232 Systematic Investor Series ft. Nick Baltas – February 26th, 2023

[00:00:00] 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. If you’re new to the show, I hope that today’s episode will trigger your curiosity enough to check out the back catalog and listen to the past episodes that you may have missed, like my conversation with Rich last week where we went into a lot of details as to how to avoid overfitting your trend following models when designing your strategies.

Also, allow me to encourage you to check out the CTA series where Alan and I have managed to line up the real decision-makers at most of the largest CTAs in the world for some very meaningful conversations on many of the most important topics concerning our rules-based approach to investing. I hope you will enjoy and appreciate all the time and access we have been given by these true industry leaders. So head over and check out these conversations after you’re done listening to Nick and me today.

Nick, it’s great to be back with you this week. It’s been a little while. How are you doing on this Saturday morning in February?

[00:01:26] Nick

I’m doing very well, Niels. Thank you so much. It’s a pleasure to be back to the show, I must say. I’m doing very well. It’s a very heavy traveling period for me, and up until the end of March, I’ll be going around the globe, seeing investors, speaking to clients and so on and so forth. So at the very least, a Saturday with family at home is all I’m asking for. How are you doing?

[00:01:47] Niels

Exactly. Yeah, I’m doing well, same kind of situation. Lots of traveling at the moment and then we’re both being ripped out of our family time for at least 60 minutes or so to have this conversation. So I really do appreciate that.

[00:02:02] Nick

It’s well worth it.

[00:02:04] Niels

Exactly. Now, we’ve got some great topics actually today and a couple of questions that we’re going to tackle. Before I do that, I haven’t done a market wrap for a little while. So let me just quickly run through what took place this week because it was another interesting, although holiday-shortened week, and it was not free from the wild unexpected economic data that we’ve been seeing since the start of the year. Activity continues to surprise to the opposite as the inflation backdrop. Existing home sales fell 0.7% from the last month’s measure while the second look at the fourth quarter GDP was revised also a bit lower to 2.7% from 2.9%.

However, the most eye-popping data for the week was the personal consumption expenditure price deflator, the Fed’s preferred measure of inflation. It rose to 5.4% year over year. Now, that measure when paired with the above-expectation CPI released mid-month has taken the steam out of the nascent bond rally that started this year. Interestingly, the usual suspects of inflation hedges like gold, silver and the self-proclaimed Bitcoin have all taken a dive along with stocks and bond prices and currencies against the US dollar. So it seems like inflation isn’t what it used to be and that many investors are left scratching their heads.

Now, reacting to all of this data, trade has continued to push bond yields higher. The 2-year note closed the week at 4.81% as it pushes above the Fed overnight rate on its way to 5%. Similarly, the deeply inverted 30-year bond rose marginally to 3.94%. Speculators are looking for Fed funds to peak at 5.4% by August of this year but there was talk on Friday that rate will need to go to 6% to contain inflation. Now, equities continue to react to the Fed’s renewed hawkishness and ongoing economic strength with the S&P 500 closing the week below 4000, nearly 18% below last year’s high.

The NASDAQ has gapped lower at the open for 10 out of the 15 days since the February 2nd high, so there’s definitely a change of character in the NDX since the rocket rise we saw in January.

All right, Nick, with that out of the way, what I’d like to do before we dive into all of the topics was really to get your thoughts and comments on what you’ve heard from the CTA managers that we have had on the show in the last few weeks. Now, I understand you may not have listened to all of the conversations of course, but I am curious since you’re also obviously overlooking lots of money in this space. I’m curious to hear what you found, maybe a bit surprising, most interesting, wherever you want to go with this.

[00:05:00] Nick

Yeah, sure. First of all, I would have to congratulate you on the work you’ve done there. I do not think that something like that has happened in the past, in the sense that it’s an industry that somehow we think we have a good sense about and somehow we kind of understand what they’re trying to achieve. But laying them out, over the course of a month or two, all of them expressing their views specifically after this great year was really, really, really interesting by all means.

It’s almost as if you’re trying to listen to the episode and then a new one comes out and you’re like, “Oh, I need another two journeys to work and back so that I can listen to the new episode.” I think I’ve listened to probably like 80% of them, seven or eight of them so far. I think I have one or two left. So I guess some first reflections on those episodes, number one, I would probably say that it becomes very apparent, what we have discussed in the show and you have discussed in the show several times, that being that all trend followers are CTAs, but not all CTAs are trend followers in a way.

Obviously, this was primarily focused on trend, but it becomes apparent how whether that is portfolio construction, whether that is dynamic sizing, whether that is kind of blending different premia, whether that is blending different strategies, how different managers have been trying to navigate over the last years or decades the space of delivering returns, risk-adjusted returns. So that’s the first point I would make. I don’t think there is a particular consensus, but greatly, trend is the driving force. But then what you do on the back end of it seems to be very diverse.

So there are some, I wouldn’t call this agreement because I think that’s a very bold statement, I would say differences in perspective. That’s the first point I would make. The second point I would make is that, notwithstanding the obvious survivorship bias, it’s incredible to hear people being in this business for like 30, 40 and 50 years. That gives credit to the work they have done. That’s the second point.

The third point I would make (which comes also a bit closer to my world) is this, when we’re discussing about how we design trend, when we’re discussing how defensive or not defensive or reactive or not reactive they are, the whole discussion pivoted towards a sophisticated discussion with the end client on the utility function, on the expectation, on what we’re trying to solve for. I think no profile can solve for any objective, but certainly, some customization is required.

But for that to be achieved, there is a long process and discussion with the end investor as to what is the objective function they want to solve for. So I think that was mentioned by a few people. I think Marty was the one that spent a bit of time on this one. I couldn’t relate more to it. I couldn’t agree more. So this is some sort of first reflection, but by all means, extremely good job.

