As trend following begins to reassert itself, Niels and Nick Baltas dig beneath the surface of recent CTA performance - where the signals are working, why fixed income remains unresolved, and how speed is revealing deeper structural divides. But this episode goes beyond attribution. What if the industry has mistaken correlation shifts for changes in signal speed? What if the very idea of a “trend beta” is flawed? And what if compounding durable returns isn’t about chasing performance, but protecting against path? This is a conversation about design over style, clarity over convention, and the quiet decisions that shape real outcomes.

-----

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 Nick on Twitter.

Episode TimeStamps:

00:50 - Catching up

05:02 - Industry performance update

13:18 - How the speed of trend following strategies is changing

24:13 - Is there a way to define speed differently?

32:19 - Do we have the right definition of a trend index?

38:20 - Brainstorming ideas on how to improve trend following benchmarks

43:50 - The best car needs the best brakes

49:26 - Baltas framework for enhancing trend following performance

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 Nick Baltas and I, Niels Kaastrup-Larsen, where we each week take the pulse of the global market through the lens of a rules-based investor.

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

Nick, it's wonderful to be back with you this week. It's been a little while. How are you doing? How was your summer?

It's been a while. Good to be back, Niels, good to see you. The summer has been good. You know, you cannot take a Greek out of Greece in the summer. I think I have a 43 out of 43 of my summers, 100% hydration. Spent a couple of weeks there with family and friends. You know, got back over the past weekend.

You know, we had a bank holiday here in the UK. So, it’s not very slowly picking up things, actually, I would say quite aggressively so. It's been a busy week so far. But yeah, summer was good. Summer was good. How about you? Tell me.

Yeah, I was, I was up north, just got back a few days ago. So, our summer was not particularly hot, so to speak. But I do gather that in the south of Europe, certainly when you see the weather forecast on the news, sort of southern Europe, Spain, Portugal, but I imagine Greece as well have been, you know, once again hit by some pretty nice but very hot weather.

Yeah, there were some 40 plus days but thankfully on those days we happened to be on an island. When I was in Athens, it was actually around the small 30s. It was bearable. I think my genes can survive that. But I can certainly gather that some of the more northern European tourists might feel that the 30 degrees is a bit of suffering.

Yeah, it seems like we see more people with number plates from really southern Europe coming to Scandinavia now each summer. So, I think some people are really escaping the heat nowadays.

I'm not surprised. You know, when I was in my teens, we did not have that many days that the temperatures would go up. I think part of the climate change is kind of impacting us in the same way that the UK summer has had a number of 25 plus, 30 plus days this summer. We did not have that, not 10 years ago. So, I think we do see the effect.

Yeah, I did read an article about the suffering people are feeling when they take the tube nowadays during the summer in London. I remember my own 12 years in London on the tube and back then it was probably cooler, but it still felt pretty hot in the summer, I have to say.

You need the air con. Now, you need the air con.

You know, I don't know if you have anything on your radar other than something else from your holiday, which, feel free to share. But in the meantime, speaking of hot, what caught my attention, of course, this week, was that the whole debate about Greenland is heating up again.

And of course, as a Dane, that does catch my attention. And apparently now three people have been kind of identified as trying to specifically influence the relationship between Denmark and Greenland. People related to the US administration, put it that way. So, that was a little bit interesting.

And also tied to the US administration (I couldn't help noticing) is the fact that there now is a direct attack on another Fed Governor, Lisa Cook, trying to essentially change the composition of the board of governors. And there's only seven of them.

And I think if they get Ms. Cook or Mrs. Cook, off of the board and appoint another one, I think they're pretty close to having, “five of the seven”. So, interesting times ahead based on those two observations, I think.

Now, let's go to something that's been a little bit cool in the early part of the year, heating up a little bit over the summer, and that is of course trend following. Because July felt, a little bit to me, like kind of a turning point for CTAs and trend followers. And there were signs of a few constructed trends and we only have, as we are recording, today and tomorrow left to trade for the month of August. So, you know, it feels like August is a bit of a continuation with some better momentum environment for our strategies now. In fairness, probably not many managers are knocking it out of the park, maybe one or two, but still performance looks pretty solid when I look at the kind of early, early numbers.

And of course, let's cross our fingers that the next 24 hours will go without any big announcements from Washington to spoil the party, as we saw at the end of July, of course, with the copper announcement. We do see, or at least from my vantage point, continued strength in the equity part of the CTA portfolio. It's being supported by a few select commodities, I would call it, in particular, of course, the continued rise in prices of things like live cattle has been important, gold, silver as well.

