. I remember one day I kind of basically asked on purpose, okay, here’s an academic paper. I kind of summarize it – a pretty good job. And I was like, okay, give me, like, the ten most important references that I should go and read. On the back end of this paper – complete disaster.

Because I know the topic, either it was mixing names or was mixing titles, but obviously not being an expert. And you read it through like, wow, these are the 10 references. Let me try Google them. They’re nowhere to be seen, nowhere to be seen.

So, I think, anyway… I don’t think I’m coming to some… I guess, I don’t know the outcome of those considerations. But kind of summarizing that part, all this proliferation of tools and machines and data has gotten me, maybe because of my business activity to, I guess, doubt a bit more what the outcome, or the alpha, or the value that it is from an investment standpoint. But then using it as an individual, I end up doubting the outcome, but I don’t want to go against the technology that I can actually feel it’s so powerful. So, I’m kind of trying to balance that part. To put it this way. Does it make sense?

[00:05:51] Niels

It makes sense. And I think…

[00:05:53] Nick

So, yeah, it’s on my mind. It’s on my mind because I’m using those tools more and more, and I’m getting those new data more and more, and I’m asking the team to test those data more, and more, and more. So that’s why it’s kind of, I guess, part of my day.

[00:06:04] Niels

It’s clearly on your radar, that’s for sure. I think we can tick that box today. But I think you bring up a few really important points.

Firstly, you bring up this balance that we, as managers and other people in our space, always have to deal with. It could be, for example, the simple one could be, should we be long-term or should we be short-term.

But the other one is, as you say, the data, what kind of input? And of course, these narratives that we’ve heard become more and more changed over the years, for example, with alternative data, and so on, and so forth. So, the question will always be, what should we focus on? And saying ‘no’ to certain innovations is as important as saying ‘yes’ to the right innovations. So, I fully understand the challenge.

And also, by the way, this thing about not necessarily that ChatGPT can give you correct facts, I think a lot of people say that you should be really careful. And it’s interesting that you bring that up because on my radar, this week, was a news story that I think was in the FT or another paper this week, and it basically says that the United Arab Emirates is set to become the first country in the world to provide free ChatGPT plus access to all of its citizens. A move that can fundamentally transform education.

So, you’re kind of hitting on, on a really important point and that is, okay, yeah, it sounds great. You can have your own personal education through ChatGBT plus, but what if it’s giving you the wrong facts? I mean, it’s interesting, right?

[00:07:49] Nick

I think it is. I mean I did see this piece of news and I was like, wow, here’s where we’re going now. I guess governments and governmental entities are deciding upon the tools they provide to the society and I guess they’re happy to pay a subscription for these type of services. It’s almost like social services now that you’re beyond health and security. And that now is like a premium access to ChatGPT.

[00:08:17] Niels

So, what I also saw, or at least I think I saw, I seem to remember this, that is that there was an article about students in, I think it was China, getting ready for their entry exams for universities. And as far as I remember, it said something about it in the article that actually China was shutting down all of these tools for one week while the students were taking their exams so they couldn’t cheat, or let’s not call it cheating, not cheating, but they couldn’t get any help, let’s call it that, a very different way of dealing with things. Right?

[00:08:52] Nick

It’s fascinating. So, my undergrad, I did that between 1999 and 2004. And I remember, at the time, I started doing engineering, I remember, at the time, past the first or the second year, most of the courses would be like an open book, open computer, open laptop, whatever policies. Basically, the exam was like, bring whatever you need as long as it doesn’t breathe.

[00:09:17] Niels

Right?

[00:09:17] Nick

So, bring whatever you need like books, references, papers, you know, your computer, your laptop, whatever. We don’t actually care. These are the questions. You have all the information you have in front of you and accessible to you, just go on and do it. And now you can see, precisely to your point, how completely the opposite route this can potentially go because obviously those tools can replicate or replace cognition in this regard. So yeah, I do remember those days. You had all the information accessible, and still it was hard.

[00:09:48] Niels

Yeah, and still it was hard, of course, but now the difference is you have all the information and someone “who can do it for you” if you were using the AI tools during your exam. So, it levels… It ups the game.

[00:10:06] Nick

It doesn’t breathe, by the way. And it doesn’t breathe.

[00:10:08] Niels

Not yet. I mean you do think, well, maybe China is… maybe it’s a proper approach, saying, well, if you want to show what you can really do to get into this university, you kind of need to do it on your own. Maybe that’s not a bad approach.

[00:10:26] Nick

For sure.

[00:10:26] Niels

So, my last point today, it’s a little bit political, which is something I normally don’t talk about here. But, since I am spending this week in Denmark, I couldn’t help notice that once a year we have this event where all the politicians and kind of CEOs of companies, etc., etc., they get together with the people. It’s called the people’s meeting, annual meeting, or something like that. So, they get together for three days and they discuss things, you know, politics of course in particular.

But the reason I’m mentioning it is that what I do find so refreshing, actually, is to see a democracy like this where politicians, yeah, they can disagree on politics but they’re also friends. And they can sit down and have a coffee and a beer afterwards. And they can intertwine with, you know, normal human beings, so to speak, citizens, and hear their views points, etc., etc.

