237 Systematic Investor Series ft. Richard Brennan – April 2nd, 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.

[00:00:22] Niels

Welcome and welcome back to this week’s edition of the Systematic Investor Series with Richard Brennan and I, Niels Kaastrup-Larsen, where each week, we take the pulse of the global market through the lens of a rules-based investor. Now, if you’re new to the show, I hope that today’s episode will trigger your curiosity enough to check out the back catalogue and listen to the past episodes that you may have missed, like my conversation with Cem last week, where we discussed the reasons behind the unprecedented moves in the volatility space and how they’re impacting different strategies including trend following.

Cem also shared his thoughts on the critical importance of positioning and why volatility strategies have been facing some tougher times in the last couple of years. So you don’t want to miss that one for sure. Also, I would encourage you to listen to the midweek episode where we had a thought-provoking conversation with Rana Foroohar, global business columnist and author of the book Homecoming: The Path to Prosperity in a Post-Global World, really a conversation that will no doubt make you think about where we are heading as a planet.

Lastly, I am excited to invite you to enjoy the CTA mini-series where Alan and I have had the privilege of speaking with the decision makers of most of the largest CTAs in the world. We dive deep into the most pressing topics and have gained unprecedented access to these industry leaders who share some exceptional insights to the CTA world. So head over once you’re done listening to Rich and I today and check out some of those episodes.

Rich, it is wonderful to be back with you today. Unexpected, I will say, because of course there was someone else who was meant to be here today, but unfortunately, Andrew couldn’t make it and you very kindly just raised your hand and stepped up to it. So thank you so much for doing that. I hope you’re well. I guess you’re still enjoying some summer weather where you are.

[00:02:16] Richard

Yes, Niels. Still good summer weather over here. It’s becoming more summer weather over here. But yeah, it’s all good on my front, Niels. I’ve had a pretty interesting month, which we’ll be discussing today on the podcast, and I’ve been loving your Top Traders Unplugged series. It’s given us so much fresh food for thought for different research projects that I’m rapidly engaged in now, thanks to all of the contributions by your great guests. So it’s a great time to be a trend follower.

It’s been an interesting month, especially for those with fixed income, but my battleship is doing well. So yeah, very happy.

[00:02:56] Niels

Great to hear, and yes, we have an interesting lineup, of course, inspired by the last month’s event. But that’s great. It certainly helps us try and dive into some of the aspects of trend following that maybe we haven’t seen for 18 months or so. But before we do that, let me just maybe give some thoughts about what’s been going on in the last month or so. It’s only been a few weeks since the last bank failure in the United States and Switzerland and the great banking crisis of 2023 already feels like a distant memory.

Now, did the Fed and the Swiss authority’s actions prevent a sure disaster? Not sure. Or was the crisis as severe as anticipated? Not sure either. Really, it’s impossible to know the answer to that. But their early action appears, for now at least, to have averted any potential meltdown of the banking system. But the question is, was it too early and is the equity and bond bear market still on the cusp?

The recent banking crises have been resolved but question remains about whether it’s too early to breathe a sigh of relief. The bear markets in equity and bonds could still continue, and if they do, the impact on the global economy may reveal more surprises in the banking sector. SVB Financial’s exceptionally high level of uninsured deposit is considered an extreme, but the underlying problem of a mismatch between the banking industry’s assets and liabilities is real.

Should interest rates rise again, the problem only gets worse and SVB is not the only bank that invested in supposedly safe government paper when interest rates were near zero. The crisis in March may signal a new era of illiquidity, and if it unfolds, it could be too big to bail. Now, during the great financial crisis, regulators attempted to bail out financial firms as early as November 2007. But as we know now, the broader US stock market still declined 54% by March 2009. This suggests that timing is crucial when it comes to rescue actions.

This time around, the Fed is attempting to intervene even earlier, with its balance sheet already at almost $9 trillion, significantly larger than during the GFC when it was less than $2 trillion. In the week of March 15 this year, the Fed’s assets increased by more than $150 billion. That’s the largest weekly increase in history. Of course, only time will tell if they’re successful.

Now, let me bring you in here, Rich, just to touch on some of the bigger things before we dive into all the nitty gritty of it that kind of caught your attention. You are also a seasoned market watcher, trend follower, but there was a few interesting moves that we may have forgotten about.

[00:05:51] Richard

Yes, so the story in March, obviously, was mostly about fixed income. As you said just previously, it’s clear that the Fed are taking this sort of immediate action whenever there’s a disturbance in their realm. It’s interesting. I was looking at the performance returns of a lot of the CTAs and the trend followers in particular over the months of March, and there is considerable dispersion in those returns.

So the real question for this particular correction in March relates to, I believe, the level of exposure that a CTA or a program has towards the fixed income sector, because the way I saw it (particularly with my portfolio), was that if you weren’t heavily exposed in fixed income, the corrective moves of the high volatility and fixed income was actually quite restricted to the fixed income sector. It didn’t flow through to the other markets. So in my particular portfolio, I was very fortunate in that I had little exposure.

Now, that wasn’t through design, Niels. It was through the fact that there were very few products in the game I play that are available in the fixed income area. But I would, in general, say I’d probably logically prefer to have a smaller allocation to fixed income than say commodities because I see that there is big diversification potential in the individual commodities as a group and probably a bit less diversification potential in fixed income as a group, because fixed income tends to be solely related to interest rate movements and credit risk.

There’s a limited realm of difference in the fixed income sector, where in the commodity sector, you get the difference between grain production, harvest, weather events, heavy metals, all of these different things which give greater diversification potential. But in my particular portfolio, because I didn’t have this big exposure to fixed income, I found that my portfolio did quite well for March.

Two of the trends that I’ve been riding for quite a few years now, being orange juice and London sugar, just continue to extend their trends. So they’ve been holding up my portfolio very well in the month of March. Really, there was not much more news from my front, from my battleship, in relation to the events of March, but certainly, for those programs that were in fixed income, there’s a story to tell.