[00:08:02] Niels

Well, that’s very kind of you to say, Nick. I think it’s interesting. You pick up on a couple of points. Certainly, the last one is probably the one that I wouldn’t say surprised me but I was really reminded about how the industry has evolved and are looking much more like solution providers. It’s not that I necessarily agree with it, because I think sometimes it can also be very hard for the clients to actually know what’s best for them. They might believe they know what’s best for them, but I’m not sure it always is in that way. But it doesn’t change the fact that our industry has moved in that direction. I’m sure, at your firm, it’s very much in focus trying to deliver different customized solutions. So very interesting points.

We’re going to dive into some of these points because, as I gave you the heads up, I’m actually really interested in your thoughts on some of the questions that were put to these managers, because of course, you are in the same category as these guys so we want to hear your thoughts as well. Before we do that, I also wanted to maybe ask you a broader question and that is a little bit since we spoke in November, if there’s anything in particular that has caught your attention. It could be whether the conversations you’re having have changed now that the year of 2022 finished, so to speak, in a certain way, the way the year of 2023 has begun.

So I’m just curious. I know you’re having lots of conversations at the moment, for example. So yeah, whatever you can share would be quite interesting.

[00:09:47] Nick

I would say that obviously 2022 was a great year performance-wise. So far, 2023 has been a bit more of a mixed bag if I were to say. If we look into all the indices or various variations of trend, it’s kind of flat; a bit up, a bit down depending on the variation. But I would say the average trend follower has had a flat year, possibly less well in January, possibly a bit better in February. It all averages out to a flatline. I wouldn’t say that this raises any concerns about performance, but I think it’s more of a reestablishment, possibly, of trends on a forward-looking basis.

The question, however, that comes around, is that following this great year, where are we? What is the expectation on a forward-looking basis? By all means, I don’t think that I can have any crystal ball here, but at least looking at the environment, my view hasn’t changed from the one that we had back in November, that being that this persistence in uncertainty, by all means, is an ingredient of trend formation. Of course, we might argue, where is the trend? Well, subject to that being formed, that increased uncertainty will make it more likely to persist.

I think several of the speakers in the CTA series kind of suggested that this macro environment is the one that, if anything, can make trends, should they start appearing to persist. I don’t think that the investment thesis has by any measure changed. I think I made that point back in November. Looking into these types of strategies should not be an act of regret by all means. But at the same time, timing them, and possibly, we’re now front running ourselves (possibly you were planning to ask me about timing, I don’t know). But I don’t think it is easy to time it.

So if there’s any question or discussion that we’re having over the last two months with clients, it’s whether this great year is about to continue and whether this is not too late. By all means, had we had the same conversation a year ago, we would be like, “Yeah, okay, after a number of years of underperformance, it probably isn’t the right time.” By all means, what would we know? So, I think that’s the point that I’m spending most of my time discussing with clients, whether it is the right time now or not, how does the environment evolve, and on that basis, whether it makes sense to look into the strategy.

The last point I would say is that (and I think you also discussed that. Who was it? Probably, Kevin. I think it was Kevin, when you were discussing about the inflows and outflows of the industry over 2022), if I gather (and I can share some of my own experiences), it was a year of great performance but that was more about bringing back the attention into the theme as opposed to necessarily leading to substantial inflows. Of course, there have been inflows here and there, but the question is whether that has been reflecting their performance in its totality.

I think we’re yet to see that in the broader industry and I think that part of the background is the vetting that all the biggest allocators are currently doing at the moment; whether it’s now the time and possibly it’s not the time, and so on and so forth. So that’s another point I would make.

[00:13:06] Niels

That’s a great point and I certainly see a lot of the same things in my conversations so far this year. We’ll get to that a little bit more later. Let me quickly do a short trend following update, which I don’t think I did last week. But as you rightly said, Nick, after a soft start in January with a lot of the longer-term trends kind of correcting, February so far has been much more constructive for trend following strategies mainly, I would say, still, fixed income being quite dominant in the portfolio with new rising yields across the globe so to speak.

But there have also been a few other things that seem to be doing quite well in the currency sector and even grains and metals, to some extent, have done okay. Equities? Yeah, probably a mixed bag, I would say, across the scope. Of course, also, just speed of trend will have had an impact, because we’ve seen a little bit of a change in sentiment in stocks the last few months. So, depending on how quick you’ve been to react to that defines your performance so far. Energies the last few weeks, probably not a lot going on in that area, and I would also suspect that by now, exposure in energies are fairly muted across the board.

But overall, a solid comeback after a soft start, as it should be, so to speak, when you just look at what the markets have done. I think maybe worth mentioning is just to say that I think these strategies continue to be a very meaningful diversifier to both stocks and bonds. So that actually hasn’t changed since 2022 and we will see how investors will react to that in the upcoming investment committee allocation decision so to speak.

My own trend barometer finished the week at 41, so it’s kind of neutral so far this month. We’ve only got a couple of days left, we’ll see. But if we look at the indices we like to track, the BTOP 50 Index is up about 1% for the month. It’s up about 1% for the year. The SocGen CTA Index is up about 1.4%. That brings it to up 0.6% for the year. SocGen Trend is up about 1.2% (and this is as of Thursday, yesterday; Friday was a good day so it’s going to be a little bit better for all of them). The Trend Index is down about a 0.25% so far as of Thursday, and the SocGen Short-Term Traders Index is down about 15 basis points, both for the month and for the year I think, so it’s pretty muted so far.

Now, as I said, we’ve got a couple of questions that I want to tackle and then we’re going to dive into some of the topics with you, Nick. But I wanted to address these because they were prompted by, one, people knowing that you were coming on. But also, there was a question as a follow-up to last week’s questions which I think you’re perfectly capable of answering. It’s a question from Adam. He has a question that relates back to the episode I did with Rich last week.

