Why Most Trend Following Improvements Should Fail ft. Rob Carver

[00:00:01] Intro

Welcome to Top Traders Unplugged. In markets success doesn’t come from predicting what happens next, it comes from being prepared for what you can’t predict.

In each episode we go deep with some of the world’s most thoughtful minds in investing, economics, and beyond to understand how they think, how they prepare, and how they decide, and the experiences that shaped how they see the world. No noise, no short-cuts, just real conversations to help you think better and invest with confidence.

[00:00:35] Niels

Welcome and welcome back to this week’s edition of the Systematic Investor series with Rob Carver and I, Niels Kaastrup-Larsen, where each week we take the polls of the global market through the lens of a rules-based investor. Rob, it is wonderful to have you back this week. Hope you’re doing well. How are things in the UK? Is summer finally there?

[00:00:52] Rob

Summer is finally here. The sun is shining. England won their first football game last night. I’m not an especially big follower of the football or a huge football fan, but it’s the World Cup so I guess I should at least pretend to take an interest. So, yeah, there’s a general mood of optimism in the country at the moment.

[00:01:11] Niels

You know, having lived in the UK for many years, I always remember every time there was like some kind of European Cup or World Cup, it’s always this strong belief that this time it’s coming home. Right? We’re still waiting 60 years on.

[00:01:27] Rob

You know, I do statistical analysis for a living, and I’d say the odds are not great, to be honest.

[00:01:33] Niels

We’ll see. I have a feeling we might talk a little bit more about this in a second. But anyways, I just wanted to say we’ve got a wonderful lineup of topics today. Probably more than we expected because I didn’t see all of the ones you had sent me. We also have a question that we’ll be tackling today. So, yeah, strap in and you know, hopefully it’ll be a fun and educational one.

Anyways, before we get into all of that, as usual, I will ask you what’s been, I wouldn’t say keeping you busy necessarily, but what’s kind of come across your desk that you found interesting?

[00:02:10] Rob

I mean, there’s been a lot going on the last week, couple of weeks, obviously. We hopefully have some kind of peace deal in the Middle East which the terms and conditions of which are being debated as to whether it’s a good deal on both sides or very weighted towards one side, whether it’s better the deal that they had before. But we’ll leave that to one side.

So, that does mean, I guess, that hopefully there’ll be some crude flowing out the Strait of Hormuz, and that will then have an impact on markets. Although it does feel like markets have been pricing that optimistic scenario for a while yet. So, we’ll see.

I guess the other thing in markets that’s kind of interesting has been SpaceX. And, in particular, there’s been some debate about whether SpaceX should be allowed into the index club because these indices have quite strict rules. And we’ll touch on that, actually, later in the talk today, because I think we’re going to have a talk about indexing in the CTA space.

But in the olden days, you couldn’t get into an index unless you were profitable and they made adjustments for things like the free float of your shares. So, for example, Tesla wasn’t in the S&P 500 for quite a while, even though it was a big enough company because it just wasn’t profitable.

But in their desperate urge to get SpaceX into their indices, most index companies, apart from S&P, in fact, who have held out, have torn up the rules and they’re like, open the door and please come in. And yeah, I’m not sure it’s an entirely responsible move. I’m not a fan of SpaceX, not a fan of the man that runs it, and we shall see how that works out.

And then I guess, finally, the other thing we must mention, I suppose, is that the change at the top of the Fed has finally happened and there is now a new chairman. But with the current inflation environment, there doesn’t seem to be much appetite for cutting rates. So, maybe we won’t see the true colors of that gentleman just yet. So, yeah, been busy.

[00:04:14] Niels

Actually, all of those points could have been on my list, and they probably were until I realized they were on yours. That’s fine, but just maybe to add a few comments to it. So, yeah, actually, you know, let’s start in reverse. Kevin Warsh is now, you know, had his first meeting as FOMC chair. And you’re absolutely right, kind of interesting that the tone was a lot more hawkish than what he has been, you know, saying before getting the job, so to speak.

No surprise in terms of some of the things he had mentioned in the press conference about shrinking the balance sheet, and so on, and so forth. But maybe something I was thinking of, he does mention that he wants to kind of revamp the way the Fed communicates. And for example, forward guidance is not a business that the Fed should be in, according to him. But then there was something else and that was this thing about reworking kind of the data the central bank releases, as far as I can tell, and maybe getting rid of some of the old fashioned survey methods.

And I was kind of thinking, when I saw that, I thought kind of interesting because we do have strategies, in our world, that have come more recently, that are economic trend or we have systematic global macro. Now, they do rely on this kind of data, as far as I can tell.

And so, I thought it could be interesting to see if they are affected in some ways if suddenly you don’t have access to this kind of information. Of course, for the pure price-based trend followers we don’t care. But I don’t know, any thoughts?

[00:06:13] Rob

Yeah, I mean I find it quite disappointing to be honest because there’s already been quite a serious degradation of the quality of statistical data in the US partly as a result of the DOGE cuts frankly. I know we’re trying to avoid politics but, you know, that’s one of the departments, one of the effects of those kind of fairly indiscriminate cuts to government spending, also a sort of culture of secrecy and sort of lack of transparency, which I find really worrying.

But also, what really didn’t help was the government shutdown last year because that resulted in a lot of statistics being late, or delayed, or not collected at all for quite some time. So, as with my sort of economist hat on, it really worries me. I mean there’s been a debate in the UK as well about the quality of government statistics. You know, if you’re trying to run the country, and you’re looking at your economic indicators and they suddenly go blank, well that’s like a pilot flying with, you know, when his instruments disappear.