And we do continue to also see some weakness in parts of the grains, in particular wheat, with a bumper crop being expected in many parts of the world. So, that's obviously weighing down prices, but it's actually pretty good for trend followers as far as I can tell.

Now the tricky sector has been something like fixed income. And I get the feeling that CTAs are not aligned here. I get a feeling that some people are short, some people are long. When I look at the daily fluctuation sometimes, and I look at what happened in fixed income, it seems like there are basically two camps at the moment. So, it'll be interesting to see how it all turns out.

And of course, fixed income, in general, has been pretty horrible for trend followers because they've been stuck in large trading ranges for quite a few months. And similarly, maybe you could talk about energies in the same way. I'm not sure about positioning here, but certainly from a trading perspective, oil has kind of just been gyrating up and down for quite a while without too much real follow through either way.

So, those are kind of my observations about August. I'd love to hear your experience in terms of either relative contribution or changes in exposure that you've seen in your portfolios.

I mean, I think broadly I would be relatively in line with some of your comments. So yes, August has been a positive month, broadly speaking. And that is across a variety of programs that we have in terms of speed, in terms of equity participation, and so on, and so forth.

Broadly speaking, the slower the better for the month, which seems to be the theme of the last couple of years. Perhaps also related to the discussion we have had back in the days on the V shapes that we saw in April, the V shapes we saw last August. So, month-to-date the slower has been better but not substantially so to substantially change the overall profile.

So, we're talking about between 2% and 3% of positive returns, you know, for a volume of 10 to 15 depending, again, on the variation. Return drivers, to your point, are indeed quite similar and partly reflecting also what we've seen year to date. I think equities, obviously, with a significant April impact, have had significant negative contribution at the time but have been now contributing, capturing those positive trends, as you nicely pointed out, gold for sure.

I think gold is probably the number one positive contributor year-to-date out of a universe of like 80 plus markets. And I would agree with you on the fact that the interest rate space is by far the most challenging. You know, last week Yoav was speaking and he actually made the point, and I certainly agree with him. This is a very disappointing year for fixed income.

The reality is there's no particular trend. It's really kind of going around and whip sawing with no specific direction. If I try and see the difference between different speeds, month-to-date, on the fixed income front, there's a bit of a discrepancy. I think the slower has done actually slightly worse than some of the faster ones. But you know, the absolute numbers are actually quite small. The contribution to the month performance from the fixed income components is very, very muted in a way.

Yeah, yeah, I completely agree.

So, not much to say about it if I just literally look into the numbers. Yeah, we're talking about like a month-to-date contribution of negative 20 basis points. It's quite small. The primary drivers have been equities and currencies, actually.

Okay, interesting.

For this month. I mean, some other reflections. specifically in the CDA space. for the month of August, has been very strong performance of carry components. Which again, is sitting outside of trend following, but carry continues to have very, very strong month, or months, or maybe a year. It's supposed to be one of the best years for carry across the classes. But this is beta neutralized. That's beta neutralized.

That explains, actually, a little bit of some very early numbers that I've seen where I thought that's a little bit of an outlier, in a positive way. And that, I think, would explain that.

Correct.

Yeah.

Yeah, so, those CTAs that deploy some of carry or perhaps stop loss or reversion exposures, I would expect them to fare better month-to-date as well as year-to-date. And I think some of the reasons could come from those components, specifically the carry one.

Okay, cool. So, my own trend barometer actually yesterday finished at 34. So that's still a weak reading from a trend perspective but also from a relatively short lookback period, which is what the trend barometer uses. It hasn't been a great month.

And I think that's kind of confirmed when I look at performance numbers because the BTOP50 index, which obviously is a very broad index, is indeed up 0.92% as of Tuesday, down 3.29% so far this year. So, you know, not a bad year really. SocGen CTA index up 1.3% but still down 6.25%. That's fine. The Trend index definitely stronger, up almost 3% so far, down 7.39% so far this year. But then comes the little bit of the outlier and that's the Short-Term Traders index. It's down 1.25%, and it's now down 6.29%. and that's a very low vol index. So, if we've turned that into kind of CTA or trend index, vol, It's actually suffering and by far the worst of these indices. So, it kind of speaks to this idea of how different speeds are behaving quite differently this year.

I think it also aligns to my earlier comment that, month-to-date, the faster you've been, the relative underperformance you've had, you know, still positive, but you know, to your point, faster signals this month have not actually delivered as much as slower ones. So, that's quite in line with my, with my data.

Sure, okay, cool. So, in the traditional world, MSCI World is up 3% as of last night, for the month, and up 14.3% so far this year. And the US Aggregate Bond Index up about 1% in August, up almost 5% so far this year. And the S&P 500 Total Return up 2.35% and up 11.14% so far this year.