And you can trust that what’s going on right now in different countries, where we are heading in the complete opposite direction, and at a time where, you know, democracy is certainly changing and is under pressure. So, I just thought, you know, if there was anything that should inspire people, it’s just to see how it’s being done for three days like this. It’s actually, yeah, I think it’s quite optimistic when you see it like that.

[00:12:04] Nick

Interesting, interesting, okay.

[00:12:06] Niels

All right, let’s move to something maybe not quite as optimistic, the trend following performance review.

[00:12:13] Nick

I’ll let you run with that.

[00:12:14] Niels

Well, actually, June is not so bad. So, June is a little bit kinder to our style of investing, for sure. Although I will say trends are pretty scarce still. There are very, very few bright points or bright spots in the portfolio. But those that there are, are coming from slightly unusual places like the livestock sector, which is like lean hogs and live cattle, they’ve actually had some really good trends I’m sure many managers have enjoyed. Equities, of course, here we are. I mean DAX set all new all-time highs and US markets close to all-time highs again.

And this is, of course, a little bit unfortunate in that we had Liberation Day. Because a lot of the long positions got cut around the low. Managers locked in losses and didn’t have a lot of exposure to enjoy the real liberation, so to speak, the upside momentum came back. But I do think this will also open the debate, a little bit, about time frame.

But I generally think that long term trend followers, at least, probably and thankfully didn’t reverse positions and go short on equities. That’s not my impression. Of course, shorter-term managers may have done so. But generally speaking, it was a little bit better environment.

Before I go through the performance numbers, obviously, I’d love to hear your thoughts on where we are from your perspective and the kind of strategies that you follow.

[00:13:54] Nick

So, I mean, look, I would agree with the fact that kind of April was another V shape as the one we saw last year with the dollar/yen unwind. And certainly, a V shape is like two reversions, one after the other, because the first one is when the market reverts and then as it hits kind of some intermediate bottom, then it bounces back again. So, the two reversions, one after the other, so, obviously following trends.

Yeah, that’s exactly the recipe that you’re going to get hit twice, specifically if you’re quite quick in responding. And that in itself kind of crystallizes some losses. So, it is not a surprise that both last year in August, and maybe we can even go back to SVB, but that was a bit more isolated, I would say, in the fixed income space.

But anyway, those three events, you know, one per year, I’m not sure if the regime is different. And I think we had this conversation a couple of months ago. I’m not sure whether the markets are reacting very different to information. I’m not sure if the information in the market is a bit more short-lived and more noisy than what historically was the case. I must say that the response, at least from the interactions I have had from investors, has been yeah, we get it. It’s a V shape.

You know, obviously if you’re allocating into trend following, you should not expect a V shape to give you performance specifically when, over April, every single asset class, in a way, was incorrectly positioned – you know, anywhere from oil, to rates, to the dollar, and obviously equities.

So, there’s, I think, I guess, a not positive but certainly a well understood acknowledgment of what’s going on. Which has led to little activity in terms of experiencing outflows. Actually, none in this regard.

What is more important is to acknowledge the fact that maybe education, maybe understanding, maybe the experience so far has been pretty much in line with expectation and the value that hopefully these types of strategies bring, in a prolonged downturn correction, to your point, I don’t think we have seen a crisis year-to-date or year-on-year that would suggest that there is some crisis alpha that is in need.

So, they have a specific role to play and, because the scenario has not materialized for that role to be played, they still remain part of the portfolio. And also, we get more and more incomings and specifically for these types of profiles.

So that is case number one in terms of client interactions and response to the event. I think the one that certainly, and maybe we’ll discuss it a bit later, that, but certainly can raise some questions and maybe we should not necessarily react emotionally to it, but more about study in more detail, is whether those V shaped scenarios start becoming more frequent. And that frequency should make us somehow rethink as to how we design the strategies or maybe how we manage risk or how we distribute risk. I think we’re going to touch upon that later on. But that’s maybe the second part of my response or my answer, I guess, to your question.

[00:17:09] Niels

In terms of what’s going on right now in the trend space. My own trend barometer finished at 39 last night. So, it’s still a little bit weak. But it’s all, on the other hand, been relatively stable the last few days. We’re recording one day earlier than normal. So, these numbers that I have are actually since Tuesday evening.

And the BTOP 50 index was up 1.17% for June, but still down 3.39% for the year. SocGen CTA index up about 1% in June, down 7.6% for the year. SocGen Trend up 1% as well, down 10.4% for the year. And the Short-Term Traders Index seeing a little bit of a reversal of fortunes, down 64 basis points, and now it’s down 3.32% for the year, which again based on its low volatilities is starting to be meaningful for that index.

MSCI World up 2% in June as of last night, up 7% for the year. The S&P US Aggregated Bond index flat for June, but up 2.31% for the year. And the S&P 500 up 2.2% in June, and up 3.29% so far this year.