And for us it’s kind of a problem, yeah, if you’re using those indicators sort of as inputs into any kind of process to forecast prices, I mean, there are third party services. In fact, this was one run by a good friend of mine. I won’t mention the company because I don’t want to advertise but Syed, if you’re listening, thumbs up to you.

So, there are third party services that collect data and kind of forecast what the government statistics will be if, you like. And that will work if that raw data is still out there. And in this area of big data and AI maybe, arguably, the need for that sort of old-fashioned statistic is gone. But I’m of the general belief that more data is always better. And deliberately getting rid of collecting data strikes me as potentially dangerous, both for the economy and also quite irritating, sure, if you’re trading on it.

[00:08:03] Niels

Now you did mention the football, so let me just add a little comment to that, Rob. First of all, full disclosure, I did not watch England play Croatia last night. But you know what, to me, was the most surprising thing about the game? And I have a feeling you may have watched it… You’re completely quiet.

[00:08:25] Rob

People watching on YouTube will know this is my poker face.

[00:08:28] Niels

Okay? So, I wasn’t surprised by the result. I was surprised that there was not a single yellow card given out during the game because usually those teams are, you know, they go at it. So, anyways, the most fun comment I saw, not about that game, but about the first round of games, was the Turkish commentator who has been suspended because he mistook the two teams he was commenting on and that was Iran and New Zealand. He switched the teams, and he switched the players, for four whole minutes, basically referring to them as the opposite team and opposite players.

I thought that was kind of funny. But anyways, probably not.

[00:09:13] Rob

We’ve all made mistakes.

[00:09:16] Niels

We’ve all made mistakes, absolutely.

[00:09:17] Rob

I mean, I sometimes I get my kids names mixed up so, you know.

[00:09:20] Niels

Okay, I wouldn’t want to make that mistake. Okay. All right. Okay.

So, let’s move on to a quick trend following update here. My trend barometer finished at 36 yesterday, so it’s weak for sure. And I think that kind of ties in with what we see in the numbers so far in June. If I, just from my vantage point, see where there might be a little bit of headwind at the moment. You know, clearly some of the equity markets have been somewhat weaker. We’ve also seen some of the precious metals or the metals markets, I should say, I’m not entirely sure. It could be aluminum that I’m thinking of here that’s been a little bit weaker. And also some of the grains have been.

What seems to work right now is perhaps some of the short rates, in particular after the hawkish stance from the Fed because I think CTA generally have a short stance in terms of exposure and then actually some of the currencies could have worked out well. And I see the dollar has gotten a little bit of new wind under its wings more recently.

Any observations from your point of view in terms of markets from a trend following viewpoint?

[00:10:37] Rob

Yeah, I mean, I’m just having a quick look at my sort of performance, and risk, and numbers, and the last few weeks have been a little bit positive, which is nice. I’m not quite up to the levels I was at earlier in the year. And in terms of the sort of risk I’m running, yeah, quite low.

So, 9% analyzed risk which kind of equates to your trend barometer numbers, I suppose. The biggest risk I have, actually, is in equities. I’ve got a long equity position and that, I guess is pro-deal, if you like. Although, actually, I’m actually flat in energies, and oil, and gas. So, I’m not really positioned for anything going on there, I would say. So yeah, it does feel very much like the world is still waiting.

You know, both if we just look at the quantitative numbers but also kind of just assess the feeling in the markets, you know, this deal’s on the table but it’s been priced in probably anyway. So, maybe this is one of those classic cases of you know, buy the rumor, sell the fact. And then now, you know, now it’s actually happened people will say, well, okay, is this really as good news as we hope it is, or is it still going to be a long time before things normalize, and there’ll be supply shortages of oil, and potential kind of knock on economic effects to that.

[00:11:57] Niels

Yeah, very true. I mean, looking at kind of the portfolio of markets that tend to resemble, I think, most of the liquid markets that CTAs trade, I think the big market move, so far in June, has really been in the oil complex where we’ve seen kind of 15% drops in price. But you know, even some of the precious metals market – silver was down 9% so far this month, platinum also down about 9%.

But, you know, to offset that, as you pointed out, equities 7% up so far in June for the Nikkei, at least on the futures contracts. So not a bad start to June if you’re long that market.

[00:12:37] Rob

Yeah, that’s actually my biggest long. So yeah.

[00:12:40] Niels

Yeah, I think performance wise, I’ll run through it now as usual. I think yesterday was a mildly positive day for CTAs. But you know, again, probably not much of a day so to speak.

But anyways, the BTOP 50 index, as of Tuesday, is down 85 basis points for the month, up 8.76% so far this year. SocGen CTA index down 81 basis points for the month, up 9.5%ish for the year. SocGen Trend down a bit more 1.25%, but up 9% for the year. And the Short-Term Traders Index down only 0.25%, and still up 5% so far this year.

In the traditional world, MSCI World is down about 1% for the month as of last night, still up 9% and change for the year. The US Aggregate Bond index is pretty flat in June, down slightly, and up very modestly, I should say, 0.50% so far this year. And the S&P 500, despite all the excitement around SpaceX, it didn’t lift the index (of course, it’s not part of the index), it’s down 2% so far this month, and up 9% so far this year.