Okay, so let's leave the usual stuff behind and let's talk about two topics initially. Let's see how much we get to today. But two topics, initially, that you brought along, which I find very interesting because on the surface they look like, are we going to talk about things that we've talked about before? Or is this where there’s a new part?

And actually, when I saw your thoughts, I thought this is interesting. This is kind of a new twist to an old narrative. And the narrative here is really that CTAs have gotten slower over the past decade or so, maybe even more. But you question whether that's really a shift in speed or whether that's a shift in structure. And I think this is interesting because I think most people will just, by default say, oh, that's a that's a difference in speed.

So, I'd love to hear you lay this out and hopefully we can get into some level of detail about this because I think investors need to understand this maybe a little bit better than the way we often just sort of casually talk about speed.

Yeah, thank you, Niels. So, I mean this topic came to me, maybe, partly as a consequence of some of your discussions with the rest of the other speakers in the show. And you know, we keep on talking about the characteristics of trend following strategies; how fast the signals have been or have been shifting towards, or maybe they become slower, and how the industry is adapting, and so on, and so forth. And most of the empirical analysis to document those patterns pretty much comes from, let's build a set of bottom-up strategies that go from a very fast signal to a very low signal. And let's start doing some ex-post correlation analysis, maybe on a one-year rolling basis or a two-year rolling basis and see what has been the shift in those correlations throughout the years. These are the correlations between those different speed programs in, let's say, the SocGen trend index.

And you know, we keep on making those comments that you know, some of the programs have become slower, and so on and so forth, that pretty much comes from this type of analysis. It's an ex-post analysis with those kind of stylized simulated returns and how they correlate and how they cross sectionally rank on the base of that correlation to the benchmark.

The point that I wanted to flag or maybe just bring forward without necessarily having specific proof about it, is that purely by looking into a correlation metric, beyond the fact that we can make the case that something is moving more than what it did in the past, I don't think we can necessarily argue that the speed of that program changed versus, for example, that program having had other features being introduced.

Let me make a very basic example. If I'm a momentum investor in equities and suddenly I become faster or I introduce a value signal in addition to my momentum signal, it is highly likely that my correlation to some benchmark will shift very similarly. But the reason for the shifting is very different in those two cases. In one case I change the speed at which I'm following trends. In the other one I'm just interacting with those speeds with a negatively correlated signal. And that's now a value signal.

So, it might appear as if I'm behaving slower or faster, but this does not come as a consequence of a change in speed. It can simply come as a consequence of me introducing stop-loss rules, introducing carry signals, introducing reversion signals, amending my program, amending the algorithm not at the speed of trend but on other components around it.

Maybe we're adding different markets, and the different markets behave very differently with a given speed, which, in itself, changes those correlations. So, the fact that the SocGen Trend index, in itself, is a living creature of realized returns of real programs, and it's not a simulated history of today's algorithms, that in itself can put a question mark on this type of analysis.

So, to be clear, and maybe I make a pause here to hear your thoughts, a shift in realized correlations is a necessary but not sufficient condition to infer that the speed has changed. We can maybe hypothesize it because the data suggests that this is the case from a correlation standpoint, but it could be other reasons that interact with strength, in a variety of ways, that make it look as if it's faster or make it look as if t's slower. So, that's the point that I kind of wanted to bring and discuss with you.

I think it's a super important point actually. And it also is something that relates very much to my own day-to-day work. Often at Dunn we're classified as a long-term trend follower, which is true, but only if you look at sort of signal generation. If you look at other things like risk management, as you say, you can have risk management filters that react much faster and it'll look like you have short-term models in there, but you don't. You just have more agile risk management, for example. So, I think this is very relevant, but it's also very difficult for investors to maybe fully get their arms around.

I also agree, I don't have any experience with this really, but I also agree that there is definitely and there has been an introduction (we know that) of other strategies in the trend programs; value carry, other things. I don't know if, from your experience, whether you have any feel for whether, say, a carry or a value model actually makes a strategy look slower, to your point. I mean, have we become slower or is it just that these extra factors that are playing there?

And then the other thing that I was thinking of… And the topic actually, or the question actually just escaped me. So, maybe I'll leave it at that. I'll try and think of what I wanted to ask you as well. But those are the things that I see as well.

I agree with you, also, on the risk management component. Even changing the way that volatility is estimated or the covariance estimated, or how you penalize some of your exposures might, in itself, appear, ex-post, as a shift in the speed. Speaking from personal experience, we do a lot of research on a continuous basis and I'm sure CTAs do the same, if not more, as they’re day-to-day as well. I cannot speak on their behalf, but I cannot, at the same time, feel that people go to bed and wake up and say, let me make it faster, let me make it slower, let me make it much faster or much slower. This is the anticipation that I'm having about the markets and the macro environment. Let me do this and do that.