[00:08:36] Niels

First of all, let me just give people examples of how unusual (let’s put it that way) the moves were. For example, a one month bill in the US, which at the time (so this is around say March 8th, 9th, so this would have been an April 11th 2023 bill, so about a month), on March 9th, was trading at a yield of 4.58% just before the SVB meltdown. Now, the same bill (again for April 11th this year) traded as low as 3.55 on March 27, and (hold your hat) it’s closing out the month now (I wrote down) at 4.77. I wonder if that is correct or whether I meant to say 3.77. Yes, I think it is 4.77 which is just mind-boggling. But of course, that has something to do with where the Fed’s rates are and of course at maturity, they are going to be very similar. So it is probably 4.77 which makes it even more mind-boggling.

Anyway, then you had the two-year US Treasury notes. On March 8th, they closed at 5.07%. Then they rallied and the yields dropped 123 basis points to March 23rd, down to 3.84%, and that closed out a month at 4.08%. So this is just to show, I guess, that we really talk about some extraordinary moves in that space. Now, then I picked up on something you said, which I think is quite interesting. You said this thing about that you like commodities, and so do I, of course.

But when I think about what’s been going on in our world for the last decade or two, you could kind of argue that some of the biggest outlier moves have actually been in the fixed income space. What I mean by that is, nobody expected yields to become negative and all of those things. So, if you think about the extraordinary moves, and on top of that, then the reverse, nobody expected yields to rise so quickly as they did last year. So in some ways, you could argue that there has been some great outliers in fixed income.

[00:11:11] Richard

Yeah. Well, there has certainly been huge volatility. We’ve never seen this volatility before in the fixed income sector. Something which is a very poignant point in that if you’re developing your rules-based systematic process using an extensive backtest, you’ve got to be careful because there are regimes we’ve never experienced in the backtest which we’ve been presented with today, that clearly, there is no precedent for what’s happened.

That, to me, shows the importance of not just relying on your backtests but doing these what-if scenarios to say, what if this happened, what if this happened, rather than relying on what history has presented. It’s good to apply a logical sensitivity test to some of the assumptions in your backtests so that you don’t slip into that potential trap of finding that you overfit to favorable regimes of the past. For instance, with bonds, we know that we’ve had a stellar performance of bonds in relation to trend following over the last decade or so.

There’s been numerous opportunities in fixed income for trend following opportunities. But if you could imagine, if you are relying on a backtest and you aren’t taking heed of the fact that your backtest might be over-optimizing your portfolio construction to have a higher percentage of fixed income in that portfolio because of its very solid performance, you could potentially find yourself in a trap when those regimes change on a pin like they have done in March, and you find goodness me, I never expected this level of correlated behavior to occur when the bonds snapped like that.

These are all things that you’ve got to take care of in your assumptions about what that backtest tells you, if you know what I mean.

[00:13:12] Niels

Absolutely. Actually, it kind of raises an interesting point, because you talk about how trend followers, without a doubt, have benefited from a very persistent trend in fixed income over the last couple of decades, not a trend that didn’t have its corrections and reversals and counter trends and all that for sure. But it was certainly an area where trend followers, in general, were able to harvest a lot of returns.

It reminded me about a paper written before 2014, for sure, by Roy Niederhoffer about how trend followers would be unable to make money when interest rates started to go up, and he argued that. I’m teasing this because we interviewed Roy last week, it was a fantastic conversation, and of course, because I spoke with him about this in 2014 the first time he was on the podcast.

Of course back then, there was no argument because we had not seen a period of really consistent rising yields so there was no real way of telling whether what he wrote back then was right or not. I’m not going to spoil it. I’m not going to take the thunder out of the episode. All I will say is that watch this space because it was, at the time, something that really I think influenced a lot of investors as to the outlook for longer-term trend followers. Of course, back then none of us knew that interest rates would continue to fall for another six years or seven years, whatever, which also is an important point. But it was just fascinating to have the opportunity, and Roy’s a great sport so he definitely commented in depth about that particular statement in the paper.

Let’s leave that for now and let’s turn to the month of March because it is month end as we are recording. Now, we don’t have the final numbers because they take an extra day to come out. Of course, you know roughly what you did. You follow some of the intra-month numbers. Of course, I know how the month turned out from our side. But if I was just going to put some words to it, of course, influenced by, somewhat, what I experienced inside our firm on a day-by-day basis, but it was an interesting month, because on the 8th of March (I’m sure there were others than us), but actually, we were making new all-time highs on the 8th of March.

Little did we know and little did our models know that something was brewing over in Silicon Valley. Of course, fixed income markets truly experienced some exceptional moves in the following days, which of course, sometimes will catch trend followers and other strategies for that matter on the wrong side of the move. Of course, in our case, in most trend followers’ case, we suddenly ended up giving back what we had made in February and a few extra percent, but not too dramatic as I’m sure we’ll see in a few days.

But also, I want to put the numbers in context before they start to come out and just remind people that even though it was, and most likely will be, a healthy down month for many managers, in the last 12 months, there were actually up months. I can at least count two in the last 12 months that were bigger to the upside than what we saw to the downside in March. So people should also take that into account.

Now, the thing is that when we have these periods of time, it reminds people that there will be give-back periods when it comes to investment strategies like trend following and sometimes they can happen over just a few days, just like you can have some really exceptional upside performance that occur out of the blue. It kind of goes to the point that we talked about so many times on the podcast, and that is, this is why you can’t time trend following. You really just need to have it and stick with it and make it part of your core application.

The other thing that actually is somewhat positive, when we look at a month like March, and that is, it gives us as an industry, as managers, a chance to showcase how different models, how different risk management regimes are designed to respond to these extreme markets. Because as you pointed out, Rich, these moves were really not in any of our datasets.

You could argue a firm like DUNN, because it’s been around since the ‘70s, that the last time we had such a move was I think in 1982, that maybe we had something similar in the database. But I don’t even think that financial futures on bonds were available in ‘82. I think it was just commodity futures back then. I think the fixed income futures came later. So I’m not even sure that we can fully say that even a firm like us have seen this before. So it was quite interesting.