Now I did mention, I think, that we have one question that came in from Garner to you, it came in a couple of days ago. And Garner writes, when deciding whether to add size or remove a strategy, how do you balance Sharpe against path risk like drawdown depth, drawdown duration, loss clustering, and whether losses happen when the rest of the portfolio is also losing? Love the show. I look forward to every episode and listen as soon as they drop.

Now obviously that line had to be included and we’re very grateful for that. Garner, thanks so much. But over to you, Rob, on the question.

[00:14:39] Rob

So, this is essentially a question about portfolio optimization. So, one thing I should say up front is that I rarely drop a strategy, as I’d only do so if there was very strong statistical evidence that it stopped working. And actually, I’m currently in the midst of writing a whole bunch of research about optimization. And the blog post I was literally working on this morning, before starting this podcast, is actually about finding structural breaks, which is a fancy way of saying working out systematically if X or Y is no longer working. So, to get into the sort of meat of the question, the answer is, to be honest, I tend not to look at much beyond Sharpe.

So, let’s take drawdowns, for example, drawdown depth and draw duration. And, actually, we’re going to be talking later in the pod about a nice paper from Man Group about drawdowns, but I’m not a big fan of drawdowns as a sort of statistic for evaluating performance.

You know, generally speaking, if you sort of ignore autocorrelation effects, then your drawdown is going to scale with the performance of your strategy and how long you’ve been running it for. So, you know, the more history you have, the bigger drawdown will be, the better a strategy is. The higher its Sharpe ratio, the smaller the drawdown will be. So, I don’t think including the drawdown really has any meaningful information.

And you know, I’m a big fan of keeping things simple and looking at things like path risk and loss clustering and stuff like that. Maybe that’s interesting and useful at sort of the whole portfolio level. But to sort of get into that on every individual component of your strategy, for me, feels like really overcomplicating things and sort of, you know, potentially even going down the route of overfitting as well.

So, the answer is no, I don’t really look at that in the sense of using it to decide how much risk I should have allocated to a given sort of component of my overall strategy.

[00:16:38] Niels

Is there anything, like one thing, like kind of your favorite thing you look at when you, if you want to say, add or remove part of your strategy? I’m referring to different strategies. You may have built it as one and you don’t really look at it as separate components. But let’s just say, in many instances, actually managers probably have a few different types of trend following perhaps or types of strategies inside the overall program. And if you were in their shoes, and you had to kind of pick one go-to thing to look at, to evaluate, would that be anything that springs to mind?

[00:17:18] Rob

I mean, probably skew, to be honest. Because, obviously, I think that the issue becomes more when you’re combining trend following with other things like carry, because, you know, virtually every CTO out there is not a pure trend follower. Yeah, they’ll have, you know, a 20% or a 30% allocation to what you might call divergent strategies… sorry, convergence strategies which aren’t trend following. My own allocation to those is 40%, which is higher than CTAs because I don’t have an end client pool to worry about. I don’t have to fit into an allocator’s box. I can do what I think will give me the best performance.

So, you know, it could be quite useful to use skew as a proxy for deciding which of those two buckets something could sit in. Because actually it’s not always obvious. So, you know, there are very obvious sort of negative skewed strategies. Things like, for example, if you had a systematic short vol strategy or any vol trading strategy which has a bias towards being short vol, that’s going to have negative skew. So, that’s fairly clear and obvious. But, you know, does carry have negative skew? Well, actually it depends on the asset class.

So, carry in emerging markets, FX, definitely has negative skew. Carry in developed government bonds has actually mildly positive skew. So, it’s not necessarily obvious which of those two buckets things should sit inside of. So, that’s the only one I’d sort of bring in. But even then, I wouldn’t use that formally. I wouldn’t have what economists would call a utility function that combines, say, the skew and the Sharpe ratio to come up with a single metric as to how much a strategy should have.

But I might use a classification mechanism to say, well, you know, I want to limit, say, 30% of my portfolio to negative skew strategies and no more. Therefore, I’m going to just categorize things and then do my allocation within that side. I’d use it in that way. You know, again you get into issues around statistical robustness around skew estimates, which is not a conversation to have today. But it does mean I’m wary of using skew, in of itself, in sort of a very constitutive way, as a mechanism. In the same way, I’ve already said I’m wary of using drawdown for the same reason.

[00:19:30] Niels

Fair enough. Good stuff. All right, let’s move on. So, the first thing we’re going to do is we’re going to kind of talk briefly about two papers that you sent to me that I completely missed and therefore I don’t know what’s in them, but I do know where they came from. And the first one came from an article in Alpha Architect, it’s called the Factor Zoo Has Hundreds of Animals But Only a Handful of Species.

So, over to you.

[00:19:56] Rob

Yeah, so, for those who don’t know this (and anyone who’s listened to me talk for more than about 10 seconds will know it because it’s one of the first things I usually talk about), a factor, you can call it a risk factor or a return factor, essentially is a sort of source of return. So, you know, classically, in the equities world, you had so-called value factors and they were based on simple things like price book ratios, and dividend yields, and price earnings ratios. And then the factor universe expanded, and they brought in things like momentum (which is different from trend following because it’s a relative factor, not an absolute factor), and quality, and so on, and so forth. And then, you know, people like AQR started talking about macro factors.

And then we, in our own universe, we could think of, say, trend following and carry as factors themselves, as well. So essentially there’s all these explanations for where returns come from. And if you’re an academic, you view these as risk factors and you say well, these explain why you’re earning so much money because you’re exposed to all these risks. And if you’re a practitioner, you just say, well, just give me the money, then I’m happy to be, you know… And they call this sort of earning a risk premia, if you like.