I think the response is more on the risk management side. Maybe let's reduce the level of risk we take. Maybe let's reduce the net or the gross exposure. Maybe let's have a more prudent way of diversifying our signals and less so let's become just faster for the sake of being faster or being slower for the sake of being slower.

So, I think some of that discussion as to what is the speed at which we can maximize the correlation might be masking other types of phenomena that take place in the model evolution and model design that we just shrink down to a single number and that's the speed. I could be completely wrong, just to be clear. That's my disclaimer here.

Let's use another extreme example. If I take a correlation of a momentum trader to a value investor, I can equally say that it's a contarian, but in reality it's a value investor. So, it might appear as if they take negative return signals, but in reality they might take fundamental information to position themselves in favor of convergence back to fundamental value. So, the correlation beyond just the statistical nature that it has cannot tell us, necessarily, what is the behavior of a particular program.

So, I think we should look at it with a pinch of salt while acknowledging the deficiencies it can have. I can produce information, I can produce variations of strategy that actually show that the SocGen trend index has become faster, not slower. And that, again, depends on the markets that I can put together, depends on how I calculate the signal, depends on a variety of ways.

If we use an exponentially weighted moving average with the same look back window, the story might be slightly different. So, I remember this Quantica paper, it was like a couple of years back, maybe two years back, three years back, that they did this maximization of correlation per year and identifying what's the half-life that does so. So, what is the half-life of a signal that achieves the maximum correlation to the SocGen Trend index?

And I think they ended up with something like 66 business days of half-life for the last 15 years. If I were to claim that this is the maximum correlation, and then we argue that we have become slower or the industry has become slower, I don't think that 66 business days over half-life is what we consider as being slow.

So, again, I'm just putting a bit of not doubt in a bad way, but more like a transparency and scrutiny on how we feel about the speed and how active we feel the industry is. When in reality I think the research direction moves more towards how can I protect my portfolio, how can I anticipate the signal to noise dropping, how can I anticipate a macro environment that becomes less trendy or more trendy, and less of should I make like a three month speed, a four month speed or a five month speed? But I could be wrong. I could be wrong.

Okay, so, a couple of things. First of all, I'd love to know whether you think we should rethink how we define speed in a trend model. I mean, it seems like what we're using is not necessarily a very good indicator. And I'm not sure that there is really any universal definition of speed anyways, frankly.

The other thing is that I think a lot of modern portfolios, CTA portfolios today, you know, the speed is not constant anyways. I mean we do dynamic select parameters on an ongoing basis and therefore they will shift, over time, quite naturally.

But again, if we don't really know how to define it, what should investors do? Because they love to put managers in buckets. Oh, he's a short-term manager, oh, he's a medium-term manager, or whatever. But then they may be surprised by the way they actually behave during a certain period of time. So, is there a way to define speed differently?

The signal in itself is a version of how trendy a market is. And if you believe in dynamic risk scaling it can also be an input into your relative risk scaling between the assets. And that's about it. Whether it's slightly higher or slightly lower, we can have the conversation. So, I don't think there's a unified way of defining speed, but condition upon a single definition. We can have a family.

As I mentioned, I think a lot of investors like to classify managers in certain categories, which is fair, and speed is one. But it reminded me about the conversation I had, which you may have listened to or not, with Rich a couple of weeks ago, where we talked about, well, you know, trend followers are often labeled the same because long-term their correlations are pretty high, but really they're not the same.

And you could probably define four or (including my extra definition) five types or categories of trend followers; you know, the replicators, the core diversifiers, the crisis risk off-setters, the outlier hunters. And then I added one called the pure trend followers because I do think there's a little bit of a difference there.

And so firstly, I don't know if you listened to it, if you did, I'd love to hear your thoughts.

I haven’t, that one.

But it does beg the question, and that is how can we help investors maybe look at the universe of trend followers in a different way in order to make sure that they actually get the right trend follower or CTA for their purposes. And of course I'd love to hear what you think your own strategies, that you design, what kind of one kind of trend follower are you, Nick?

But, I do think it's important. I actually think it's a conversation that we need to evolve a bit more if we want a broader participation of investors in the space so that there are fewer of these ‘surprises’. Oh, it didn't do exactly what I expected. Well, maybe because you were looking at it the wrong way, for example.