Of course, another thing it highlights (and I don’t know if we dare to go there, Rich), is this thing about how models react to dynamic position sizing versus static position sizing. It is an interesting time, completely, from an objective point of view. No right, no wrong, just to how they’re different. I think when I reflect on what’s what we’ve been through and what I shared with my clients as well, is just that it seems to me (and I don’t know about you), that it feels similar to the three-day counter-trend period we had between November 26th and November 29th 2021 when the Omicron variant was announced to the world.

Back then, it certainly felt incredibly uncomfortable to see performance move that quickly, not least because it came at the back of a softer performance period for the industry over a few years. But it also showed that these strategies are designed to be incredibly resilient, something I think people underestimate and don’t give us as much credit for as I think a lot of these designers and researchers should get.

Also, I think that might be the reason why, as far as I know (I’m sure people will correct me), this is probably the longest-running true alternative or hedge fund strategy that we have where you have continuous track records for about 50 years now. I don’t think there’s many other strategies that can claim that.

Of course, finally, as a little hope out there, and that is what happened back in November 2021, which was unpleasant for those three days, it actually also marked the beginning of one of the strongest performance periods for trend followers on record. So, lots of things to take away from that. I want to hear your thoughts, but let me just give people where we stand in terms of the numbers.

So my own trade barometer finished the month at 25. Definitely very weak confirming that it’s not going to be a good month. We know that. The BTOP 50 Index is only, as of Thursday night, down about 5%, down less than 4% for the year. Not a big deal in my opinion. It was up 15% almost last year. The SG CTA Index down about 6.4% as of Thursday for the month, down about 5.2% for the year. It was up 20% last year. The Trend Index down 7.58% for the month, down 7.16% for the year, but it was up 27% last year. The Short-Term Traders Index down about 2% for the month, down about 2.4% for the year, and it was up 11% last year.

Now, the great thing for those who use trend following and CTAs, within a multi-asset portfolio where they have stocks and bonds, they probably didn’t lose any money this month because MSCI World was up 2.8%, up now 7% For the year and a quarter actually. The World Government Bond Index up 2.55%, and the S&P 500 up 3.51% for the month, up 7% for the year. So if you were having, say, a 20% allocation to CTAs, which by the way we’ll come to later in the show, you probably didn’t feel much pain, if any, in a month like this.

Anyways, Rich, that was a long monologue from me so I’m going to take some tea and you can chime in.

[00:22:18] Richard

No, it was good to put a context on it because I was just listening to the figures you gave for SG CTA Index and the SG Trend Index. So, down around 6.5% or 6.4%, down around 7% for the Trend, that tends to me to say that there’s going to be dispersion in those groups maybe with some that have double-digit losses for the month and some that have small amounts for the month.

So when we get these events, when I talk to people, a lot of them say, well, as medium to long-term trend followers, when we get these short, sharp corrections, there’s not much you can do about it because our models are slow to acclimatize to those short, sharp correction so inevitably, we get these painful shock reversals. If the trend extends, then we get the benefit of Katy Kaminski’s crisis alpha coming with these enduring trends. But during these short, sharp corrections, we can be exposed.

But a lot of them say, well, that’s just part and parcel of the way we trade and you shouldn’t worry about it. Now, I tend to think, well, look, this isn’t the first time this has happened as you’ve said. We got a similar event, not in fixed income, but we got a similar event in vol-mageddon in February 2018. As you rightly said, that Black Friday, I think November 2021, I remember crying to you the day after that event with the impact on those short, sharp reversals for the Omicron variant on my portfolio. I think that was about right, if I can remember correctly. So these things, they occur. We’ve seen these things occur.

Interestingly, when I went back to your Winton episode with Top Traders Unplugged, he referred to a similar event occurring back in 2007 where they took note, Winton took note of that event and took adaptive steps to address that event and they found that when vol-mageddon came along and when the bond shock recently came along in March or whatever, they were much more prepared for that event. This is what I think we need to do.

There is some learning we can do from these events because if for instance we find that we do face double-digit losses in a particular month, when we’re looking at our monthly returns, the idea of a trend follower is because we cut losses short and let profits run. We never want to see adverse months get too extended. We want to see large moves in beneficial months like 10%, 20%, 30% returns for a beneficial month. But we don’t want to see minus 10% or minus 20% for a month. We would far prefer to see a minus 4% or minus 5% for a month.

So if we find that we have these events and we get these double-digit adverse exposures, it tells us information. It tells us where we clearly were overexposed to that volatility shock or that corrective shock. That’s why we got this large impact on our P&L. This inevitably, to me, tells me that there’s usually a way around this when you look at the correlations in your portfolio, or the leverage in your portfolio.

This should really make us say, well, to reduce the impact of these corrective events, let’s look at the leverage of our portfolio. Is it too high; or is it too high, in particular, in relation to this asset class? What is the reason for this large double-digit loss in this month? These are all great questions we should be asking. Because if we are risk managers, which we all say we are in our trend following world, we want to treat these short, sharp corrective shocks seriously and not just say, “Well, that’s too bad. That’s what happens with a trend follower.”

We want to try and learn from them and see if we can adapt our processes in some way to make our portfolios less correlated or less exposed to a particular asset class. There are, I think, certain learnings that we can take from this into how to make our portfolios robust to stand up better to the short-term corrective whipsaws, as well as take full advantage of the major crisis events from extended trends. I don’t think we should just relax on our hammocks and say, well, we couldn’t do anything about that and we’re going to catch up over the long term so c’est la vie. What do you think?

[00:27:18] Niels

I’m going to push back a little bit on that. I’m not pushing back on the idea that we should always learn, and of course, this data will be in the dataset, certainly, so our models will automatically learn from it to some degree. But I do want to push back about this idea that we can remove the unpleasant short-term periods and it does not have an effect on the long term performance. I actually think it does. So it’s finding that balance, because I certainly remember on our side that we were asked after vol-mageddon to say, well, isn’t there something we can do about it.