So, there are lots and lots of these things and some of them are very correlated. And the exercise that they’ve done in this nice little paper that’s been linked to, in the Alpha Architect website, is essentially as a classifying group. And sort of say, well, actually, if you look at this vast universe of factors, there’s actually only a few that really explain what’s going on. And you know, it’s an interesting piece. You could argue it’s perhaps only of purely theoretical value.

But actually, you know, one of the conversations we have a lot is about diversification and the more that these sort of… the fewer kind of real things that are driving the world, the less help diversification is going to be. So, it is a useful and interesting thing for someone to think about, even though you may think it’s only a kind of, you know, purely an intellectual debate in one sense.

[00:22:08] Niels

Sure, excellent, good stuff. And the other article, or post I should say, that you send over is from a Substack, Beyond Passive Investing. It’s with trend following. Trend following actually adds to a risk premia or what trend following actually adds to a risk premia call. I’m not entirely sure if there’s a name of who wrote this…

[00:22:36] Rob

That’s a very good question. I’m just clicking through.

[00:22:40] Niels

It’s fine. I think we referenced…

[00:22:41] Rob

All I can tell you is it’s someone who’s got a physics PhD, but it doesn’t look like they’re sort of outing themselves if you like. So, I would imagine it’s someone who works in the quant industry and doesn’t want everyone to know.

[00:22:54] Niels

That’s fine.

[00:22:55] Rob

Which is fair enough. It’s a very nice Substack and it’s a very nice series of posts, actually, they’re doing about trend following specifically. This is actually the third in a series of four posts that they’re doing. So, it’s well worth looking at the whole series. I’m just picking on this one, because it came out quite recently.

So, one thing, the sort of source of this post, really, is this fact: if you say trend follow equities, you will be, on average, net long equities because equities tend to go up. So, that means that (again, putting my economist hat on) you can kind of say that, well, actually a big source of your returns is your exposure to equity, the equity risk premia (getting back to his idea of factors and premia).

And then you’ll be earning some additional return above and beyond that due to the fact that you’re trend following. Because you’re trend following sometimes you’ll actually be short the market when it’s, you know, going down, hopefully, fingers crossed. And therefore, you’re earning some extra return during those periods when the market isn’t just going up and you’re just being long the market. So that means that (we talked about this before on the podcast) when you think about how you’re designing your trend following portfolio, you need to ask yourself what your actual end goal is.

So, if all of your money is in a single portfolio of trend following, and you’ve got no money doing anything else anywhere else, then absolutely this is a good thing. Because if you believe there’s an equity risk premia (and most people seem to believe there is) then absolutely, it’s good to have a portfolio that’s got some exposure to that, and you can earn that as a source of return.

However, if your trend following portfolio is, say, 20% of a bigger portfolio, and that bigger portfolio is, say, a 60/40 mixture of equities and bonds, then you’re already earning the equity risk premia in spades elsewhere in your portfolio. So, you could say, well, maybe what I should do is sort of change, modify, tweak, adjust my trend following portfolio so that it has more of a kind of a neutral character.

So, what that would mean in practice is, rather than being long the market when it’s going up, you’d probably be only long the market when it’s going up a lot. And that would bring down your average exposure to equities, your average beta. And the same would applies in bonds, obviously, and any other asset classes you’re trading to a lower level or even to zero. And you could argue that would give you a better overall portfolio when you combine it with your 60/40 than if you’re doing this individually.

So that’s the topic and we’ve talked about it before, and I just thought it was worth mentioning this paper because it kind of… I’m certainly not going to go into it in detail. It’s quite a complex paper and quite detailed, but it does give another interesting twist on that particular idea. So, it really caught my attention because that’s something we talked about in the podcast before in length.

[00:25:58] Niels

As I mentioned, I haven’t read the paper so I should be very cautious in saying something, although I will just say, I’m not a big fan of these people who say, oh, you need to adjust your trend following strategy because it’s going to fit with something else you have in your portfolio. I think if you feel you have too much equity, you sell the equity. You let the trend following program run because, at least for many people, the interaction with other markets in the portfolio will be affected if you start reducing certain sectors artificially. So, that’s just my gut feel.

But… And before you jump in, I can see you’re almost jumping out of your chair here. But the other thing, and this is just from quickly, quickly looking at the conclusion where it says, ‘what trends add, precisely’, and the one thing I can see is that it says that it adds the markets you don’t trade yourself. And I think that’s a great point, actually, because we know that there are so many wonderful markets that you can trade but most people find them difficult to access or risky. So, actually, I think that’s probably one of the key contributions in addition, of course, to also being usually unbiased whether to be long or short.

And again, if there is a prolonged crisis, why wouldn’t you want to be fully short equities, for example, whatever. Anyways, yeah, you can’t contain yourself anymore.

[00:27:32] Rob

Yeah, I guess that’s an argument for sort of, you know, not necessarily changing your strategy, but tilting your sector allocation within your CTA away from financials that most people already have exposure to and towards things like commodities. And actually, again, looking forward, one of the papers we’ll talk about today gets into that.

[00:27:49] Niels

I do agree with that. I do agree with it, having a large allocation to commodities.

[00:27:53] Rob

I think what we’ve got to bear in mind, Niels, is for most people the choice isn’t between, say, having 60/40, you know, most people don’t have that option, necessarily, or for whatever reason, of reducing their equity allocation and putting more into the CTA and getting the risk premia that way.