Adam writes, “There was mentioned of two types of systems, a number of different signals applied to a single market or two, a global system which is applied to several markets. I’ve been running a model which is backtested on a single market, performed well on a backtest 10 years, and I wanted to apply it to more markets, about 20 in total. It also performs well on paper, but it’s only working on 12 out of the 20 markets with positive returns on average, whether equally weighted or liquidity weighted. The portfolio returns is stable, but 20 components have ranged from some being up more than 100% and some being down more than 50%. Because the average, let’s say, is positive, should I stick with it still? At what point would I assess whether to modify the strategy for the eight underperforming markets and so on and so forth?”

So thanks for that, Adam, that question, because I think it’s quite a good talking point in general, because, as systematic managers, and especially if you choose to apply universal rules, meaning using the same rules on all markets, then clearly you’re going to have periods where some markets are simply not making money and other markets are doing really well. We see that all the time.

So I’d love to hear your thoughts about this in general, Nick, how you think about that challenge and maybe also if you have some “golden rules” where you say, “Well, hang on. If it hasn’t made money for 10 years, for example, I’m going to change something or I’m going to leave it out.” So I’d love to hear your thoughts.

[00:17:57] Nick

Yeah, that’s a very good question. We can address it in a number of ways. The first one I would say is that I think there’s a philosophical point to be made on how we think about a trend following strategy from a construction standpoint, and I think we might discuss that a bit later. But when we’re looking to each and every market in isolation, then we effectively have a collection of single strategies.

Conversely, we can think about a selection of markets that are put through a portfolio construction engine and then we deploy risk as a function of let’s say individual market riskiness and possibly correlation structure and possibly the strength of the signals. Now, what is different between those two is that in the former case, the assessment is done purely in isolation market by market, but in the latter case, there are other dynamics coming into play by the principles of diversification, overall risk balancing or risk distribution.

In the former case, when it’s more about single markets and them being assessed in isolation, it can become a parameter optimization exercise in a way. In other words, if a signal works in a particular market but doesn’t work in another market, what does it tell me? It can very well be that we’re lacking one or lacking the other and I don’t know exactly which side of the coin we could possibly take. So I tend to be quite reluctant, looking in each and every market in isolation, if there is an underlying premise that you expect trends to work.

If I look at it from a portfolio perspective, an additional market added into an optimization would, by construction, make the risk contribution coming from the rest lower Because it’s a zero-sum game. So, if I add one, let’s say, equity market more, then I would have to (all else being equal) reduce the contribution I’m getting from the other equity markets. Therefore, here, it becomes more of a story of risk distribution and whether that addition can help me from a diversification standpoint.

In addition to it, when we say that some market is not working, what does that really mean? It means that over the longer term, we have a negative return. Surely. But what if the periods that it has a positive return are the times that all the rest are doing badly? So here’s a diversification dynamic which when assessed in isolation is not obvious. You just say, “Oh, this market over the longer term has been detracting performance.” But how about that market having done those few periods of doing well when everything else was doing badly, and this is only allowed when we look into the portfolio in its totality, not in isolation.

So I don’t think the answer is easy. The other point that I would make is that something doing badly and being removed from a system that somehow we feel confident about is introducing a bit of a survivor’s bias. Do we have a crystal ball? Probably not. So doing the asset selection purely on what performed well recently or over the last 10 years, unless it’s based on some economic or fundamental reasoning or possibly a liquidity reasoning, I think the hurdle to go past, at least for me, is extremely high in terms of removal.

I tend to look at it from a portfolio setting. I do not tend to look at it from a market-by-market perspective. At least that’s my view, but by all means, I’m happy to be challenged further.

[00:21:46] Niels

No, I don’t think there’s any challenge there, Nick. I think that’s very well said and I think it’s an incredibly important point you brought up. Because actually, you could say exactly the same about trend following in a multi-asset portfolio, that even during the periods where trend following doesn’t necessarily knock the ball out of the park, we still may actually have a positive contribution overall to the portfolio. So that in itself, I think, is quite important. What I will just say to you, Adam, is a couple of things.

One is, it’s hard for us to assess when you say 20 markets because that could be 20 equity markets and bond markets or is it really 20 very diversified markets. So what I will say generally, why don’t you test the classical 50 markets across financials and commodities, see what it looks like? The other thing I will say is that, unfortunately, even though it might cost a bit of money to get hold of, 10 years’ worth of data is not enough, unfortunately, especially because the last 10 years have been so different from the last 40 or 50 years.

So you should go back if you can. Get your hands on some back-adjusted data for 30, 40 different commodity markets over the last 30-plus years to see whether that pattern you see really holds up, and then as Nick said, you need to then select a portfolio of markets that you want and can trade. Secondly, then look at the portfolio returns because they may actually be pretty good once you combine them and that’s obviously part of the secret of trend following.

I often hear people talk about trend following as something they can just apply to equities and it’s the greatest thing in the world. I’m thinking, no, no, no, no, no, no. Trend following works because it’s applied on a diversified portfolio, not on a single market, not on a single sector, not even on two sectors. So yeah, just a few things for you to go home and do further, Adam, and hopefully you will get some more insights that you can use.