Let’s just maybe remind people what happened in vol-mageddon. Vol-mageddon was in February of 2018, and it was a period where probably losses were a little bit bigger than what they were in March of this month. It just felt unpleasant. I can only put this in relative terms because we don’t want to talk about specific performance, but I just want to say that on our side, we gave back only about half of what we had made the previous four months. So over a five-month period, we were still up pretty healthily.

You could say, again with last year, yeah, we’re giving back, as I said, a little bit more than what we made in February, but last year was a fantastic year on our side. So should we do something to dampen the opportunities for delivering really strong returns when there are lots of trends? Of course, it always has a bigger impact when it’s fixed income because there just are more fixed income markets that we can trade; there aren’t that many orange juice we can trade other than one, for example. So obviously, fixed income and equities to some extent, for some people currencies, can have a big impact.

But I will say, you can’t just focus on the short-term picture and say, yeah, we should do something to address this period or that period. We do need to look at how are we going to deliver depending on what your objective is, of course, but if your objective is to deliver the best CAGR you can with reasonable volatility, with reasonable drawdowns, etc. If it means that sometimes you’re going to have a double-digit down month (which by the way is only random because it’s a calendar month), it could have been two down months or 6% and we wouldn’t have noticed it, right? It just happens to be one month.

Also, going back to this thing that we like to talk about, Rich, and that is keeping things simple. I think if you try to engineer too much about fixing a vol-mageddon, fixing an SVB collapse, fixing this, fixing that, then you are starting to optimize more, to some extent, or putting in more bells and whistles to do that. That’s inevitable. So you have to weigh that as well.

Now, for those who can do it without giving up anything in the long run, fantastic. I think very few people can do that, but that’s just my personal opinion.

[00:30:32] Richard

No, I do agree with you. But I’m just thinking that it’s always important to look at correlations in our portfolio. Not simply what historical correlations have told us, but more importantly, look at ways to break down the correlation structure of your portfolio. We’ll probably get into that later in one of the topics we might be talking about where we’re dealing with ensembles of systems, but that is one way to naturally break down the level of correlation in the portfolio.

So as a trend follower, Niels, I say to myself, well, if I did get a double-digit… The good thing is, with a trend follower (this is a plus to the trend following community), if we find that a month, we do find we have a double-digit downturn, let’s say we’ve got a 15% downturn, a horrible month for us. We haven’t seen that frequently in our historic record, but it’s possible. What you do find is that with that event occurring, your risk is released from your portfolio because you’re exiting your positions.

That’s a good sign because if that regime of whipsaw continues, going forward, you don’t find that the next month you also have a 15% impact and the next month and also a 15% impact because you’ve released risk from your portfolio from that one event. We don’t find that staying in that particular whipsaw regime makes us necessarily go backwards any further than we did with our initial hitch, if that makes sense. So that’s a really good sign with those trend following portfolios that deliberately release risk from the portfolio.

So then the question I ask is, was the portfolio using stops on their positions? If not, maybe they would find that isn’t totally released, the risk isn’t totally released from their portfolio. We’ve heard a lot of these discussions with the different program managers through your series. Everyone has their different nuances. There’s a lot to consider in these questions because they’re not easy to respond to.

But what I’d like to say is, if I got this major hit, I don’t want it again. Is there ways to maintain the necessary properties that I’ve been getting from my program, but at the same time, address those weaknesses without necessarily having to go to tail risk protection, without necessarily having to go to aggressive short-term models? Are there other ways we can do it? Is there ways through reducing levels of correlation in our portfolio through further diversification?

So the idea being that we become less and less exposed to a single asset class with a broader diversification approach, provided there are uncorrelated return streams. Is there ways to embed some risk mitigation mechanisms within the portfolio itself to ensure that we’ve released all the risk from that corrective phase? There are these steps.
Secondly, when we utilize that data from those corrective periods like vol-mageddon and the events we’ve talked about, when we update our models, refresh our models through an adaptive workflow process, that might take into account some of the necessary tweaks we’ve got to do in our systems to make the impacts less impactful in future events.

So, the thing I’d be calling out for is, don’t just assume that these are a necessary pain of trend following. Consider ways that through an adaptive process we can get better at handling these events. That’s all I’m saying.

[00:34:29] Niels

Yeah. I agree. As you speak, I’m thinking that yeah, it would have been nice to have a bit less exposure to fixed income last month for sure. I agree with the idea of further diversification. I agree with that. But I also think that further diversification, to some extent, will dilute returns, meaning the fact that some managers made so much money last year, well, it was because they had a decent exposure in some of these fixed income markets.

We know of one manager that did exceptionally well, and all of that performance and more came from three short-term interest rate contracts. Talk about concentration, right? But it made a lot of money, and if they managed to get out, well, that’s still a bet you would want to take if you want pure trend. If you don’t want pure trend and you don’t want your 50% plus return in 2022, and you’re happy with something that is much smoother, and you can certainly obtain that through more diversification, I’m sure, then that’s fine as well. There’s nothing wrong with it.

I think what was a little bit unusual, which we haven’t seen for a while but I don’t think it should be expected not to happen again, what was unusual is that we have been talking about for almost 10 years on the podcast about this term crisis alpha, and indirectly, we always talk about it as an equity crisis. Then what shows up is a fixed income crisis.

But this is what I said in the beginning and what I love about what we do and why I think we all have so much passion for this space, and that is, well, even if we haven’t seen this for quite a while, the models were perfectly fine. They’re designed to deal with these unexpected changes. This is what divergent strategies are built to do. They thrive on change.

A convergence strategy, I haven’t seen any numbers yet, but if there were some convergence strategies in fixed income, I don’t really want to speculate as to how serious that could have been.

[00:36:47] Richard

There would be some big negative skew events.

[00:36:50] Niels

Again, that’s the feature of their strategy. Maybe they’ve had a good 10 years of return, but you know, they have a different risk-return profile than we do. So it’s obviously all about mixing these things up. There will be some short-term managers for sure that did really well last month, and kudos to them. They suffered for a decade or more, and now they’re having this moment of sunshine, and that’s great. That’s great.