Yeah. So, my immediate response, or maybe my impulse response to your question is that trend followers, or at least the strategy in a broad sense, achieves two objectives: some form of defensiveness in prolonged equity drawdowns or market macro driven drawdowns, and some long-term positive returns. And this is achieved with some level of decorrelation to benchmark markets, not locally, but more longer-term. Which, to a certain extent, explains why longer-term CTA managers are more correlated than locally. Locally the differences become more prevalent, but long-term, whether you are 10 out of the 12 months long equities or 9 out of the 12 months long equations because you have had an equity rally, at the end of the day is going to realize a positive correlation long-term, but not around those transition points.

So, I think what really matters is what trend following is used for. If that’s used for a defensive mandate, it's a very different discussion to be had. Are they used for an absolute return vehicle as an overlay together with other absolute return strategies, maybe hedge funds, maybe private assets and so on and so forth? That's a different discussion as well.

And sometimes you ask the question and then the response you get is probably different to what you're expecting it to be. Because sometimes you get this question kind of posed to investors and say, okay, what do you need it that for? And they might as well say, you know what, I actually care about that being my inflation hedge, I care about that being my prolonged drawdown kind of defensive sleeve.

And if that is the objective that should prevail, then obviously we do not want to have a drawdown. And it's painful to have to go through the drawdown and explain what's going on. But it's a different objective.

I think to me, a response to your question would go along the lines of how do we set the expectations with investors, how do we monitor those expectations, and how we eventually manage them? Because on their side they have their own stakeholders, internal and external, to convey those points, to convey those discussion points.

And if it's a single line item, it is completely assessed in isolation and it has to be seen alongside the rest of the portfolios. So, that's how I typically go about when it comes to those strategies.

That's why in the very beginning of our discussion, I mentioned about carry, I mentioned about reversion dynamics, and I mentioned about beta neutral implementations. Why beta neutral? Because if beta is what you're getting from trend following, and seemingly it's a good beta timer, or maybe not (we can again debate upon that), you don't want to add more beta to it, whether it's adding to it or maybe negating it. You want to add something diversifying to it. That, in principle, it should help at the periods that are not performing like this year.

But speaking of beta, actually, it's kind of a nice segue into our next topic that we wanted to talk about. We often talk about the trend beta as if it's something we can just download from an index like the SocGen Trend index. But let's be clear, that's an index. You know, that index isn't some pure representation of trend, it's a snapshot of the 10 biggest managers.

That means that it's shaped by AUM, by survivorship, and by whatever model compromises came with that scale of the program. So, the real question, you know, are we measuring trend following or just what it takes to run a billion dollar plus CTA today?

And again, I know you have some thoughts about this, so I'd love to dive in. It's obviously something that, when it comes to indexing and replicating CTA returns, it comes up in our wonderful conversations with Andrew and Tom. So, I'd love to hear your thoughts on this.

Yes, and in reality that's where this topic came from, by the way, those discussions. And, okay, maybe I share my thoughts here in as much of an objective manner as possible, but everything is subjective, clearly.

I think the way that human beings behave and investors behave is in the form of dimensionality reduction. The fact that things like principal component analysis resonates quite well is because it allows us to reduce information to a few tangible items.

So, thinking of a benchmark is both allowing us to condense information into one or two items, but also, in itself, acts as a benchmark to bid, maybe to track, maybe to outperform. And the point I was trying to make here is that what is really the trend beta? And I've heard, obviously, your discussions with Andrew and Tom speaking about the trend beta and replicating it and the trend alpha. I've had my personal discussions with investors that, you know, speak about what is the trend beta. And while I do get the sense of what is the need here, and that is a representation of what a trend following strategy would look to be defined somehow, I would suppose the question, is the SocGen Trend index a trend beta in itself?

Clearly, I don't really have visibility on how this is put together beyond the fact that we have, you know, the average return of the largest end managers, on an annual basis, and so on, and so forth. But that in itself entails some survivorship. So, it's only the managers that stayed alive that can be used in that average. It is the manager either outperformed, so their assets grew by outperformance, or they managed to attract capital, so they grew by simply attracting capital. And that aggregation of those 10 managers is now perceived as the industry. But is that the industry or is that perhaps the best in managers? And the threshold to clear is actually quite higher than just a beta. So that's point number one.

Point number two, if we think that this is a core trend exposure, we're implicitly stating that all the design choices, beyond just a simple trend (which in itself is very hard to define), but anyhow, let's assume that everyone has a transcript that says that's your trend thing, do your risk management however you want, but this you keep as is.

All those design choices around risk management, maybe reversion signal, maybe stop losses, we make a supposition here that they average to zero. If they do not average to zero, the average of those 10 managers is now trend beta plus a bit of alpha, or maybe a bit of negative alpha. I don't know if they add value or they deduct value.