So, we’ve got a few key topics we wanted to talk about. But the first one, it’s kind of, you know, a continuation of what we just talked about. And that is the fact that, you know, people are (certainly who listen to this podcast) fully aware that CTAs have gone through some challenging periods the last 12 to 14 months for most managers, I would say. But actually, I was looking at the numbers. Some managers are actually somewhat further away in time from their all-time highs. So, I do see some managers that made their last all-time high as far back as 2022, although for a lot of people it’s only been 12 to 14 months.

I will also say, just to balance the conversation, that it’s also actually the fact for trend replicators or CTA replicating funds, at least some of them are in pretty long-dated drawdowns by now. Maybe not… They’re obviously off their lows, but the drawdowns have been going on for quite a while. So, it’s not just the underlying managers, it’s also people, as it should be, I guess, trying to replicate these strategies.

You kind of mentioned that it ties into a conversation we’ve had some time ago where you were seeing or thinking about sort of static versus dynamic rebalancing. And I have to be careful here because you’re thinking more about sort of time varying volatility targets rather than position sizing.

So, take us into that world and I may bring you back to the old world, but let’s talk a little bit about what you’re observing.

[00:20:16] Nick

Yes, so, and for the avoidance of, I guess, any doubt, to your point, we’re not going to discuss how we think about positioning at the individual markets and whether that should be a static or a dynamic allocation, dynamic scaling between them, for example. I think both you and I agree that there are benefits to this dynamic risk scaling. So, that’s not part of the conversation. Not that we should be dogmatic, but it’s not like the focus that I would want to kind of bring here.

It’s more the fact that, at the portfolio level, building a strategy with some inputs, in terms of signaling some inputs, in terms of risk, call it covariance structure or something along those lines, volatility estimates, correlation estimates, ultimately we built a portfolio that has some realized volatility profile, and that realized volatility profile, more often than not, tries to be within some ranges or at times we even target a certain level of volatility just because this also allows the end institutions to at least scale the specific exposure in this regard. Because if you know that this portfolio is targeting, or realizes close to 10%, you know how to scale it when you put it next to your kind of 60/40 portfolio.

So, I think what I wanted to bring into the discussion is some feedback I’m receiving from my conversations with investors whereby there could be value in changing dynamically, through time, that level of volatility that is targeted. And you and I have this discussion as to whether this is maybe VAR or maybe realized vol of the portfolio.

But broadly speaking, if historically we have been thinking more about a constant risk or a constant portfolio volatility, is there a scope for that to be dynamic? And why would it be dynamic? Because ultimately, reaching a certain level of volatility is a function of leverage, which in itself is a function of diversification or correlation dynamics, which in itself is a function of how directional or mixed the positions are.

To your point, if your barometer is 90%, then, more likely than not, you have a lot of directional exposures. Maybe your long on equities or your short on bonds. So, the headline risk there is pretty much equity risk or interest rate risk.

But at times whereby the signals are quite mixed, and let’s say your barometer is at 30, maybe you’re long half of the equities and short the other half. And in this regard, the portfolio can be more diversified, but this can have implications on the overall correlation structure, on the overall leverage, and therefore the overall portfolio level can be hit in terms of vol. So, that’s where I want to kind of tilt the discussion.

Let me pause here in case you had any remarks and then we can kind of continue this discussion.

[00:23:11] Niels

I do think it’s very interesting and maybe the only thing, again, just to maybe keep things sort of clear, the way I see it (and you and I just touched on that briefly before we hit record) is that you have kind of the school of targeting vol. There are certainly, definitely, a lot of managers doing that.

You have the school of people targeting VAR, but where the vol actually can fluctuate a fair bit, but where you have a risk budget you want to adhere to. And then, of course, you have people who (not going into that direction right now at least) say, okay, we just take an initial risk and we let the vol take care of itself.

And, of course, we’ve seen that in the last year or so with some managers doing that to an extreme where the volatility of the portfolio has gone like from 30 to 90 at times, which is, in my opinion, a bit wild. But, now I’m going to let you continue and then we’ll see where we go.

[00:24:17] Nick

Okay, so maybe it’s a bit of an empirical observation. Maybe it’s a bit of, I guess, of an outcome of my conversations. It feels that over the last year, or two, or three, there has been some attempt to move away from a static volatility target into something that becomes a bit more of a dynamic one. In other words, today I’m targeting 10%, tomorrow I’m going to target 8%, the day after I’m going to target 12%.

Now, why could that be relevant? Because ultimately, the level of volatility, as just mentioned, is purely and I guess monotonically related to leverage. The more leverage I have, the more vol I have, or to hit a higher level of vol, I need to deploy more leverage.

That doesn’t necessarily make the strategy more defensive or less defensive. You still have your trends that you somehow try to deploy.

So, the question then becomes, is an up-tick in the performance or a down-tick in the performance magnified by a higher level of leverage? And is that the right time to have it?

Because you said already, year-to-date, shorter-term, or medium-term, they’re all negative – some a bit more, some a bit less. That basically tells us the following. However quick or slow you are, the turning point is a turning point. A V shape is always a V shape.

Maybe you can benefit less if you’re a medium-term trend follower, but if you’re like quite slow, you’re probably okay because you just get the whole V shape and then you kind of move on with your life. If you’re super quick, maybe get some of the turning point down, and some of the turning point up.