But then again, you have to ask yourself, could you, as an investor, have endured a 10, 14-year drawdown and then have a couple of really great years in the last two or three years because things are changing? Maybe, maybe not. So it’s all about preference. It’s all about striking that balance, and it is for investors to find those managers that they are most likely to hold on to. I think that’s what it comes down to. Because again, there’s no wrong or right per se here. It’s all about preference. As we’ve talked about many times, the best strategy for people to be invested in, the best investment that they invest in is the one that they can hold on to and sleep soundly at night with.

All right. Do you want to dive into your topics? It is somewhat a continuation of what we talked about just now. I think topic number one, I think we’ve probably dealt with that. But then we move on to topic number two, and I want to set the stage a little bit because we wrote about it in our blog post that we posted a couple of weeks ago, Rich, that people can go and find on the Top Traders Unplugged website and they should read it, hopefully. But it is actually also a question that Alan and I have been talking with all these friends of ours from the industry about.

I guess the way you could put it is to say, okay, we know Sharpe is important, but which Sharpe is the one we want to optimize for? Is it our Sharpe, as a manager, or are we trying to design strategies that will actually improve the Sharpe of the portfolio the most? That, I think, is a very relevant question and I’d love to hear what you found when you crunched the numbers and we can talk about it.

[00:39:20] Richard

Yes. Look, this was an interesting exercise. What we did, we looked at the performance of (as many as we could get) the program managers in the SG CTA Index. We called that the TTU CTA Index, if you remember, Niels. But there was a couple that were excluded from there like ISAM and maybe one other where we just weren’t able to get the data, but for all of the others, we were able to get the data from their inception up until current day.

So interestingly, some of the programs in the CTA index such as PIMCO have got a pretty short track record, three or four years or so, but some of them of course have a 20-year track record plus. But what we found was that when we looked at the performance of each of those programs from their inception to current day and recorded their Sharpe (so we had around, I think in this example, close to 20 programs here which we evaluated their program Sharpe in isolation), then we looked at the Sharpe over a similar date range for the 60/40 portfolio. Then what we did was, we said, well, let’s blend 20% of every individual program into the 60/40 portfolio and let’s look at what the blended Sharpe of an 80% allocation to a 60/40 portfolio and a 20% allocation to the nominated program, let’s see what that would generate in terms of its program Sharpe or portfolio Sharpe, and then let’s see if that portfolio Sharpe has improved or it has increased or decreased from the performance of the 60/40 portfolio.

Now, what was really surprising to me, Niels, was that no matter which program we took, every single one of those programs, those which we can regardless trend following programs after we’ve heard of a lot of them from your TTU Index.

[00:41:29] Niels

I would say with a couple of exceptions. There’s a couple of short-term managers in the list we did.

[00:41:35] Richard

Fair enough. We’ll chuck them in as well. But when we look at it, we see that in every instance, the Sharpe ratio increases. Now, what is important is that it’s a material increase when we look at those with a longer-term track record. So the period from inception to current day, if we’ve got a 20-year period, for a lot of them, we’re looking at 30% increases plus in the Sharpe ratio from a 60/40 portfolio with an inclusion of 20% of that trend following program in their portfolio.

That’s a massive increase; a material increase in Sharpe. What was surprising to me, Niels, was that we didn’t get any negatives here. It was across the board. Every single program showed this improvement in Sharpe ratio.

[00:42:25] Niels

Just to clarify, not everyone improved by that much, right? So we had a lot of people kind of between 20% and 30%. improvement. We had one that didn’t improve, but that’s a shorter-term track record we were using, and we had some that only improved slightly. I think when people go and read the blog post, they might be surprised to see who they were. But anyways, we can’t talk about that specifically here.

[00:42:52] Richard

But no negatives.

[00:42:53] Niels

Exactly. There were no negatives. That’s very important to stress.

[00:42:56] Richard

So that impressed me for a start. Then I thought, well, okay, for those people that have got a slight concern over the use of the Sharpe ratio, even for a portfolio (that’s a discussion we can probably get into a bit further in this conversation), but I want to see if the CAGR improves as well as the Sharpe ratio. So I had a look.

I didn’t record it in that report, but we do find that the CAGR is improving in those programs as well as the Sharpe; with the improvement in Sharpe. But the example that we gave is where we looked at your program, Niels, DUNN Capital’s World Monetary Agricultural program, the WMA program, and we find that there are two things going on. We see that we’ve got a significant material improvement in the Sharpe ratio. But also, when we look at the compound annual growth rate, we also get a significant improvement in the 60/40.

[00:44:08] Niels

Total portfolio does improve. Yeah, that is correct.

[00:44:11 Richard

It’s quite a material improvement. What that is, is what I’m calling the lifting power of trend following. It’s not just that trend following smooths the portfolio program and improves the Sharpe, in most instances, it’s offering lifting power to their portfolio as well, where it’s improving the overall net wealth at the end of that time series.

[00:44:35] Niels

Yeah, and actually because we are restricted in terms of talking about specific performance numbers, but what we can say as you say, yeah, there is an improvement, but there’s actually also quite a large improvement, a very significant improvement in terms of lowering the drawdown compared to a 60/40 portfolio. This is of course why we all love trend following because it is not necessarily magic on its own, but it has a magical effect on a traditional portfolio of stocks and bonds without a doubt.

[00:45:04] Richard

Yeah. This is just an exercise we did looking at a 20% allocation where we just pulled a number out of the air. What happens if we do a 20% allocation? But in our previous podcast, and I remember when I talked to you about what is the optimal allocation of trend following towards a portfolio that mimics the S&P 500, and I came back to you and said 65%.

Well, I mentioned to you how I was listening to that Winton podcast the other day and one of their programs, they’ve got three programs, one that is 100% trend, another that is 75% trend, and then another that I think is much more multi-strat in form. But anyway, they were talking about how, when they talk about a material allocation to trend, that preserves the great properties of trend following in the actual portfolio that you’re allocating towards; you do need to allocate a material allocation. In this case, they were using 75% allocation with equities.