The next two questions are from Tim. Tim writes, “This week, I have a question on the topic of portfolio volatility, but it is not the typical question about volatility targeting that you discuss with your guests anyway. Having met Nick Baltas in person and having read his work on different ways of using volatility in the trend following portfolio construction, I would like to get his views on the following. In your recent podcast with DUNN boss, Marty Bergin” (so that’s my boss), “Marty made an interesting comment along the lines of DUNN adding risk when markets are trending and reducing risk when markets are less trending. So other discussions on trend following have pointed out that trend following the trend followers may not be the best idea. Personally, I tend to disagree and I’ll explain why. Of course, it might not be ideal to chase performance and end up buying into the strategy at a high. Similarly, with trend followers, it is important to hold on through drawdowns and not sell them at the low. However, having done some simple calculations on the historical performance of some trend following strategies, and this works for DUNN as well, increasing exposure when the strategy performance is positive and vice versa would have added to the overall strategy performance. The observation would imply that trend following the strategy performance curve may have some benefit. What do you and Nick think about the varying of the portfolio volatility/exposure based on its own recent performance history?” So there we are, Nick.

[00:25:15] Nick

It’s a great question. Effectively, the question says I build a system (that being trend following) and should I see it performing, I kind of add to it or I increase the risk or I increase leverage. All of these are kind of equivalent. So why do I love the question? Because the underlying premise of trend following is that it’s in every market so when it performs well, it becomes more likely it’s going to continue, and therefore, that’s at the micro level.

Now, the question is at the macro level. If all of those markets, in conjunction, are doing better, does that tell me anything about the aggregate exposure and whether I should have more at the top level above and beyond what I have at the individual level? So in a way, it’s very similar to the dynamic sizing point, that being do you dynamically change the exposure to markets as a function of many things but one of those things being the recent performance? Because if the signal becomes stronger, you might want to deploy more risk to those markets that have been trending more.

Therefore, in aggregate, the question is whether you should be doing dynamic sizing of trend itself. In a way, it’s the same thing. So what are the benefits and the downsides? The benefit is that to the extent that at the underlying asset, we do have outperformance following good performance, it can lead into a hypothesis whereby the aggregation of markets performing well may suggest that we should be increasing the exposure. What’s the downside? The downside is just basically miss the trend if it starts kicking. In other words, suppose that we are in a period of no significant trends.

If that is the case, then the premise of sizing the top-level exposure as a function of that would mean that we don’t necessarily allocate too much into trend. Should there be an ignition of trends, it will take us some time until we get there. Now, I know that I’m not answering the question straight away because this is actually an active research topic for us in the present moment, as in, we are looking into whether it makes sense not to have a volatility target at the portfolio level, but rather have a dynamic volatile targeting mechanism which itself is informed by the average trendiness in the markets.

I think it’s a good hypothesis to be tested. It is unclear to me whether that is a premise that we can systematically introduce into the portfolio. Possibly, this is the element that at the top level, above and beyond just trend, is what an active manager can decide upon. The overall location into a trending strategy becomes more of macro view rather than a systematic view on its own performance. So that’s the only point I would hint here. But setting that aside, I think it’s a beautiful question.

[00:28:29] Niels

Yeah, absolutely. Let me address it a little bit since I do work for the firm that was referenced in the question. First of all, I think where we were coming from (in terms of saying, ideally, we should be increasing risk when there are lots of trends, we should limit risk when there are fewer trends), we came to this 10 years ago. In 2013, we changed from always having the same risk. Now, we were never targeting volatility. We know volatility of the portfolio will change as volatility in the markets will change, as correlations between positions will change.

So that’s not what we’re doing, but we want to target the risk we take. From 1974, when the firm was started, and up until 2013, we didn’t have a way to systematically distinguish between a good environment and a bad environment, but we knew that clearly, when you have a trendless environment, you shouldn’t necessarily run at full risk, so to speak. So that intuitively made sense. We just didn’t know how to systematically distinguish between a good and a bad environment until 10 years ago when we implemented some research that we had been working on for quite a while to do that.

I think what we’re doing is not necessarily trend following the trend follower because I do think that there is some truth to the fact that you shouldn’t do that. But when you start looking at your signals and your strength, etc., it might just give you clues as to the environment that you’re in. You have to combine that with a few other things of course. It’s not just looking at signal strength. But I do think it has certainly, in our case, helped significantly in terms of delivering the same performance but with a lot less volatility.
Of course, for regulatory reasons, I can’t mention the percentage improvement we’ve had but it’s certainly been meaningful. So let’s leave it like that.

Now, I want to go on to the second question that Tim writes, which came in like a day later, where he says, “Incidentally, I remember asking Nick six years ago when I met him about the AUM, that the three big groups of systematic strategies, i.e., one, CTAs; two, risk parity; and three, what vol control we’re running. My recollection is that he said” (and of course, I don’t know if you remember this conversation but), “he said that there was a bit more than a trillion in AUM, with risk parity and bulk control each 40% and CTAs being the smaller group about 20%. With what has happened in markets, in recent times, i.e. risk parity performing badly, trend followers winning, I would like to think these relationships of total AUM have changed. If you have a minute to ask him about this also, I believe that would be of interest.”

So is that something you track, Nick, at all, in terms of where you think the AUM is in different types of systematic strategies?

[00:31:45] Nick

That goes back to the discussion we had back in November. I don’t know if you remember, we had this discussion about whether systematic investors, namely, trend followers (but also as a byproduct risk parity involved targeting) can cause a drawdown to become more extreme, and so on and so forth. That’s where the question comes from. Because in assessing whether this type of investing can make a drawdown larger, of course, we have to understand how big that market is, and obviously, on the back end of that, how are they positioned, how they estimate risk, how much consumption they have, liquidity…

I guess the point that I had made back then was that progressively, as we go from CTAs down to risk parity down to volatility targeting, it becomes much harder to claim that we can estimate the size of the market. For CTAs and trend followers, we can have some numbers. There are databases around that. It’s like a 300, 400 mark, but we kind of know that it’s somewhere there in billions. The number hasn’t changed (at least at the top level) substantially over the last 10 years, it was a bit higher. Towards the end of Q3 last year, there was a 400. It’s now down to 380. But give or take, the CTA industry has been around this 330 to 400 mark for the last 10 years.