So, there are nuances in how performance plays out. And I think what happened in April is that, you know, if you’re super quick, you lose less in the beginning, but then you don’t recover as much. If you’re too slow, you’re basically going down with the market, but then you recover with it.

So, the end point might be the same, but the trajectory is quite different that gets you there. But all of them are kind of negative. So, said differently, there is no predictability of the turning point.

So, the question then becomes, how levered are you entering into that point? Because if you’re twice as levered, because you’re going to be hit by the turning point, then you’re going to experience like double the loss you would have otherwise had, had you not increase the leverage.

So ,here’s the question, or maybe the consideration, or maybe the, I guess, the debate. Are those events that we have gone through, those V shape events we have gone through the last couple of years, hinting towards more frequent V shapes?

And if those V shapes are more frequent (and maybe I can quote Andrew here, speaking about the signal to noise, which is extremely low in those environments), are those environments inviting us to consider lower levels of leverage, in other words, lower levels of volatility? I think that’s the consideration here.

Now, how can somebody achieve that? I mean, it’s hard, right? Because we can probably say, okay, let’s just reduce exposure when the performance is negative. But then, can you time trend? I think it’s hard, let alone the path dependency it can create.

Because if you start reducing your exposure, then what are you assessing yourself against? The reduced exposure you just have had, or the ideal you should have held but you decided to move away from? So, I don’t think it’s an easy answer.

But I believe, and I feel, and even if I look into the index itself, the SocGen Trend index, it feels that over the last year or two, the levels of realized vol are slightly lower than what historically was the case. You know, they maybe fluctuate between 7, and 10, or 11, and historically you can see values between 10, and 12, or 13.

So again, there’s a lot of averaging there, by the way. So it’s hard to actually make a case. But this is the point that I kind of wanted to bring, whether that level of vol is now as important to be static or that should be somehow dynamic in a way that quantifies or expresses the overall signal to noise that exists out there. That’s basically the whole point.

[00:28:18] Niels

Yeah. I certainly do feel that I remember that there was, for a long period of time, some managers who chose to target volatility, a certain level of volatility, call it 15%, I do remember that. I also think you’re right in saying that there’s probably been a change so that more and more people… because I also have seen it in the naming of programs, they call it dynamic something or whatever. So, I think you’re right in saying that more managers today probably have some kind of dynamic changes, adjustments going on in either their vol level or VAR level.

I think there are a couple of interesting points relating to that. One is why would they do it, right? Why? What’s the motivation for doing it?

So, of course we would say, well, it must be because then they can maybe reduce some of the drawdowns and maybe they can have more risk-on when there are lots of trends, maybe, ideally. So, that’s one thing.

And then of course the other very interesting question, which we may not be able to answer here, and that is how do they do it? What drives the decision to suddenly target a higher vol or a lower vol?

Those are the interesting things from your observations that I would love to hear your thoughts on.

[00:29:47] Nick

So, on the first part, why would somebody do it? I think, well, I don’t just think, I think economic theory suggests, and I think it has been proven by all sorts of social experiments, that investors and human beings are exposed to the so-called disposition effect. Said differently, they get, you can call it like the cumulative prospect theory.

We value less $1 of a gain than the pain we experience if we lose $1. Right? So, I don’t know, you give me a dollar, I become twice as happy. I give you a dollar, I become four times more sad.

So on that basis, this asymmetry of experiencing gains and losses can eventually lead to a hypothesis that says I’d rather be under levered at the time that a drawdown is happening because then I can mitigate the loss even if I lose some participation on the up-side because I happen to have reduced my exposure and then the market kind of moved in that direction.

So yes, I mean, if I place myself in 2022, and I do not deploy as much those trends that were prevailing at the time, not grossing up the portfolio, yes, I would have had a ramp up that would be quite slower to the rest of the crowd. But you know what, in a year that the rest do 20 and we do 15, it’s probably okay. But a year where the rest do minus 15 and we do minus 20, that’s a bit more harsh. So, I think that can be the reason to, purely from an investment standpoint and managing an allocation, the motivation, at least to me.

Now, how that can be done. That is hard because, to my earlier point, I’m sure that any type of past performance adjustment will perhaps show us better performance recently because, guess what, we kind of know how it went. You know, the market went down, and a bit more down, a bit more down, and the drawdown kind of happened over the course like a few days. So dynamically reducing leverage through that period for sure is going to give you a better line.

But you know, if we go back and backtest it, I’m pretty sure we’re going to find situations whereby not being as much allocated into emerging trends would have gotten you into a worse outcome. And it’s now this trade-off that we have to kind of acknowledge.

I don’t think this is a perennial kind of structure and that can allow to outperform no matter what. So, I think there is this nuance here on how this can be done, or maybe there is some aggregate signal to noise that could be utilized.

But the counter I have to it is that, historically, there is no association between the overall trendiness or your barometer to subsequent performance. Except when this is too high, by the way, not too low. Here we’re talking about that being too low.

But the problem, historically, has been when this has been too high, because if it’s too high it means that there are so many trends, but in reality there are very few principal components. And if you have too few principal components that are probably highly correlated, then one going south, it’s enough to take all the rest with it. So, underperformance becomes more likely to happen. So, it’s quite interesting because it’s not even in the data.