You know, if you are limited in what you can put into the CTA sleeve, if it’s only, say, a 20% allocation or even a 5% allocation, then at least from a theoretical perspective it does make sense to try. And you’ll get a better portfolio overall if that 5% or 10% or 15% or 20% has less exposure overall to equities and bonds than otherwise. But I agree, the best thing to do, if you can, is to increase, is not to do that, but then to have more in the CTA bucket. But, as I said, unfortunately that’s not always an option.

[00:28:36] Niels

Yeah, I mean, just one final comment here. If you only have 5% to trend, I don’t think you would need to worry about the equity allocation within that trend, honestly. But there we are.

I will leave it at that because we need to move on very conveniently to something that caught my attention and that you and I agreed that we should just maybe touch on, and that is that the CFTC (which is the regulator for all CTAs and CPOs in the US and even outside the US, if you’re in this business you’re probably most likely registered with the CFTC), they just came out this week and they approved (or maybe it was last month, it could have been last month), they approved what we could refer to as a perpetual futures contract.

So, what is that? Well, first of all, it’s the first of its kind and, secondly, it’s basically a futures contract that, unlike every futures contract so far that has an expiration date, this one has no expiration date. So, that makes it very different. Do you have any thoughts on this? And I should say maybe that the first one they’ve approved is linked to crypto.

[00:29:53] Rob

Of course it is, because, you know, this is where this idea came from. Right? I mean, actually, in fairness, the idea has been around for a long time, longer than crypto has, but the first place it’s been implemented in any science is in the crypto universe.

[00:30:07] Niels

But you are a crypto advisor, as far as I remember.

[00:30:09] Rob

I mean, I am, I hasten to say, I’m not a crypto expert though, but yeah. So, perpetual futures are very interesting. So, let’s start with a normal future. So, I just rolled some futures today because it’s kind of mid-June, and a lot of futures are expiring because they expire on IMM dates. So, we just had an IMM date. So, I’m rolling from, say, June to September.

So, you know, you’ve got futures that are, say, quarterly and, you know, that means every three months, if you like, the bet settles Right? So, the bet I’ve made through an exchange with somebody who’s on the other side of that bet, at that point we settle that bet. And let’s keep it simple and assume it’s a cash settled future. So, at that point we look at the prices and we see, you know, who was right? Did the thing go up or down? And we sort of settle our bets.

So that’s kind of, sort of in a simple, very simple way, the way that features work. Now, of course, they don’t actually work like that because of this thing called margin, a variation margin specifically. So, what that means in practice is that if you and I made a bet about, say, the price of bitcoin and I believed it would go up and you believed it would go down, then what would actually happen every single day is that the, you know… So, let’s assume we’re doing this through the CME rather than through a weird crypto exchange. We’re betting on the sort of standard bitcoin future, which I think is monthly but not quarterly. But let’s ignore that for a second.

So, even though the future doesn’t expire for a few months, in practice, if the price goes up you would have to pay me some money every day that actually goes via the exchange (you don’t sort of send me a check directly), and vice versa. And that sort of variation margin, that’s how that protects us from the fact that we don’t want to get into a situation where the future is expiring, and I’ve won and you’ve lost, and you suddenly turn around and say, well, actually, Rob, I’m not going to pay you. Tough. You know, there’s not enough money in my brokerage account to cover it. So, that’s what variation margin does.

Now, I want you to imagine the same bet, but let’s compress the timescale. So, instead of doing it every three months, we’re going to do that… we’re going to bring the expiry of the bet down to one day. So, the future now no longer is in three months away, it’s in one day.

So, what that means is that rather than kind of “settling our little bet every few months”, we’re going to settle our bet every single day. So, every single day we’ll look at the price, and we’ll do the sort of bet settling. And this means that the margin payment, rather than being a kind of gradual flow over the period of settling of the bet, is a thing that just happens once every time the bet settled.

So, although these are called perpetual futures, what they really are, in a sense, are daily futures – the futures that settle effectively every single day. And, in fact, if you look at sort of thing about cash flow, it’s going to be identical to the cash flow variation margin. And the sort of two fundamental important differences are, whereas, with the kind of quarterly bets every few months I need to decide if I want to continue having that bet. If I do, I need to make a new bet starting for the next expiry. It’s the same thing on the other side because it’s a perpetual future, it’s always on.

So, basically, unless you say something otherwise to the contrary, then when that daily bet settles a new daily bet will be opened up automatically for the following day. And the important key thing about that is, whereas, at the moment there can be a big difference between the price of the future and the spot of the underlying thing, with these perpetual futures, those two things will always be the same, basically. Under some kind of hand waving, no arbitrage argument, essentially, every time the bet settles (which could be daily or could even be more frequently than daily), the prices must be identical in the same way that when a normal future expires, it will expire at or settle at the cash price.

That’s the technical side.

[00:34:10] Niels

Yeah, can I ask a question?

[00:34:11] Rob

Or we can move on to discuss why.

[00:34:15] Niels

Just a question before we do that. And I’m not an expert in all the mechanics and the pricing of futures in terms of the theory here (I’m sure you are), but if I’m not mistaken… If I’m not mistaken, for example, interest rates play a role in the pricing of a futures contract. And so, in a sense you can think about the perpetual one as being oh, so how do I do this with the interest? How does that… What do I use? But are you suggesting that because it’s looked upon as a daily future settlement that maybe that is not an issue?