Now, risk parity, to me, can be anything from the risk parity in equities and bonds to an allocation framework that can be applied to anything. In a way, dynamic position sizing trend is a form of risk parity or possibly risk budgeting. So, is there double counting here? I do not know, but certainly, risk parity is, to me, partly an allocation framework, partly an investment theme on the asset allocation side. If we just stick to an equity bond complex, just looking at the numbers, 20% drawdown last year certainly has impacted that, but obviously from ‘16, we were doing this estimation of market size, it’s possibly netting out to similar numbers.

Again, I’m not following those closely, but just reading out how equities and bonds have performed in the last four, five years, we can basically make a hypothesis as to how just the risk parity on equities and bonds has evolved.

I guess the last one, that being the volatility targeting, that is even more obscure than the former two. Because we can easily argue that volatility targeting is just a mechanism that we can deploy in any form of a strategy in order to control its overall volatility profile. By all means, there are consequences to it. Some good, some bad, depending on the fundamentals of the strategy.

So volatility targeting a positively skewed strategy is going to impact, negatively, its performance. If it’s negatively skewed, it’s positively impacting it. If it’s not skewed, at the very least, it’s going to flat out the volatility profile. But having said that, when people talk about volatility targeting, I think they refer primarily to the FIA market (fixed income annuity market), these types of markets whereby the end investor is effectively a receiver of an option. That option to be priced requires a relatively flat volatility profile, and therefore, volatility targeting allows you to get there.

But having said that, we can easily claim that a CTA, that does dynamic risk exposures and then has a volatility targeting, it’s already in itself a combination of those three features. So really, getting an estimate of that market is not easy. I think at the time when I was writing this report, I was kind of saying I have conviction for CTAs and then it drops dramatically as you move away from it. But again, the industry talks about volatility targeting under the premise of making a down move more excessive. This is primarily risk-targeting, volatility-targeting equity exposures, or the FIAs with relatively traditional benchmarks.

So I guess, more than a minute here, I managed to spend a couple of them, I’m not monitoring those very closely. Has the ratio changed? Not easy to tell. Certainly, the CTA industry has been relatively stable and possibly a bit nudging up on the equity bond. Last year was definitely detrimental to it. On the FIA market, it has grown over the recent years, but it has grown possibly away from traditional markets and more multi-asset, more alternative and the likes have been put in place. So it’s not too easy to come up with a number.

[00:36:50] Niels

No. On top of that, for those who may not be aware of that, when we talk about the CTA industry being, let’s call it 375 billion (plus-minus), 40% of that is attributed to one firm, which is Bridgewater. I’ve never heard them really being classified as or describe themselves as a CTA. So it’s very hard to get a real gauge on these things. But let’s leave it at that. I appreciate the questions from Tim.

To turn the topics around a little bit now, Nick, because I couldn’t help myself, of course, knowing that we were going to speak today. I thought, “Nick oversees a lot of systematic strategies, including trend following, so why not take the opportunity to ask some of the same questions or bring up some of the same topics that we’ve been discussing with our friends in the CTA industry for the last few weeks?” So even without my wingman Alan, I’m going to try and dive into some of the topics because I really do find them fascinating and I think your perspective will be incredibly insightful as well.

I want to kick off with a topic that has been discussed many, many times, and it is crisis alpha. Now, yesterday, we published a conversation with Katy Kaminski, who of course coined the phrase after the GFC, and we wanted to revisit the term crisis alpha, as it seems to have been lost in translation a bit. I don’t know if you heard the episode or not, but before I talk about what Katy reminded us about, I wanted just to ask you, what do you think crisis alpha means and how do you think investors should think about it?

[00:38:51] Nick

I haven’t heard, yet, what Katy says. I don’t know, so let’s see. But I know her obviously and I know the term she coined for trend followers. Crisis alpha, what it means to me, crisis alpha (if we just go through the words one by one), it is effectively performance that comes when a crisis hits. Now, I think the subtle point is what constitutes a crisis. I personally look into trend followers simply because they’re price takers, simply because they expect the move to first document itself before a position is deployed. The first point I would make is that they’re not going to deliver a reactive profile on a day or two of a significant market reaction.

So it is not a February 2018 type of a profile. If the market, following months of outperformance,is experiencing an intraday 5% or 10% drop, by all means, positive exposures would be hurt. But now should that move (once documented and captured by signals) continue, then we end up first of all entering into a more longer-term correction, a more sustained drawdown. Somebody called that a wear and tear type of an environment. This is more of a macro-driven crisis. This is when the macro fundamentals make a crisis prevalent. In that environment, we’ve seen trend followers delivering this alpha, this convexity.

So to me, crisis alpha is more about, A, defining the crisis, but then subject to that being a longer-term correction, this alpha is pretty strong. I was looking at some numbers since you mentioned it. I just reminded myself. I was looking into the worst years that S&P has had in the last 25 years. The worst three years (I don’t remember the ranking, but the performance was equally bad) was 2002, 2008, and 2022. So over the whole year (and obviously, there’s some sort of path dependency here because I’m just picking points that begin from January to the end of December, but at least on a calendar basis), the three worst years have been those three, ’02, ’08 and ’22.

The best three plus one years for trend followers, over the last 22 years, have been 2002, 2008, and 2022. The plus one is 2014, which was obviously primarily driven, at the time, by the oil crash and the dollar and rates moves. Now, what am I saying here? All I’m saying is that I did not quote a flash correction. I only quoted three years that became the consequence of a more macro fundamental down move. It was the dot-com crash, it was obviously the GFC, and it was obviously last year, the Goldilocks reversion. So to me, that’s crisis alpha.