The data almost says the opposite. It kind of says look, if everything is actually performing quite well, maybe it’s the time to be a bit more humble to those gains and start crystallizing some of them.

So, I think it’s hard to get that, but I cannot not witness the fact that those V shapes have been happening now more frequently and certainly they raise some eyebrows when there is a trend of follower, I guess, in scope. And I’m not even going in the direction of diversifying those signals and bringing carry or reversion dynamics.

We’ve discussed it at length here in the past and suffice it to say that year-to-date all those other components have actually been extremely, extremely helpful, at least from my perspective, to diversify those trend exposures without reducing the overall correlation profile to the benchmark. I’m purely looking into the trend as an entity and how the volatility target, in itself, is a means to magnify gains, but perhaps magnify losses too.

So, these are the two kind of responses to your question – why and then how.

[00:34:11] Niels

Yeah, and of course we don’t know the inner workings of how people do it. I know how we do it, at Dunn, and how we’ve also introduced kind of dynamic risk budgeting more than 12 years ago now. And we, from our point, obviously find it very valuable. We see that in the data. But I do agree that’s probably something that’s more industry wide.

And of course, you’re going to get critics who would argue that that’s a little bit against sort of trend following history, that you start changing things if it’s not purely driven by signal changes, you’re going to get that. But in order to do it, you do need to introduce other factors, whether it’s volatility, whether it’s correlations, etc. etc.

Actually, Rich and I talked about it a little bit last week. The correlation conscious versus the (I can’t remember the other word of it), certainly people who just look at price as the only…

[00:35:13] Nick

I think we discussed that as well, in the past, the agnostic disparity. That’s pretty much the same concept.

[00:35:20] Niels

Yeah. So, of course, if you can find a way to somehow reduce your downside and your volatility but maintain your returns, of course, that’s the holy grail. But it’s easier said than done.

For what it’s worth, Nick, I do think the last 12 to 14 months, it does reveal something new, some differences between managers that we may not have seen before because a lot of times managers can look quite highly correlated. So, you kind of think, well, all trend follows are the same.

But clearly when you go through and you look at what’s been happening in the last 12 and 14 months, not all trend followers are the same. And so, I do think, as much pain there’s been, there’s also something interesting and good, and certainly for the investors to look at when they do their analysis to see how these models, how these portfolios, different ways of taking risk budgeting, etc. etc., how that’s turned out.

Now, we’ve got a few other things we wanted to touch on, maybe not in quite as much detail as this one, but there were actually a few articles, one in the latest Hedge Nordic publication, about systematic strategies, that I recommend people go and download and read – always good stuff that they publish.

And the first one, it was an article (and we’re not going to get into too much detail), but it was an article written by Christoph Junger, who used to be the, the head of alternative investments at one of the Danish pension funds. I had lunch with Christoph not that long ago. Super nice guy.

And he wrote an article recently, I think, about what he would do different, if anything, if he had to run a portfolio where there were no constraints.

If you could ignore all the rules, no regulations, no liquidity concerns, no career risk (which is a big one), and just focus on the long-term returns, how or could you build a better portfolio? Or maybe you just would build one that’s more risky, so to speak. And so, that was kind of an interesting article.

I know both of us only sort of briefly read through it, but if you were going to say a few words about it and also maybe kind of wearing the hat where you actually deal with a lot of institutional investors facing these constraints, what is your experience? What are your kind of takeaways from these things?

[00:38:08] Nick

The article effectively says that basic asset allocation or policy portfolios would typically have something like 60/40 equities and bonds. One example here is like 60 equities and 30 fixed income and 10 real assets, just to get a bit of an inflation hedge.

And that could, in principle, be very different if you start acknowledging the value-added that we can have from private credit, and CTAs, and infrastructure, and so on, and so forth – so, diversifying components. And the point that the article is trying to make, it’s a bit more like a thought piece. It’s not quantitative in any form. It’s almost like a thought piece that basically says, look, in reality there’s so many constraints in place. People try to be sensitive to costs. It makes sense. They try to be conscious about liquidity, specifically when we have to fulfill certain obligations. And I’m thinking about now pensions and insurance companies. At times there are governance complexities. At times that are kind of horizon management. Technically we care about the long term, if we’re talking about a sovereign wealth fund, but in reality, short-term losses can raise eyebrows.

So how do you manage that or the regulatory constraints? And that has to do perhaps with utilization of leverage, or there are capacity constraints and some of the investments are not liquid enough to allow for high allocations.

What else do they have here? I’m just going through… like benchmark constraints. There are investors that we know, who are benchmarked against some ideal portfolio that is to be held. And obviously, and importantly, there’s a reputation risk as well.

So, the point he’s making here is that look, all those constraints form a very constrained sphere in this hypercube of possibilities that allows a manager or an asset owner to simply move in a very controlled fashion across all those dimensions. And the question he kind of poses is that look, if we had no constraints how we would have allocated?

And it’s no surprise that, obviously, if you are not having those constraints, then probably you’d have much more into private equity and CTAs (and I’m just kind of reading through here), and farmland, or infrastructure, and so on, and so forth. And clearly there would have been historically a benefit to it.