[00:34:52] Rob

To get very technical, the interest will be in there, but it’s going to be such a tiny, tiny component, you can pretty much ignore it. There may be scope for very smart people to do some kind of, you know, arbitrage. And, I guess, if you’re market making this stuff, you might have that in your model.

But for all intents and purposes, you can pretty much ignore that in the same way you cannot all the other factors that affect the difference in futures and spot like, you know, for equities, the dividend yields; for bonds, the yield on the sort of cheapest to deliver bond; for commodities, you know, the storage costs, the insurance costs, and so on, and so forth. You know, for all intents and purposes the futures price will equal the spot price. And that’s a simplification for people who don’t, for whatever reason, you know…

So, as to why you would do this. Well, you know, there is actually a corner of the financial world where something similar already happens, and I’m going to preface this with a lot of legal stuff because it’s in spread bets, and contracts are different, which are not legal in a lot of jurisdictions. They are in the UK but are not in the US and a lot of other places.

So, everything I’m about to say now is certainly not financial advice, and certainly if you don’t live in a jurisdiction where those are legal. But in the spread betting, in the CFD world, you have the option of trading what’s called a dated bet, which is basically like your future and actually the spread bet provider will hedge their own exposure with futures. Or you have the option of doing what’s called a daily bet, which settles every single day, and that’s going to look a lot like the sort of cash flow future.

And the reason why people would do the daily bet is, well, for starters, a lot of the people, the customers of these firms, for good reasons or bad (and I believe it’s bad reasons in most cases), they aren’t keeping bets open overnight, they’re closing them during the day, they’re day trading. And if you’re day trading it’s cheaper to do it with a bet that closes today because you don’t have to worry about overnight funding costs which, because these brokers are not peer to peer, they charging massive spreads basically, and it’s a very expensive way of doing things.

Whereas, of course, if it’s done through an exchange then it’s effectively a peer-to-peer trade. So, the only spread you pay is difference in bid and ask. So, it’s a bit of a nice environment. So, I think it sort of simplifies things. But also, frankly, if you’re someone who’s trading, you know, multiple times a day then it’s an easier sort of structure and way to work out yourself and you don’t need to worry about rolling.

[00:37:47] Niels

Well, that brings me to the next point. And that is, clearly, if this became the thing that many more… and the types of futures that we would trade become perpetual futures, clearly, that has some massive implications for firms involved in the futures industry that rely on volume because I imagine the volume would go down significantly. There would be no roles, for example. And so, it’s not surprising the CME Group and maybe other futures exchanges are not too happy about it.

So, they have essentially, as I picked up this morning, yesterday, sued the CFTC over this decision to approve a perpetual futures in the US. And I’m not going to get into the technicals, but they apparently argue that this should be classified as a swap.

[00:38:44] Rob

So, there’s an important point here which is to ask ourselves, well, why did futures exist originally? Well, originally they existed. I mean it’s the first futures were in the commodity markets. You know, exchange traded financial futures didn’t really sort of start until, you know, the early 1970s after the collapse of the US gold-backed dollar system and that’s when the CME started listing, you know, FX and then subsequently interest rates and equities.

They originally started in the commodities markets because they were there purely initially as a hedging instrument. And if you’re a farmer hedging your crop, well, you absolutely want a dated future. You know, if your crop’s going to be sold in June next year, because that’s when the harvest is, you want a June 2027 future available to you. You don’t necessarily want a perpetual future. It’s not necessarily a lot of good for you.

Now I do appreciate that the vast majority of trading, even in commodity markets, is not done by hedgers, or by physical producers, or physical consumers as well. It is done by, you know, by speculators, people like us, if you like.

But you know, there’s still an argument to be made that there has to be some kind of real price discovery going in on in the market. And for that to happen there needs to be a place for producers and consumers of the underlying commodities to go to hedge their exposure. And you know, I don’t know where that’s going to go. I mean, we talk about swaps briefly, will it start going to banks and doing to OTC swaps instead?

You know, that would be not-great for spreads in either market because, the capacity and the liquidity will be split apart. So, although as a person who trades futures and has to go through a painful process, you know, every few days of checking whether I need to roll, stuff and rolling it, and that’s a lot of code I wouldn’t need, I think, sort of for the futures market generally, I’m not sure it’s a great development, to be honest.

[00:40:47] Niels

I wonder if it’s all going to end up with these massive market maker houses, “quasi hedge funds,” or whatever we call them, platforms, who seem to have been taking on a lot of volume in many areas of finance. But we’ll leave that for that.

You know, I was thinking, while you were talking about this thing about farmers, maybe it starts in crypto because there are no more of these yield farmers. Didn’t you have yield farmers back in the early days of crypto, 20% risk free and all of that good stuff? They’re gone, I think now. So, now we can have perpetual futures.

Okay, let’s move on to something very important because every time our friends over at Quantica publishes the paper, we pay attention. And the Q1 paper came out very recently, only to sort this morning. It’s called Chasing Trends or Chasing Performance, and is, of course, a very appropriate topic. And as usual they’ve done a good job in writing about it. Could I ask you to talk a little bit about some of the things that they address in the paper?

People should go and read it themselves, but we will address some of the key topics because I do think it’s something that a lot of people (I wouldn’t say a lot), some people think about this idea of, maybe we should just trade the trends that are the strongest or whatever best performing. Let’s leave cocoa out because it didn’t perform for 15 years, and bang, you missed the biggest trend it’s ever had. Things like that.