That’s why we’re looking into trend followers as a return generation during a market downturn, but not a flash correction. We can get that as well with some sort of reactivity points, but I’ll leave it there. Is that in line with, Katy?

[00:42:30] Niels

Well, you know what? This is very interesting and I’m actually really happy that you haven’t listened to the episode. But I will encourage you to listen to it on your next long-haul flight, which I know is coming up soon. Because I think what you’re expressing, to a large extent, is actually exactly what the translation of her work went into, meaning a sharp focus on performance during equity crises.

I have been on record saying I thought the term was fantastic. It still is. But I also said that I spent a lot of time in the last 12, 13 years debating with people what was a crisis, what wasn’t a crisis, and did we make money, did we not make money. So of course, we had to talk to her in-depth about this. What was really interesting is, of course, she reminded us how she came up with it and it’s not exactly what I remember.

Listen to this. She basically said, the question she was given by a pension fund was, what is it that makes a strategy succeed during stress? Of course, I’m paraphrasing. People should go and listen to the conversation. Essentially, what she came back with was she said there are three features the strategy needs to have. It needs to be liquid, it needs to have the opportunity of being opportunistic, and it needs to have no bias. Okay, fair enough.

Now, why does it need to be liquid? Well, because if there is stress, you need to be able to efficiently and quickly reposition your portfolio. That’s point number one. Why does it need to be opportunistic? It’s because you need to be able to go long and short depending on where the opportunities are. She gives a great example. She says during COVID, the best trade was to go short energy and long bonds. But during 2022, the best trade was to do the opposite, to go short bonds and long energy. Very, very powerful stuff.

Then, of course, it needs to have no bias. A good example of that would be that many people said trend followers have just made money from being long bonds for the last 20 years and so once that trend changes, you’re not going to be able to make money. Well, last year was nice in the sense that it proved that we have no bias, no emotions, we just went short bonds and there we are. That was the best-performing sector that we had. So, as I said to Katy, I love the term now, again, because you just need to explain to people what it really means.

It’s not about whether you make money specifically during a certain crisis, although your points are well taken in terms of three equity years, down years, in the last 22 years. All three years were periods where trend follows and CTAs maybe more broadly did well. But actually, we need to talk about what are the characteristics you need to have as a strategy or as an asset in order to even have the potential of delivering crisis alpha. There, you can, in my opinion, say maybe there aren’t that many other strategies or other assets where you can say that they have the ability to deliver crisis alpha.

Maybe I’m not looking at this in a broad enough picture, but for example, we were told for a long time after 2018 that the best way to protect your portfolio would be with long volatility strategies. They became incredibly popular and so on and so forth. But actually, when you look at the three characteristics, it probably doesn’t meet all of them for example. It is stuck in the same sector. It cannot move out of volatility, while with trend followers, we can go into grains, we can go into energies, we can do bonds, stocks, whatever.

So there, you could say maybe it only has one or two of these characteristics, and so on and so forth. The same with bonds. You can only be long bonds, so it’s not really opportunistic, it can’t go short, and so on. So when you hear that argument, Nick (so now I can really put it to the test here),when you hear Katy’s argument like that or reminder, how do you feel about crisis alpha now?

[00:47:00] Nick

I think we’re in full agreement. What I did not say is how those three best years for trend followers were delivered, and then the underlying drivers were so different. I think that was your point, right? It’s this dynamism and how, for example, bonds did help a lot by being long in 2008 and then by being short in 2021. It was like the dollar moves last year from the summer to November that gave performance but that wasn’t the case back in 2008. It was probably going short equities.

Last year, going short equities wasn’t a good deal by the way, because we had this very weird market with very short-lived short squeezes but certainly in a bear market, relatively unprecedented by magnitude. The equity rally in March, in the summer and then obviously towards the end of the year, that was relatively bad performance attribution-wise for trend followers. That wasn’t the case in ’08. So the opportunism if you like, that’s one of the points that Katy made, or the lack of bias is certainly there if we start looking through why in the three years that seemingly top level were good years, where the performance came from.

So I couldn’t agree more with that. The other point I would make is that precisely to your point, long vol has its own good attributes, and that is it’s instantaneous, effectively contractual reactivity, and this is something for a different type of an environment. So by all means, a holistic, defensive portfolio might as well just combine features from the volatility space but also from a trend following space. I don’t necessarily think that they solve for the same objective. There is no long volatility program with a flat or positive carry profile with long term positive Sharpe.

But by all means, I think we jointly believe (I would imagine, together with Katy and all the other guests), that trend following is a positive long-term Sharpe ratio strategy with defensive characteristics subject to defining what defensive means and subject to defining what’s the utility function we’re solving for. So I couldn’t agree more. I couldn’t agree more.

[00:49:21] Niels

Yeah, it’s great. I was so happy to be reminded about these things because I think it’s certainly going to help me in the conversations that I’m having about this, to make these distinctions and to talk about other things that people think about in terms of risk mitigation, and just to make that simple test. Do they have these three characteristics? All right. Now we’ve opened the kimono box here diving into some of the topics and questions that we’ve been talking to our friends and peers with. So I want to continue down this.

One of the other topics that we talk about, Nick, that I’d love to hear your thoughts on is this pure trend versus diversifying trend, and the evolution of our industry. I’m going to dive into Cliff Asness’ point later, but just generally speaking. I know you provide solutions in some way so maybe you fall in both camps. But can you talk a little bit about this idea of whether you think our industry has evolved to be more trend-plus type providers, so to speak, rather than pure trend, and how you look at it from designing strategies as well?