So, where am I going with this one? Again, it's almost like a philosophical point that I'm kind of trying to bring up. But at the end of the day a trend benchmark is different to a back tested performance of today's models. Statement number one.

Statement number two, there is no absolute scrutiny on whether those 10 or 20 or however many managers are just core trend managers. That in itself is very hard to define. And maybe here, maybe Tom has some good pointers to share, next time around when he's, when he's around.

Because if they are not just pure trend followers, you know, call themselves trend followers at heart, then I would question what is the impact of all the risk management features, the correlation adjustments, the volatility targeting, the gross and net leverage adjustments, whatever the features that come into play, how do they aggregate into an average? So that's basically a stream of thoughts that I have in the trend beta space because I also get the question, can you give me the trend beta?

How do you define the trend beta to start with? Maybe it's the index, maybe it is not the index, but if it is the index, then it's not even the historical performance of that because I'll get a model today ,and I'm going to backtest that model, and that's very different to the realization of that model if you were to kind of run it live. And obviously this is evolving through time.

So that evolution in itself makes that index deviate from the sense of a benchmark because of all those active decisions that happen within it, both in the composition as well as in each one of the components. So, I'm pausing.

Well, what I wanted to try and also make clear to people listening to us today is that this may sound trivial, right? Do we have the right definition of a trend index? I mean, but actually it's not that trivial. There are big investors where they are paid bonuses based on their outperformance of these benchmarks, in terms of their own portfolios. So, these are not completely trivial questions.

I'm sure you're also familiar with the other SocGen index, which is not based on managers, but it's the trend indicator that they do. And I look at that on a daily basis just to get a feel for how that's behaved.

And although I don't know long-term, if it's, you know, the proper design or the proper choices in terms of lookback, and so on, and so forth. But in terms of the daily direction, it seems to align pretty well with what I see from kind of pure trend performance. So that is interesting in some ways.

And maybe, we can also, obviously, encourage our friends and peers in the industry to think about whether there should be a better benchmark for what we do that is not based on the commercial success of companies. I know there are lots of other indices that I don't track. I think, actually, you guys have a index, Credit Suisse has an index, and I have no idea how they calculate these, and so on, and so forth.

It's not like there's a lack of indices and benchmarks, but the question is, do we have the right ones and is there something again we can do to improve it and, in that way, also perhaps help investors along the way? And of course, what would be quite interesting, and this is maybe something we can throw back to Andrew and that is, well, if we did come up with a better index for trend following, even if it's not based on managers, could that be replicated? Probably could, but again, that would be interesting. Then you could maybe have a replication of a true index, so to speak.

Do you have any thoughts on if one were to put something together like that, how would you even go about it?

I mean, I think it would look very similar to some of the simple bottom-up models that you see flying around in a number of academic papers for the last 10, 15 years. I don't think it would be too dissimilar.

And frankly, I think what we try to do, and I think, at least the QIA space across all banks, in a way is kind of following that path, obviously with the nuances here and there on the design, on the choices, and so on, and so forth. But I think that's the principle here.

There are some attempts by some third-party service providers in the QIA space to build aggregated strategies or aggregated representations of a specific strategy type utilizing versions across different banks that, you know, in principle resembles what the SocGen Trend index does with itself having, you know, its own survival biases.

You know, you can expect that some of the strategies that do not perform, you know, banks can decide to take them out or maybe replace them by buying new variations of their models. So, very similar, at least in nature. Now, could they be a good benchmark? Hard to argue, right? Hard to argue.

At the end of the day, what is the best asset allocation benchmark? Nobody knows. I mean, people use the 60/40 partly by convention rather than academic prudence or economic sophistication, in a way. So, it's a matter of definition.

I just find, ourselves, or at least I'm expressing here an opinion that calling a trained beta something that, in reality, could be an aggregated performance of managers who are also trying to provide alpha, could be an assumption that the aggregate alpha is zero or cancels out, leaving the index just with a beta. Which can be a fair assumption simply by arguing that all those gimmicks, in one way or the other, just cancel out, or we can make the argument that there is some alpha there and that alpha we're expecting to see it performing. But in reality, these are programs that are put together today and are tracked going forward rather than backwards in time. So, we don't have the backtested version of those models today.

So, all these are nuances, and I'm not really trying to make a point beyond the fact that I think these are important items.

Sure, fair enough, fair enough.

And that's it. I'm not saying it's a bad or a good benchmark. Just not to be misrepresenting anything here.