In this example, we’ve been discussing, it’s only a 20% allocation. But you’ll find as we increase the allocation, it just gets better up to a level of about 50%.

[00:46:20] Niels

I’ll give you another way of looking at this because it is of course a question that we get a lot on the podcast, I get a lot in my day job, and it is, of course, how much should I allocate to trend? I always thought, well, it’s kind of a little bit random because it depends on when you start your simulation of an allocation and all those things. So maybe another way of looking at it was to say, but what if you started just before or just after a crisis? Because that would be kind of extreme starting points where the allocation between strategies will become quite different.

So I do this exercise every month that I update where we look at a starting point just before the tech bubble and then just after the tech bubble finished, so the low of the stocks. I think October 2002 is the low. Then I do a start point just before the great financial crisis and just after the great financial crisis. Of course, this is just simple. We’re just optimizing for Sharpe between S&P, World Government Bonds, and then our own trend following strategy.

What you find, which I think is just so lovely, and that is, if you look at before and after the tech bubble, the allocation to our strategy within that portfolio is almost the same whether you started before or after. Now, of course, it does change with time. It is interesting, right? But it does change, of course, a little bit depending on as we continue to add more data to it. Before and after the global financial crisis, there’s a little bit of variance, a little bit more variance between before and after the allocation to DUNN. But it’s even more extreme between stocks and bonds because at the moment, there are no bonds.

If you started after the GFC, so in March of ’09, and you optimized for Sharpe, you should only have stocks and trend following, no bonds. Anyway, what I’m kind of trying to communicate to people, and it goes back to an earlier point we talked about as to why you need to have this as a co-allocation because one, you can’t time it, so you’re going to get caught either up or down if you try and be too clever about it. But actually, the data also suggests that it’s the most constant of the components in a portfolio of stocks, bonds and trend following, at least when I do the numbers in our case. It may not be the same for everyone. But I have a feeling it’s probably not that dissimilar.

So I just think it’s just another little funny way for us to showcase how people can use this and use it very effectively, really, as a portfolio diversifying tool. As you say, it has lifting power. It has risk reduction attributes. What’s not to like, so to speak? Of course, we know there’s a lot of people who don’t like it. But from our point of view, it’s pretty incredible.

[00:49:28] Richard

Yeah. Look, that gives a good segue into the next topic. We’ve had this discussion (well, you’ve had this discussion with Katy Kaminski and a lot of the program managers) where you’ve talked about crisis alpha. You’ve talked about Cliff Asness’ report where he said is there a responsibility we have, a moral obligation to offer crisis alpha in addition to other trend following properties.

But what I’d like to talk about here is I think we need to distinguish ourselves between tail hedging and allocation towards trend following. We do know that tail hedging strategies do offer negative correlation during short, sharp reversals like we saw in this month.

[00:50:24] Niels

Sometimes.

[00:50:26] Richard

Yes, yes, good point. But, in general, tail hedging tends to protect for both short sharp corrections and enduring adverse crisis periods. However, they give no upside benefits outside of those regimes. You’ve actually got to pay the price of those premiums during those beneficial times where you get no benefit from them. So there is an inherent strong cost associated with having tail protection in these strategies, despite the fact they do offer this comprehensive protection when your traditional portfolio is suffering. But trend following is a bit different to that.

Now, we talk about the uncorrelated nature of trend following to other alternative investments or other different traditional portfolios. But whilst we term it being uncorrelated, we find when we look at the detail, we find that during periods of extended adversity, we do very well, and during periods of extended boom time, we do very well. What we find is during this adversity where a traditional portfolio is suffering, as we’ve said before, trend following in these protracted periods of crisis do come to the aid. They provide this protective cushioning for that traditional portfolio and they do help reduce the blow of that adversity.

But importantly, that’s not the only benefit of trend following. I think that when we don’t talk about the benefits during boom times, we lose one of the big, great properties of trend following because we can find, for instance, during periods of equity boom time such as during some of the periods during 2010 up to 2020, there were some stellar performance times for trend following that were positively correlated with equities. So whilst they might be uncorrelated over the long term, there are periods where we get great negative correlation and there are periods where we get great positive correlation. There isn’t this consistent level of uncorrelated nature.

The reason for that is that when we look at the distribution of returns, we see that trend following tends to work well in those zones of the market regime that are trending towards the plough regions. So that’s why during periods of boom or periods of bust, we’re heading towards the left and right tails of the market distribution and we find that the beneficial properties of trend following kick in when we start getting these tail regions evolving. But it’s during the much more sedate market where the predictable regimes that are much more noisy, where there are no clear defined trends, that’s where trend following doesn’t do so well.

But it is important to note that the tail properties of both the right and left tail is something we want at every traditional portfolio because that’s the weakness of a traditional portfolio in that it doesn’t basically incorporate allocations of strategies that offer any degree of protection in those tail regions, either boom or bust. A portfolio that is heavily dominated in bonds, during the equity boom time, might find that they go nowhere. Equity portfolios during periods of bond boom might not go anywhere.

But the thing about trend following, it’s a fascinating strategy, and there’s very few like it that offer these protective properties as the distribution migrates out towards the tail region. To me, I far prefer it to tail risk protection because you get the upside benefit and the downside benefit with trend following. You just get the downside benefit with tail risk protection, and there is a large cost of insurance to protect yourself during those other times when your portfolio’s doing well.

[00:54:37] Niels

Yeah, I completely agree. As I mentioned very early in our episode today, last week with Cem, who of course, is a volatility expert and one of the best, we did talk about some of these challenges. Of course, at DUNN, we also have a volatility strategy. Although it’s not a long vol strategy per se, it is an absolute return strategy. We are trying to make money not just when things go haywire. It is a challenge without a doubt.

So I think your points were very well articulated. Really nothing I can add to that. Do you want to go to point four? Have we already talked about point four?