I think the few points that I would make here, and by the way, I completely agree with the fact that some of those constraints are real to the extent that they actually determine substantially how investors think about those allocations. Maybe I can quote one of our conversations years back that we’re kind of discussing how we should size trend following, right?

And you remember at the time I basically said, well, if you just take a mean variance optimizer, and you have a 60/40 portfolio and managed features, you should do 100% managed features and zero 60/40 because managed futures gives you the better return for the same level of vol. So, it maximizes Sharpe. That’s the optimal allocation from a mean variance standpoint.

What mean variance is missing here is that it’s not the absolute risk that matters, it’s the relative. And if you’re assessed the 60/40 portfolio, deviating from it incorporates all this kind of career risk, and this benchmarking risk, and these constraints, that, unless we operate under this, we might end up even losing our jobs.

So, then the whole proposition, that you and I were discussing back then, is that we should think of a managed futures and any alternative allocation under the prism of that there is a benchmark and a tracking error that any active decision will make us assessed against. So, I totally get some of the points.

But I would also make the case that precisely because of those constraints, some empirical patterns exist and otherwise they wouldn’t exist. Like for instance, the fact that low beta stocks outperform is largely driven by the fact that we have capacity or leverage constraints.

If I were to tell you you’re a pension fund and you need to deliver 10% return, then you should be agnostic between using a 10% 10 vol stock or a 1% 1 vol stock that you have to lever up 10 times. But guess what? The latter is impossible to happen because you have leverage constraints. So, from a Sharpe ratio standpoint you should be agnostic. But from an access standpoint you cannot lever up 10 times a 1% return, 1% vol. You’d rather go for a 10% 10 vol stock. And this increases the demand from leveraged investors, leverage constrained investors, reduces overall the expected return, and therefore makes the assets that have lower vol, in this regard, outperform.

And now, how the beta comes into play? Well, if you have a benchmark versus which you need to minimize the tracking error, there is this tendency of going for higher or closer to beta of one assets and that, obviously, in equilibrium, increases the demand, obviously reduces expected returns, and then high beta names end up in equilibrium underperforming.

So, those nuances and empirical observations are purely not the outcome of behavioral biases, but are purely coming from structural impediments, as we call them typically in academia.

So, some of the portfolios or some of the alternatives, that the article puts forward, are there or have historically delivered high return because the constraints are there in the first place. So, there’s a bit of a chicken and egg. That’s the one thing I would kind of flag.

The other one is that we cannot take for granted an expected return, let’s say, even for CTAs, for the sake of argument. Let’s just basically speak about the stuff that we know well.

The fact that the allocation is maybe not at the level that would have been the case had we had less, I don’t know, governance complexity. This has also giving rise to the historical returns of the levels we have seen, and maybe more allocation would have led to capacity issues.

So, it’s kind of almost like a vicious circle. It’s not like let’s replace X for Y, and we expect the Y to behave in the same way as it behaved in the absence of inflows. So, these are the only few things that I would kind of flag. So, it is a zero sum in this regard.

So, the market has to clear, and obviously demand will meet supply, and then that’s where the price is going to be determined. Liquidity is obviously a function of investor activity. So, it’s not coming from an outside world. It’s almost endogenous from the investment process.

So long story short, it’s a nice thought piece because it kind of lays out all the constraints that institutional investors have to operate under. This hypercube of options becomes much more constrained in a sphere of how little can be done.

And maybe by exposing those constraints, the purpose of the article is to say maybe we should rethink a couple of those things. Maybe we should rethink how we reward success. Maybe we should think what is the tolerance for short-term risk for a long-term investor? Maybe we should think how we remunerate individuals. So, I think it tries to bring that angle, or that’s basically what I got out of it.

But pure substitution of the historical allocations to the alternatives. It’s not just one versus the other. I think, to my earlier point, it’s a circle that closes between them and there would be consequences going from one to the other, and not expect that it’s going to perform in the same way as it performs on paper. That’s the whole thing about that.

[00:45:59] Niels

And I also think, I mean, of course, who gets to decide on these constraints, so to speak, what’s the motivation of these constraints? And rarely are some of these constraints put in place for the benefit of investors.

I mean, broadly speaking, what I mean by that is that there are certain things that, you know, for example, lawmakers can make very difficult for investors to get exposure to. But actually, when you look at the underlying investments, they would probably benefit from that kind of diversification.

For example, if we go through, now, a time where we, (and this is pure speculation, of course), but who’s to say that at some point we couldn’t see maybe government dictating more about what kind of assets pension funds should hold, should they have lots of debt to sell, you know, who knows? But it is an interesting conversation to have. And, and I think Christoph did a good job in making people aware of that.

[00:47:01] Nick

Absolutely.

[00:47:02] Niels

Let’s briefly, I would say, turn to another paper, just very briefly, but it is always fun to read one of Cliff Asness’ papers. He writes some good stuff. And this time he put out a piece where he kind of revisits one of his old papers in a quite fun way.