[00:50:39] Nick

I’m not sure what I’m going to say is going to be radical here, but it’s unclear to me what pure trend is by itself. What do we really mean by saying pure trend? Does that mean that any market that goes up I’m buying by allocating a dollar or a risk dollar to it and that’s it? Is it more of a system that follows signals that monitor recent performance but then it becomes much more involved in risk sizing within sectors, outside of sectors, within asset classes outside of asset classes, top-level portfolios?

Is it a system that somehow embeds risk management features above and beyond vol scaling and somehow reduces exposures when the signals become too large or possibly when the losses in particular markets become too extreme? Is it a system that above and beyond all this, is monitoring some form of macro news that above and beyond the trend direction can be impactful about performance?

I’m not sure exactly where we stop being pure trend and we become trend plus something. Because ultimately, in the premise of delivering positive returns over the longer term with good characteristics on specific market regimes, I think it is almost a survivorship necessity to bring some other features in. Again, it is unclear to me where do we draw the line? So that’s the first point.

Related to it, I can give another example. I think that’s for another episode, but even by the fact that we vol scale some markets because we have a multi-asset portfolio and therefore we want to bring commodities and equities together and all that lot, just by our necessity to do so, what we might actually be doing here is that we’re enhancing the program simply because we vol scale. Why? Well, even (just back to the point we made earlier on why volatility targeting works) if we have a market that we go long and it happens to have more frequent negative outliers that positive outliers and it has also some sort of volatility clustering, by vol scaling them, we improve them in their own right above and beyond any trendiness they might experience or not. So is that purity or not?

Well, to me, it’s a form of risk management that happens to have, coincidentally, good features that makes, at the asset level, the allocation beneficial and then trend adds to it, obviously because that’s more of a signal that captures the continuity in the returns. So I don’t think I’m digressing here. I’m trying to address the point. Trend plus only moves away from a broader family within trend when other signals are coming into the game. But pure trend, to me, by itself is not something that we can easily define. I think it’s very endogenous. Sorry, go on.

[00:53:46] Niels

I was going to say, I think that that’s a very astute observation. You bring up a lot of interesting points there. So let me take it a step further. Because part of why Cliff Asness brought this up in his paper last year was also raising the question whether we, as industry, managers, have become too concerned about Sharpe and therefore we’ve done things to our systems to essentially make them more palatable during the bad times, but also, probably giving up something during the very good times, but it could help with investors staying with the strategy and so on and so forth. So I’d love to hear your view about, how do you balance the objective of pure trend which certainly looks different from trend-plus?

In this sense, I’m thinking about trend plus something else like carry, like value. I’m not talking about just a little bit of volatility adjustments, etc. Because here’s the question. Deep down, I worry that investors look at our strategies and they look at the Sharpe ratio of the strategy and they will compare managers’ Sharpe ratio, they will compare the Sharpe ratio to different strategies and so on and so forth. But, of course, the way I understand things is that, first of all, Sharpe wasn’t invented to be used for individual strategies. It’s a portfolio tool. What we really should be worried about and focused on is, what do these strategies do to the overall portfolio Sharpe? What improvement can we see from adding the strategies?

I don’t want to change what Cliff wrote, but there is this distinction and maybe somehow we’ve become too concerned about our own individual Sharpe. In trying to improve that, we’re actually (and maybe investors are not even aware of it), we’re actually going to reduce the overall portfolio Sharpe improvement that we would otherwise deliver. I don’t know if that makes sense, but I’d like to hear your thoughts.

[00:56:06] Nick

It makes total sense. If I were to paraphrase or even simplify, if in an up market I buy equities, I would have a better Sharpe. Now, do I actually have a better profile to be added on a portfolio that includes equities? Probably not. I think the point that I would make to the question or I guess to the theme is that I couldn’t agree more that we have possibly become too drawn to this Sharpe ratio end game.

At least to me and the way we operate, the way we try to discuss with investors and clients is on a more conditional basis. In other words, and I think Russell made that point, we’re looking into performance across specific economic outcomes because that’s what we try to solve for. An element of solutions is not, “What can I add to my portfolio so that over the longer term, it has better Sharpe?” But it’s more about, “What can I add to my portfolio to cater for economic regimes that I might be more exposed to?” That is a very, very different utility function, but also a much more well-defined one because what is a long-term Sharpe ratio?

Well, if the processes are ergodic, we can have a Sharpe ratio for the last year and then we can hypothesize that that’s the Sharpe ratio for the longer term. But this is not reality. That’s not how it works. So by all means, any Sharpe ratio is a function of the period we’re looking at. Conversely, looking into more specific economic outcomes, we can have a better understanding. Again, that is still a victim of how much data we have, but at least we have more understanding of the economic mechanism.

You made the point earlier on looking into Sharpe for 10 years, it’s probably not going to work that well for trend followers because that was not a great period for trend followers. So I think to me, looking into Sharpe is not the right answer. If anything, there are strategies that have some virtues. You mentioned long vol. Well, if I look into the Sharpe ratio of long vol over the longer term, it’s negative. Not only is it negative, it doesn’t even make any sense, because a negative return divided by volatility number for more volatility would look better because it would be closer to zero, not closer to more negative.

So just a Sharpe ratio is not enough. I think we should be looking into, number one, Sharpe ratio across economic outcomes we want to solve for; number two, Sharpe ratio improvement to what we ultimately hold as a combo. I would want (and I’m trying) to avoid just looking into Sharpe ratio.

[00:58:58] Niels

I’m aware of our time, but I do want to try and see if I can squeeze in a couple of more topics with you, Nick, and that is one thing that I’m interested in is that I think narrative has become such a powerful part of how people end up making decisions, so to speak, in the end. We already touched upon it a little bit, and that is the conversations I’m having right now. A lot of the things people say are two things. One is if 2022 was such a great year for trend following, for CTAs, then clearly, 2023 can’t be. That’s their view.