[00:42:34] Rob

Yeah. There are sort of two ways I think you could think about, let’s say, dynamically changing allocations. You can talk about it over time. So, I’ve done research and other people have done research on, for example, whether you should get out of trend following or get into trend following after drawdowns. Winton did a piece on it a few years ago, it was quite interesting. My conclusion is probably not.

And the reason is that the sort of pattern of, you know, to get 10, the autocorrelation returns is not statistically strong enough to suggest that you would generally get either very good returns after a, you know, prolonged drawdown or very bad returns, you know, so do the drawdowns persist or do they continue essentially?

By the way, there is an important caveat there, which is if you are paying performance fees, then you absolutely want to invest in a fund when it’s in a drawdown because you’ll get the first bunch of returns for free, essentially, without performance.

[00:43:34] Niels

I have to correct you here.

[00:43:35] Rob

Okay.

[00:43:35] Niels

Not necessarily. If you get your own individual share class (I just want to make sure) then you get your own individual high watermark. So just be careful.

[00:43:43] Rob

Okay. Yeah. Okay. Thank you for correcting me there. So, it may be, you know, in a specific case it may be beneficial. But otherwise, you know, if the performance fees are linked to your share class, as you say, then we’re back to the original statement.

[00:43:58] Niels

Yeah, if it’s linked to the NAV of the share class, it’s different. You can get a free ride.

[00:44:03] Rob

You’re right.

[00:44:03] Niels

Yeah.

[00:44:04] Rob

Okay. Now and the other way… So, that’s sort of time-based predictability and the other is sort of cross-sectional predictability. So, can we say that particular instruments or particular asset classes will trend better than other assets or instruments?

And you know, for example, Yoav and myself have had some debates about that subject because, you know, he has a ‘fixed income only fund’ and he’s a big fan of fixed income. Given that we both used to be head of fixed income at AHL, although not at the same time (we weren’t co-heads), and that’s not really a surprise. But you know, that’s another kind of interesting debate.

Now, what’s interesting about the Quantica paper is it takes both of those dimensions simultaneously. So, what it says is say, well, if we look at the performance of trend following in different asset classes over time, we see some very interesting patterns. So, if you look at not quite the last 30 years, but basically the first decade of the century, the second decade of the century, and then the seven years since then, which sort of, actually, interestingly is not too dissimilar from my own career in the industry. It sort of spans a similar periods. So, from 2000 to 2009, they sort of see what I call a fairly broad-based set of performance. So, in that first decade, about a third of the performance comes from equities, about a third comes from currencies, and about a third comes from commodities. Fixed income actually detracts slightly in that first 10-year period.

Then in the next 10 years, that’s from 2010 to 2019, so that’s just before COVID, importantly, but after the GFC, it’s all fixed income. And I note that is also the period in which I was head of fixed income for AHL, followed by the period in which Yoav was. So, maybe, you know, is this correlation or causation? I couldn’t say. Anyway, we were lucky in our timing there, and commodities and currencies detracted, and equities provided almost nothing.

Anyway, if we look at the last seven years, that’s COVID onwards, 2020 to 2026, it’s all commodities pretty much, a little bit of a contribution from fixed income equities which actually detracts. And we talked about cocoa, but it’s not just cocoa. I mean, the last time we spoke, or the time before, we were talking about silver, I think it was. Obviously, we’ve been talking about crude oil. So, it’s very much not about the financial assets. So, going back to the conversation we were just having, you know, the last seven years, if the commodities futures market ceased to exist, that would have been a very sad thing indeed.

Anyway, so what they say is, well, these patterns are very interesting and I think if you’re a quant, whenever you see an interesting pattern, your immediate thought is, well, you know, for two things really. First of all, is there a reason why this is happening? Are there different explanations for why this is happening? And secondly, can I exploit this apparent predictability in some way?

So, could I say, well, you know what, it looks like we’re currently in a phase or a regime where commodities are doing really well. Can I shift my portfolio away from fixed income and equities, that are kind of sad at the moment, and instead put them into other buckets? So, that’s basically what the paper is about.

It’s a long paper, it’s a detailed paper, it’s from Quantica, they never do anything by halves. There’s a page full of equations, there’s lovely graphs. So, I’m not going to go through the sort of results in a lot of detail, but I am going to sort of point to a couple of interesting results.

So, one, the first thing to say is that whenever you see good performance, there are, kind of broadly speaking, two possible explanations for it. One is that you’ve got good diversification, so you’ve got a lot of things doing reasonably well. And the other is (and this goes back to our early discussion about the risk factor paper), the other thing is that you haven’t got good diversification, but everything is doing well. So, it’s not just that you’ve got one or two markets all moving the same direction. Everything’s moving in the same direction.

It’s not a diversification story. It’s basically, if it’s fixed income, then the Fed’s just cut interest rates very predictably. They’ve gone down to zero and we’ve made a lot of money in that period followed by rates rising again in a very predictable way. It’s just been beautiful. So, it’s not been that you’ve done well by being exposed to, you know, random things like Korean bonds, or Swiss bonds, or anything like that. It’s just a big one picture story.

So, they look at this in detail and they find that, in these periods when one asset class is dominating, it’s very much about a sort of single factor that’s driving everything. It’s not a diversification story. It’s the fact that you’re getting, in fact, less correlation. But it just so happens that everything’s moving up. So, it’s wonderful.