But actually, the reason why I just want to quickly mention it is it’s on a point that you often see being used. And he actually lists eight very big firms using this particular type of “marketing” to keep people invested in their funds, I imagine. And it’s this idea of, well, you better stay invested, because what if you miss the 10 best days of the S&P, your returns would be so much lower.

And just before we started recording here, we obviously were thinking, well, actually, it literally just happened like a month ago or a month and a half ago where the tariff “pause” caused the S&P to have one of the best days for decades. And of course, if you missed that day, that would have hurt.

But I think what Cliff is saying is, well, we have to be careful because people would normally not just miss the best days, so to speak, but maybe also some of the worst days and therefore it’s a little bit disingenuous to target investors, private investors, with this kind of scaremongering about, if you’re not staying invested at all times you might miss out big time.

Anyways, love to hear your thoughts on that. But where I really wanted to just go with this piece (people can read it themselves), it’s just this idea that actually thinking of our own industry, because our returns can be quite lumpy and because we focus on monthly returns, not daily returns, then the question is, of course, could this argument at all be relevant for people investing in CTA’s? Meaning, should they just stay invested or should they try and time it? My personal experience, and when I talk to people who’ve been doing this for a long time, as I have, it’s impossible to time trend following. But anyways, I’d love to hear your thoughts on this.

[00:49:36] Nick

We have always had this discussion as to whether we can time those allocations, right? And I think purely by claiming that these are strategic allocations, we almost implicitly argue that you should not try to time them.

We can make it very philosophical and effectively say that this is supposed to be some decorrelated investment to help policy portfolios in prolonged corrections. Then if it’s indeed some sort of a statistical insurance, there’s clearly no way for anyone to time it and it has to be a strategic decision.

That’s a very philosophical argument that somebody can make. You don’t buy car insurance after you’ve crashed your car. And thankfully there are regulations, on purpose here, to force the allocation onto those, or forcing, I guess, like a car insurance policy because it’s impossible to time it.

And so, in the same way we can argue, very philosophically, that if the value that it brings is diversification, specifically at the times that the more standard (to our previous discussion), more like constrained policy portfolios have, then have it as a strategic allocation. Maybe size it appropriately and then move on.

But also, empirically, there’s, I think, enough evidence that some form of past performance is not a good predictor, but if anything, a weak predictor of future returns. So, can we time it? Can we kind of trend the trend? We know it’s hard also empirically – maybe the point I made earlier on.

However, on the volatility targeting it kind of goes against it. And I think that’s not the challenge. Because if we claim that the volatility target should be dynamic almost we’re saying, I’m moderating my exposure to it. Maybe I’m not going in or out, but I’m moderating the amount of risk I deploy. But that in itself almost goes against my earlier point.

But I don’t necessarily think that the two address the same need at the end of the day because there is a need to strategically allocate into some of those alternative investments. Maybe because of all the historical statistical properties, maybe in the virtue of the article we just discussed.

So there’s a variety of reasons why that that should be the case, but the way to respond or digest short-term losses and maybe moderating those, to my earlier point, that the loss of a certain amount creates more pain than a gain of the same amount, maybe that’s something that could be, at the margin, achieved by this vol scaling that I discussed. But to this day I haven’t had any good predictor of performance coming through, in all fairness.

We can even talk about, I think, who was it, like last time maybe we discussed it. There was some nice analysis done both by Katy and by Man Group, that suggested that most of the outperforming periods of trend following kind of started with like a slowdown or some underperformance because that was the time that trends were suffering, themselves.

[00:52:54] Niels

Yeah, right.

[00:52:55] Nick

So being there and actually claiming, oh, I need to get out now, is like the worst decision for someone to make.

[00:53:05] Niels

The last or second last trading day of November 2021, just after Thanksgiving or around Thanksgiving, we saw a massive one-day loss for CTAs and trend followers. Like something that, you know, even in my career, I thought that that is very rare that we see that. And so, that is the one example where you think, ooh, this is not great to be invested here. I should seek some protection here and get out.

However, it really was the beginning of probably one of the strongest kind of 12-month periods that trend followers and CTAs had, which was of course turned out to be 2022, or at least most of 2022, a very, very strong performance period.

And this is what I mean from just experience and also, even if you took a track record and you looked at it, it’s very hard to predict suddenly why is that month so much better when you’ve had like five months in a row of losing returns? So, I mean, I think it’s certainly in my experience. I think looking at trend as a strategic allocation for the reasons that you say, I do like the car insurance example, just make it mandatory. People should have a 10% allocation to trend.

[00:54:35] Nick

I think that’s why sometimes the way that Cliff is writing is going beyond just investment. I think it’s more the culture of investing. I think that’s what we’re discussing right now, the culture, the approach to it, setting the right expectations and having ways of measuring it. And you know, I think having a trend manager and assessing that on a day-to-day basis is hard to make a case in favor of it.

You know, there was this other Hedge Nordic kind of article speaking about skew and looking into daily skew and I don’t think you can find daily skew, positive daily skew in trend following strategies because they look into longer-term windows.

So, it’s more of an organizational setup that has a certain set of investment rules in place, and the discipline is the one that eventually leads to better economic outcomes not because they’re better just by return, but it’s also better because the expectations, as long as they’re met or as long as they can be allowed to be understood, makes the overall organization more mature in this regard.