Okay, well, let's dive into the last topic and spend a few minutes on that. It's something that, you know, both you and I have tried to kind of familiarize ourselves with because it's based on the latest LinkedIn update by our friend Dave Dredge from Convex Strategies. And he, you know, writes a wonderful update every month. This one I found particularly interesting. I don't know exactly what sort of caught my attention to really dive in and bring it up on this show. But a shout out to Dave for putting it together.

Now, what it's all about is, you know, the idea of short-term capital preservation, which a lot of investors focus on, versus how do you actually maximize your long-term compound returns when you build these portfolios? And, of course in his view you should define, you should build your portfolio very differently to the way portfolios are built today. Institutional portfolios are what I'm referring to here. And, of course, I agree completely with that.

But what's interesting about it is that the starting point of this is from a report that came out recently from one of the sovereign wealth funds in Singapore, GIC, that published their review where they disclose their annual long-term returns. And it kind of gives people a chance to look into this and to see if anything has changed, if they're starting to adapt some of the ideas and proposals that Dave has come up with, and so on, and so forth. I think, for me to really dumb it down and then I'd love to hear you maybe talk a little bit about it in a better way.

I mean, to really dumb it down. It kind of goes back to what is it? And again, I hope I'm not misrepresenting Dave here, but you know, what are people really trying to optimize for often, when we look at these portfolios? And they tend to target some kind of long-term average return. So, they build the portfolios to look safe, to stay close to this long-term average return, which often leads to slowing down the portfolio, adding a lot of strategies that essentially are highly correlated. So, they may look safe most of the time, but when trouble comes, they kind of all fall at the same time.

What Dave is proposing is that, actually, if you want to build your portfolio to have the highest long-term return, a little bit like the race car analogy that he talks about and also has talked about on this podcast, is you kind of have to build the car to have the best brakes. Because if you have the best brakes, and in this case the best strategies, to help out when things get difficult in your portfolio, if you have the best brakes, it allows you to drive faster and basically win each lap instead of just, you know, aiming for an average lap time, so to speak. So not very elegantly explained, but I think people would get the gist of that.

And since you, obviously, come from a much more institutional world and you speak to a lot of the people who sit with this problem, I'd love to hear what takeaways you have from this. I actually feel it's a really, really important discussion we need to have, especially in light of the fact that, given fiscal policy changes that we now see where governments essentially are really spending much more money than they can afford, we know that the issuance of fixed income products will just continue to grow.

So, that means that you may end up with a lot more of that in your portfolio, which is not necessarily something that's going to help you build the best long-term compound returns. Even though people perceive fixed income markets as safe and probably well yielding at the moment. Well, on a notional level perhaps, but certainly not on a real.

Anyways, let me stop rambling here and let me bring you in, Nick, to hear what you thought of this.

Thanks Niels, a lot for also raising that. And it's a topic actually quite close to my heart and work that we've done over the last few years. You made some very, very good comments here. I think you flagged the importance of cumulative returns and the importance of compounding returns.

Just to make it super simple, losing 10% would require you an 11% gain to go back to flat. So precisely because arithmetic returns, which are the ones that we realize and they're not logarithmic, will get us into this disadvantage when we lose X percent and we need X plus something more just to bring us back to flat. And I think that's the gist of it here. Like short-term losses would hurt substantially the benefit of compounding returns over the longer term.

And there are many directions, actually, we can take this discussion. Maybe we start from something very, very basic. So, mean variance, you know the Markowitz Portfolio Theory, the one that obviously has been guiding all financial economics for years and years, obviously has been evolved in a variety of directions. That's a single period model.

Basically, it says investors observe today, they care about a period, they allocate, they want to maximize expected returns for that single period, they want to minimize risk or effectively they want to maximize Sharpe ratio. That's, to your point, to the article's points is about getting out of the Sharpe ratio maximization world. And what really matters here is how we can compound returns over the longer term. Which, however in itself, brings the importance of short-term risk mitigation to the forefront.

And I think here, the difference to just a simple mean variance single period model is that we're focusing on different horizons depending on which part of the portfolio we're looking at. The expected returns are now a long-term capital appreciation, whereas the risk that we want to minimize becomes more of a short-term risk mitigation.

And I think that's where the article goes, that is the direction that it's kind of taking. And if I cannot draw parallels, it looks to me very similar to how a carry harvesting investor would perform or would operate. Like, if I chase for yield, the last thing I would want to happen is to have a spot impact in my investment.

Genuinely, I should go and buy properties that deliver high rental income for low mortgage payments and hope that the prices of the houses will not change. So, literally just get that yield, that cost of carry, or that positive carry for whatever is worth - the rental yield minus whatever the mortgage rate is, and just live there forever in perpetuity.