[00:55:20] Richard

Well, no, this is a bit of a challenge Niels because now what I’m saying is we’ve talked about how Sharpe shouldn’t be looked at from the individual program perspective, but how perhaps we should be considering the benefits of Sharpe at the portfolio level. We’re talking about portfolio Sharpe. Look, I’ve still got my concerns there. It is a far better measure, portfolio Sharpe than program Sharpe. But the reason I’ve got my concerns is this.

As you know, Niels, I like to classify myself as an outlier hunter. In targeting these outliers, I do find that my equity curve can have these significant beneficial step-ups in performance. What I mean by step-ups is we get an equity curve that slowly rising and then suddenly, with the impact of a beneficial outlier, we get this massive lift in the equity curve. This occurs during periods where outliers are prevalent, where our systems are able to capture them.

So what I did was I looked at a theoretical portfolio where I said, all right, let’s assume, over a 20-year horizon, I achieve a 1% return each month. In other words, no losing months, just a 1% return each month. Now, the reason I did that was that I wanted to get a linear progression and then say, when I compound that series, which is going to be a fantastic linear line, when I compound it, I’m going to get the best of all Sharpes from that series.

Lo and behold, when I do compound that series at 1% each month, and I compound it, I get a Sharpe of 20, which is massive. You never get that in reality. That’s just a theoretical Sharpe with a straight line when we compound it. I thought to myself, great, okay, so let’s try and demonstrate what is the impact of step-ups assuming I got that exact same result. Let’s say that I insert within that series of 1% each month, I insert certain outliers in that series where I suddenly go from a string of 1% each month to suddenly a 5% for a particular month, then back to the 1% per month, then we get a 10% for a particular single month, then I go back to that.

So you can see that there is still no adverse volatility in that scenario but I’ve just inserted outliers into that result and now I’ve compounded that result. So when I compare and contrast the compounded CAGR of the linear series that I’ve compounded versus a series with step-ups included in it, I see that the compound annual growth rate of my step-up series is significantly greater than my linear series. So the compound annual growth rate of my series with step-ups is 17.4%. Yet my compound annual growth rate over the 20 years of my linear series with compound is 12.6%.

However, now when I look at the Sharpe, I see that my linear series, which had a Sharpe of 20, has now reduced to a Sharpe of 1.65. Now, the reason for that dramatic reduction in Sharpe is because of this beneficial volatility offered by the step-ups. So that makes me concerned when I’m dealing with evaluating portfolios and their performance returns, when I’m comparing and contrasting two fairly volatile programs but one with significant step-ups and another that is a much more smoother volatility profile with less beneficial returns because the Sharpe is naturally going to favor the program with the smoother ascending equity curve rather than the one with the step-ups through the inclusion of trend following allocation into that program.

So that naturally therefore, from a decision-making perspective, will shift you away from a decision to allocate towards trend following these fairly highly volatile programs to choosing an alternative strategy that is much smoother simply to achieve this smooth Sharpe. So I don’t know whether trend following is doing itself any favors by using the term even portfolio Sharpe. What do you think?

[00:59:59] Niels

Well, as you know, Rich, it’s not always what we think. It’s more what people who look at us think that is the important part. I just believe that most people, most investors who look to allocate to different types of strategies have this very strong love for Sharpe. So even though I completely agree with you, and this has always been the problem for trend followers is that our positive volatility hurts our Sharpe, I don’t think we can get around it.

Now, that shouldn’t stop us from trying to come up with some better ways of showcasing the benefits. You and I are trying to do that with say this study of how much does it improve the Sharpe. Interestingly enough, actually, one of our listeners sent me a paper a couple of days ago, and I think it actually relates a little bit to this discussion. I may have to dig it out. Maybe you and I can talk about it in our next conversation. So yeah, it’s a challenge, like so many other challenges we have in this industry in terms of what’s the best way to evaluate the strategy, what’s the best way to present it.

I think we just need to continue to educate and make these observations available to everyone looking at it so that they can go and do their homework and say, oh, yeah, I’m not too concerned about trend following only having, and I say only having a Sharpe of 0.4 or 0.5, which is not a bad number. Because as we know, probably Long-Term Capital the day before they blew up, had a very, very high Sharpe. So as long as people remember that Sharpe is not a guarantee for anything.

I remember many conversations with Jerry when we talk about it where he makes a very valid point. It doesn’t talk about real downside risks, the risk of losing all your money or losing a substantial part of it, and that’s absolutely true. So, yeah, these things have to be taken with a grain of salt. But there are certain battles we should take, I think. I’m not so sure we should really take the battle about trying to get people to get away from the Sharpe. I just think it’s going to be a losing proposition. But I completely share your concerns, of course, in that sense.

All right. We still have a couple of points back, but I don’t know if we can really deal with all of them since we’ve already even though we had very little time to prepare, we end up as always having lots of fun talking about this topic. So I’m going to let you choose what you want to talk about of the last two or three topics that you mentioned and then we will slowly wrap things up for today. So where do you want to end up?

[01:03:08] Richard

If we’ve got to choose, given the fact that you’ve been asking the question in your Top Traders Unplugged series with Alan about the views of the program managers towards replication, I think a good one would be to look at the performance of DBMF. So I’ve managed to pull a chart together here so that you and I can get an understanding of how they’re performing lately. Because see, there has been a question that perhaps there’s been a decoupling of DBMF with the SG CTA Index.

But look, what I’m observing from the performance, since DBMF’s Inception, is that there is no- I can’t see a decoupling going on. I can still see Andrew’s program still performing over the period since inception at a slight premium to the SG CTA Index, so I can still see. But I do note that there are certain times where we do find that they lose their correlated properties. Back in February 2019 and up in around about June 2021, we find a sort of underperformance of DBMF which brings it back down closer to the CTA index.

We’ve also noticed, in the last few months, a deterioration of performance of DBMF which has brought it closer back to the SG CTA Index performance, but it is still outperforming. What I do note though is that DBMF is more volatile than SG CTA Index. So I’d be interested in discussing that with you about we talk about replication and some of the properties that replication brings. I noticed that an argument for replication is that it reduces program manager risk by being a replicator across an entire index.