To my very, very first point in today’s discussion, the fact that most of the trend investors year-to-date have not actually responded to this V shape is partly an acknowledgment that we know why it’s there. We’re not going to do much about it because we know what we have it for and that’s precisely the market regime that is going to do badly.

How small or big? That’s another question. But the sign of it is very well understood, like there’s no debate.

[00:56:11] Niels

I do want to bring up one final point before we finish our conversation today. It’s actually also inspired from something from the Hedge Nordic Systematic Strategies issue that came out earlier this month.

And it’s actually something that Kamran writes, in his editorial, where he kind of raises the question about have the markets changed, you know, have the world stopped following rules, so to speak, and maybe our systems need to be adapted somehow. Which, of course, is a conversation that we often get when things are a little bit tough.

Now of course, if you just look at the world right now and not think of it as a trend follower, you would probably say, yeah, this looks a little bit different. The things that are taking place right now, the way, you know, markets are moving just based on tweets or statements, whatever, it feels different. I’m sure that if one went back many decades, there would be other things where you could have said the same. But I think does raise at least a question that will be debated.

And, you know, again, if we think of, as trend followers, are the traditional ways of doing trend following, are they breaking down? Are they just on pause? Can we be better at separating the noise from the actual signals, and so on, and so forth? And I think he even mentions (but this is more from memory), do you need to add some kind of discretionary input?

Of course, I think you know where I stand on this, but I’d love to hear your, maybe as a sort of final point, this idea that when we go through these periods and people will challenge maybe some of the assumptions we have as, as trend followers, is there anything here that you find interesting in that debate or is it just the usual, this is kind of how it works. You go through these periods before, you know, it starts working again?

[00:58:35] Nick

It resonates quite well with me, and it goes quite in line with what we have discussed last time and this time, which is all this reflection of how I perceive information being produced, consumed, and how investors react to it, but also how the broader market responds to it.

I think this whole week that we have to go through in April, it’s not that we’ve seen it in the past, and at the time, being a systematic investor and claiming that immediacy to information is actually important, I don’t think it would have fared that well in the sense that what was the signals that could be consumed systematically that will get you better off beyond just luck? To your earlier point, how would we know that S&P would have one of the three largest returns in like 50 years on that day? I mean, extremely impossible to foresee, but also extremely impossible to argue that this is genuine in this formation of systematic signals.

So, generally, systematic investors are trying to be on top of information, and on top of price and volatility innovations, and try to be fast in a variety of ways – not necessarily fast in rebalancing, but fast in digesting this information in their systematic signals.

But at the time that those signals are pretty much noisy, then it is a genuine question if that should be more discretionary. I don’t think that is necessarily the answer, but I cannot avoid the temptation of acknowledging that this is a question to be asked more broadly as to how we think systematic investing should respond to those types of events.

And that goes back to all the points we made about the signal to noise, about scaling the overall volatility, about maybe at times being slower in the way that we react.

I think we discussed it last time that this year, simply looking into macro dynamics and looking into a falling growth and perhaps a relatively range bound trajectory for yields and inflation, probably you’d be out of equities and maybe you’d be more into gold, or maybe you’d look into some sort of a steepener.

So, there are investment themes out there that actually played out quite well, but have been, even in a systematic way, less responsive to this fast pace of noise generation and less of single generation.

So, in summary, I fully relate with some of the points made in the editorial, and I think it quite nicely reflects my own views over the last few times that we have discussed. I think it’s pretty much in line with those. I don’t think we learn necessarily anything yet. I don’t think we should emotionally react.

[01:01:27] Niels

Agree.

[01:01:27] Nick

But I do not think we should be dogmatic and just like okay, life, like whatever, noise, just leave it. Let’s move on with our lives.

[01:01:33] Niels

Yeah.

[01:01:33] Nick

Because the market is evolving.

[01:01:35] Niels

Yeah. And I think as we often say, I mean even if you’ve been doing this for decades, we’re still learning. And it is a journey of constant, small but constant innovation and improvement, that’s for sure.

Nick, this was great. I appreciate all your insights and your thoughts and the time spent for preparing for this conversation.

And of course, to the audience, I can only suggest and encourage you to go and show your appreciation for Nick’s work here by going to your favorite podcast platform. Leave a rating and review for this episode. We so appreciate that.

Now next week I will be joined by Yoav Git, he’s back. So, that is also going to be, I’m sure, a super interesting conversation and see what he has been thinking about and also writing about as he does on LinkedIn. So, if you have any questions for Yoav you can send them to info@toptradersunplugged.com. I’ll do my best to bring them up.

From Nick and me, thanks ever so much for listening. We look forward to being back with you next week and until next time, as usual, take care of yourself and take care of each other.

[01:02:44] Ending

Thanks for listening to the systematic Investor Podcast series. If you enjoy this series, go on over to iTunes and leave an honest rating and review. And be sure to listen to all the other episodes from Top Traders Unplugged.

If you have questions about systematic investing, send us an email with the word question in the subject line to info@toptradersunplugged.com and we’ll try to get it on the show.

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