The main issue that we would have there, assuming all the rest is the same, is that the spot prices drop, the house prices drop, and then the capital gain losses would be more than what the rental yield would bring to us. So, in that context, we would want to achieve this long-term capital appreciation, this long-term yield harvesting, while having controls on those spot reductions or spot movements, in a way.

So, for that to happen, even visually or I guess mentally, one can think of a path that tries to grow very smoothly and tries to moderate short-term drops, because their short-term drop doesn't have an impact on the day, but does have an impact on how the accrual or the accumulation of wealth will play out over the years. And in reality, I would even add to your point that it's not just the terminal wealth after 30 years that matters. I would make the problem even more important for investors that care about solvency on an annual basis.

If you're a pension fund, yes, you care about 30 years of solvency, but you actually want to make sure that you can pay your pensions on an annual basis. So, there are checkpoints of importance. And the same thing, for example, would happen for income funds. They want to be able to finance that income while at the same time preserving the capital that will continue to accrue returns, over the years to come, for them to be able to sustain that payment, that coupon payment out.

So, in this regard, and that's why I kind of agree with some of the points made here, is that short-term risk mitigation is very important for a longer-term capital appreciation. We've done, back in the day, some Monte Carlo simulations that we basically say if you experience the weekly or the monthly returns of a particular strategy, but you experience a sequence of those, let's say 12 months randomly sampled, or 52 weeks randomly sampled, or 24 months randomly sampled, so that you can form stylistically how an annual return on a biannual return would look from a specific investment. And of course, you can add some noise in those samples, and so on, and so forth. You can make it a bit more statistical.

What you find is the following. A strategy, or any investment, that contains some form of downside risk mitigation for the very short-term, be it long volatility, be it trend following, some form of defensiveness would not necessarily have a significant difference in the average monthly or average weekly arithmetic return, but the simulated compounded return would be substantially skewed to the right.

Said differently, a negative return, if smaller, when it appears in the first part of my random sample, has more chances of being mitigated over the year if I accumulate returns, if that return is actually quite smaller because of a risk mitigant that I have in my portfolio.

There's a lot that I'm basically saying here, maybe some of that part is a bit technical, but in reality we don't have to go down to the dynamic programming and stochastic control and how some of those single period models became longer-term and multi period models and how we can expand economic theory and those portfolio construction mechanisms into those models.

I think just keeping it high level, the value that I see in this type of article and this type of discussion is the importance of short-term risk mitigation not for the moment, but for what is to follow, not in a month, but in a year, in two years. And that goes back to the point we made possibly in the very beginning, whereby if a long volatility exposure or a trend following exposure is purely assessed in isolation, the whole purpose is actually completely missed. We have to see that in the context of what are we gaining at the times that we need it? There are many negative expected returns investments that we can have outside, or maybe negative expected return costs that we incur as human beings.

We pay insurance. In reality, insurance is a ‘negative expected return investment’. You know, you keep on paying and you actually don't want that to pay anything back because you know, if that were to pay something back, it's actually something bad happening, but you kind of have it there on purpose. And, I think that that's how we can circle the whole thing, and maybe I'll make a post very shortly. That's how I can circle the information of that report or that article. Some of our work, some of my work, and some of the experiences that I have seen over the years in terms of importance overall; short-term risk mitigation, medium-term risk mitigation for trend following purposes, for long-term capital preservation. Pausing here.

Yeah, absolutely, well said. And yeah, I agree. It's definitely something that I'm sure we'll continue to discuss. Everyone should go and read Dave's LinkedIn updates. Not just the latest one, but, but all of them when they come out. And he published them under his own name and also Convex strategies, of course.

And I think Dave will join the podcast in a couple of months, so I'm sure there will be a chance maybe to dive into it even further. That's pretty much what I had in my notes today, Nick. I think we've done an okay job in trying to cover what we set out to do.

Of course, for people who want to show their appreciation for the hard work you do in putting together these outlines, feel free to go to the favorite podcast platform and leave a rating and review that really does mean a lot to us.

Next week I am joined by Moritz. He comes back to the Systematic Investor as a guest. So that would be fun. And I know we're going to be discussing a new paper that's coming out in the next couple of days because it's coming out from the firm that I work with, which is a rare thing. We hardly write any papers. But I don't want to set expectations too high, but I think it'll be fun.

And we're going to be trying to answer the question what is the best alternative investment, really? So hopefully you people will be back for that conversation. If you have any questions for Moritz or me, you can send them to info@toptradersunplugged.com and we'll do our best to discuss them.

From Nick and me, thanks ever so much for listening. We look forward to being back with you next week. And until next time, 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.