The second thing I come across is that people really look at trend following as a trade coming back to the point about the timing of it. They really look at it as a trade and what I’m trying to say to them is you have a portfolio of $100, why don’t you take $90 and just put in all your views and your expectations and your biases and whatever and make your allocations accordingly? But this 10% here, that you want to allocate to trend following, you simply just have to be convinced by the data that it has a positive impact to your overall portfolio, then you need to allocate, and then you need to do nothing and have no opinion about what goes on inside (of course, expecting them to have picked an established manager with a good reputation, all of that stuff).

But essentially, you need to treat your trend following allocation completely differently to all the other investments you make, because otherwise, you’re not going to be successful, then you will end up trend following the trend follower, so to speak, which we don’t want. So given the fact that a firm like the one you work for has so much more clout than anyone else when they go out and talk to clients, how would you (I don’t want us to use the word convince), but how would you compel people, investors not to have too much of their own opinions and biases like, “If 2022 was good, then clearly, 2023 can’t be?”

How would you change that conversation; that narrative that goes on in the mind of the investors, and hopefully move them a little bit more towards what I was suggesting that this part of their portfolio, they really should not apply their own views and biases except that they should do the number crunching and they should be convinced and they should be comforted by the results they see in the data, the evidence, so to speak.

[01:01:51] Nick

I guess generalizing the question is effectively how much of human biases and behavioral biases we can impose into an investment thesis and how different discretionary to systematic investing is, because at the end of the day, looking into an allocation, the behaviors are driven by our own views or possibly by our macro views. To your point, it might be detrimental for trend itself. We’ll end up trend following the trend followers. I think the 2022 to 2023 point you made, aside from the fact that we just changed the year and nothing else necessarily happened.

In reality, it’s just like a data point, which is the 31st of December, somehow something happens, nothing happens. So I can try to defend that, in that sense. It’s just a day, right? But I guess on a more serious note, the way I am trying to have this conversation with investors, with regards to trend following and other types of profiles, is (and I think we made the point earlier on today about November), A, it could not be an act of regret, but B, there has to be some form of conscious resistance on timing. I think timing is hard and timing needs a lot of time to be proven skill versus luck.

Therefore, because human beings have very short memory, they can very easily attribute a good call as skill and a bad call as bad luck. Therefore, having not lived through that experience, it’s hard to be convinced. Living through that experience, can be (sometimes) a good outcome, but it can be (sometimes) an outcome whereby career risk is at stake, and we’re here to basically address the latter. Addressing the latter has to somehow be related with a more strategic way of looking into a portfolio, looking into specific outcomes, looking into the data.

As you said, running the analysis, crunching the data, vetting the people behind the engine, but have some form of conviction that A, dynamic timing is a hard exercise, but number two, that some form of trend allocation and others, but trend allocation is more of an essential add on than a good to have. Acting upon it, following a bad year or a good year, to me, is equal to just acting upon it irrespective of any short past experience. So that’s how we’re trying to approach it.

By all means, just purely looking at the Sharpe ratio and looking at the data, you’d be inclined to do like 100% in trend or 0% in all the rest but that’s not how we should be allocating money anyway. That’s my perspective. I’m not sure if that makes sense.

[01:05:00] Niels

Yeah. Absolutely. Makes perfect sense. Now, the final question I definitely could not not ask you, because I’ve asked everyone else, and that is, what’s the one thing you hear about trend following that you disagree with the most? I’d love to hear your answer to that one. We’ve had some different answers and some similar ones. But I’d love to hear what you think is the one thing that you disagree with the most.

[01:05:32] Nick

Following a few years of underperformance, that it’s dead. That’s it. We’ve seen that, right? I think it was also in the news. In the news in January, there was one article basically saying, is trend dead? I’m like, first of all, define dead. Is it like dead for a shorter period of time, for a longer term? Should we forget about it? Especially when we go into the news and we read the news, there’s always a bit of vagueness in statements people make.

I remember, what is the book? I think it’s Superforecasting by Tetlock. He makes a great point whereby when we make a hypothesis, unless we bring together with it a probability as well as a time horizon, it is impossible to be assessed, and therefore, we’re always on the winning side. If I tell you that tomorrow, it might rain, if it rains, I’m going to tell you, “Listen, I told you. You should have gotten an umbrella.” If it doesn’t rain, I’m going to tell you, “Listen, I didn’t tell you it was going to happen for sure. I told you it might rain.”

To put that into context, if I don’t even say tomorrow and I’m just going to say it might rain or it might not rain, I even make that statement more vague because I don’t even have a horizon for me to be assessed. I’m always right. I guess all the “likely” and the “might” and all this terminology people use is just to create more vague statements. I’m always consciously trying to be much more specific when it comes to probabilities and time horizons. Anyway, I’m digressing here but I think it’s important.

[01:07:16] Niels

No, it’s great, and I appreciate you elaborating on the one thing that you hear that you disagree with the most, which of course, was not a big surprise, I have to say. We have heard that from a few other people.

Anyways, Nick, this was great. This was delightful as usual. If you (listening to this episode) feel that you also are getting some value from them, why don’t you go to Apple or Amazon or Spotify and leave a rating and review. We so much appreciate that. Not only does it show us that you liked the content, it also helps other people discover this podcast.

Next week, I’m going to be joined by Alan, so maybe we will talk about some of the stuff that we feel we have learned during the current conversations with our peer group. Of course, you should also email some questions to us. info@top traderunplugged.com is where you do that. With that said, from Nick and me, thanks so much for listening. We look forward to being back with you next week. Until next time, take care of yourself and take care of each other.

[01:08:23] Ending

Thanks for listening to the Systematic Investor podcast series. If you enjoyed 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.

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.