But then I do note there were some great comments brought up in some of your episodes where they said, well, no, there’s still replication risk outside of the program manager risk. So really, we haven’t reduced the program manager risk through replication. But that’s an area I think we can quickly talk about, and the higher volatility I think we can talk about. I do like Andrew a lot. I love his fund. What I really like about it is it democratizes the ability of a retail investor to get exposure to some of these big CTAs that otherwise they’d never be able to get involved in. So that’s what I do love about Andrews ETF.

But there are certain things I don’t like about replication, one of them being I do see it as a more volatile offering than the programs that they’re replicating by virtue of the lags and the processes used in the system for replication. I think that leaves them open to a higher degree of volatility than the underlying programs. But what do you think, Niels?

[01:06:27] Niels

I think Andrew is completely fine with you bringing up these points, because of course, one of his selling arguments that he doesn’t hold back in sharing is of course towards us as managers that oh, yeah, there’s a manager risk. I think it’s perfectly fine for us to push back and say, well, sure, but there’s also replicator risk. I’ve said that also on my in many episodes, and so I think that I completely agree with that. Then you can say, and again, we’re going to try and avoid talking about specific performance numbers, but you’re right.

If we look at this relatively short period, which of course, since inception of his program, which is influenced a lot by this wonderful bond period we’ve had both through the upside and then to the downside, which is where most of their exposure tends to be because of the number of markets they trade. So is this a fair representative of the long run? Probably not. I think we can still see some new situations as we get more data from this.

You’re also correct that over the long period of time, there’s still a little bit of an outperformance. Of course, that’s natural, because one is after fees and the other one is with smaller fees, about 95 basis points, I think it is, once you include the dividend fee as well which is not always mentioned but I think it’s there. But then you have a 70% larger drawdown. Now, that’s a lot.

So nothing is for free. Even if you can save something on the fees, well, in this case, based on the data you presented in the notes prepared, the replicator has a drawdown that is 70% bigger than the index. Now you have to take that into account. You have to compensate for that. If you are willing to get 70% bigger drawdowns, well, then you should get longer or you should get better returns to the upside, I would say.

Otherwise, there’s definitely no reason for having it except as you rightly point out, which I’ve made as well as a point, I think these products are wonderful for those who may or may not be able to get into the single manager strategies which have much higher minimum. So I think those are the points I agree with. I’d love to hear Andrews thoughts when he’s back maybe in a couple of weeks. We were able to set it up instead since you kindly stepped up today for him.

But yeah, we should definitely just ask how he sees that, but more importantly also, how some of the investors that he speaks to think about that, because for sure, we have to deal with this question about individual manager risk, and we deal with that. But one thing also replicators have to deal with some of these question has to be part of the conversation. Now, it is interesting when you see the chart that you prepared when you just compare them that there are these periods where the replicator tends to just pull back to the index.

So you could say, well, then maybe the benefit of replication actually doesn’t show up all the time. It’s very dependent upon when you look at comparing the two because if it tends to pull back from time to time, you could essentially be at a point where you have to redeem and then you’re back to where you came and then maybe, compared to picking one or two individual managers, you wouldn’t see a lot of benefit in performance. But as I said, I think these products are probably for different types of investors. Although I know that there is also a push to get institutional investors involved in these strategies.

But I’ve always had my concern. This has nothing to do with this particular product. I do have a massive concern about any product. That will also be some of the managers we speak to in our series. When you have products that have no incentive to improve because it’s a management fee only product, I personally wouldn’t want to invest in that because I think managers have to have something.

There needs to be an alignment between client and manager and that alignment can’t happen if managers are only paid for managing more money not for delivering great returns and improving those returns over time. So that is a general concern that I have, not a specific one to any particular firm. So those are my initial thoughts, Rich. If you want to counter any of that before we wrap up, feel free.

[01:11:19] Richard

No, I think that covers it, Niels. I think we’re on a similar wavelength here in relation to that there are certain good parts about replicators and there are certain concerns we have about replicators. But it’s interesting, and I do wish Andrew’s firm well going forward. What I like is that I’m not seeing a clear decoupling yet between the replicator and the index.

I think that of course, during these nasty periods of correction, all of us face difficulty during those periods of nasty correction. If Andrew finds he’s in a regime that’s very beneficial for the CTAs going forward for the CTA index, I think he’ll do very well. Clearly, his low-cost structure does give him a significant advantage there. So yeah, I wish him well.

[01:12:16] Niels

Absolutely. All right, Rich, this was very kind of you to step up on very short notice and I hope that people will appreciate that we were able to get an episode out this weekend. We try not to miss any weekends, but of course, it can happen. We had a challenge with personnel so to speak, let’s call it that. We had a challenge on the technology that didn’t quite work initially either. Someone even impersonated me, which you very kindly alerted me to.

Someone has now impersonated my Twitter account by calling it toptraderslive_ with an underscore at the end, which is of course, incredibly naughty since he or she is using all of our branding. So not very happy about that. So if you want, feel free to go and report that to Twitter because we definitely only want one of those accounts I’m sure.

Anyways, enough of that. If you do love these conversations and if you do appreciate the time that we’re putting into this, then please go to toptradersunplugged.com/review if you don’t know how to, go and leave a rating and review on your favorite podcast platform. We so appreciate that and they really do help us get noticed by more people. Also, next week, Rob is joining me. He’s back. He always loves having questions from the industry, from the community.

You can of course send them to me at info@toptradersunplugged.com. I’ll make sure that we get as many as we can answered by Rob. I think that’s pretty much it for that. You should of course follow Rich, follow the right Top Traders Live account on Twitter. But of course, more importantly, maybe make sure you follow us on all the podcast platforms as well as YouTube nowadays, I guess you should say. That’s it. Have a wonderful weekend.

From Rich and me, thanks ever so much for listening, and we look forward to being back with you next week. Until that time, take care of yourself and take care of each other.

[01:14:34] 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.