Now, that means that if you’re someone who, like me, who likes to allocate your portfolio in the most diversified way possible, well, actually this is bad news because if you want to shift your portfolio into asset classes that are less diversified, that goes against a lot of things, which is a bit of a potential issue.

The other thing they say is, well, actually, if you adjust your signals according to the strength of your trend, so you’re on what I’d call a continuous run – the binary trader, actually, that factor in of itself kind of already positions your portfolio to make money out of these things.

So, basically, if you were to look at your trend strength barometer, Niels, by asset class (and don’t do this), you’d see that your trend strength barometer, in the last seven years, was much better in commodities, on average, than it was in equities or in fixed income. For the previous decade it would have been very much about a fixed income.

So, in a sense, by allocating to things with stronger trends, which is what you do naturally if you’ve got a continuous system, that allocation is already happening. You don’t need to do anything extra on top of that. It’s just unnecessary.

So yeah, it’s an interesting paper and I quite like it. And I’ve done it myself, recently, when people publish a paper and say, well here’s something interesting. Oh, by the way, if you’re wondering if you can make extra money out of it, the answer is no, it doesn’t work. But it’s still a very interesting thing to talk about.

And I think a hallmark of a good research team is they look at things that don’t work as well as things that work and they publish both because it’s just as valuable to find out that something isn’t working as to find out that something is working. So that’s not for me. I like that. I like to get to the end of papers and for people to say, yeah, nothing to see here, but I hope you’ve enjoyed the ride.

[00:51:01] Niels

Yeah, I think I’m completely with you on that. And I think this is, I mean, I would have loved if they had even included some of the previous decades like the ‘90s and the ‘80s. And the reason I say that is, I remember so clearly what took place towards… I mean, interestingly enough, when you look at the chart, figure 1 in the paper (if people download the paper and they read it as they should), figure one is very telling to me because the middle one, the middle decade, 2010 to 2019, as you rightly state, it’s all fixed income.

So, you know, the conversations we were having towards the end of that period, which of course coincides with what people call the CTA winter, things we were talking about, people were talking about, oh, have things changed, has trend following stop working, etc. etc. And I always felt that what had changed was we just had a period of time where central banks were very, very good, or very, very lucky to keep inflation down, interest rates pretty much at zero for part of that time. And that also meant that some of the economies weren’t really moving kind of in a natural form.

And so, for me, it really painted a picture, and of course it’s easier to see as soon as we got into the 2020s, is that decade was not the new norm, it’s the odd one out. And meaning, I think if we went back to the ‘90s, and we went back to the ‘80s or the ‘70s, we would see more like the period of ‘80s, or 70s, we would see more like the period of 2000 to 2009 or more of the 2020 to 2026. Meaning that many more sectors, and certainly the commodity sector, are contributing strongly.

I remember another paper, and I don’t have the date of that, and I don’t know that I can disclose who wrote it, but there was a paper out a few years ago where they had looked also at attributions by sector during crisis periods and what they concluded, and this is a very reputable firm, was that the most consistent sector in a crisis is the commodity sector.

And so, I know it’s been very popular to focus more and more of the portfolio towards financial markets because they are more liquid usually, and therefore you can manage more money, and especially if you are being paid by just how much you manage, that’s a very strong incentive. But in my view, and there’s nothing new in this, in my view, that’s never been the best solution. The best solution is to build your portfolio as diversified, with as many truly uncorrelated markets, so to speak, meaning more commodities, which is also a philosophy we share certainly at DUNN.

So, I think this is really a great paper to highlight some of the things in that area and hopefully it’ll open up the eyes of many more investors in terms of the importance of “true diversification”. Acknowledging full well, of course, there are going to be periods where, say, a sector like fixed income, maybe even a sector like equities will be, by far, the best one.

But interestingly enough, if you go back (and I was just looking at some data that I had access to), I mean, if you go back 20, 25 years, I mean even through this bull market of equities, equities from a trend following perspective hasn’t been firing on all cylinders. So yeah, I think the paper is great and people should definitely read it.

[00:55:01] Rob

Yeah, I mean I was trying to find some statistics while you were talking because I think I’ve done this analysis myself. But for sure, during the 1970s, the inflation of the 1970s was fantastic for commodities trend following. It was just, you know, just amazing. So, that was definitely true back then for sure. The ‘80s and the ‘90s, I’d have to check. But yeah

[00:55:27] Niels

But you know, what’s also great about it? It’s a perfect segue to the next and final topic we have because, as you rightly said, people could be tempted at saying, oh, let’s just move into the things that had just worked. Well, the same is actually true when you design your system and you start looking for things that seems to work better based on historical data.

This is, of course, your domain, Rob, not mine, but it is a blog post from Wisdom Trading, that you found, about overfitting, especially in the new era of AI. So, I’d love for you to talk people through this one.

[00:56:06] Rob

Yeah, I don’t know about you, Niels, but whenever I go on LinkedIn or even Twitter, about 40% of my feed is people with producing back tests of strategies that are obviously been generated by AI, which is kind of depressing. And actually… I’m very early… This is a very early book plug. Very early book plug.

[00:56:30] Niels

I was just waiting for it. I mean it took you about 56 minutes for the first book plug.

[00:56:33] Rob

Well, you know, so my fifth book is coming out in December, but that is, you know, what it is. But my sixth book, which I haven’t even started writing yet, I’ve not signed a contract with the publisher or anything, but I started thinking about. It is going to be about backtesting and fitting.

So, I’m doing a lot of, as I said earlier, doing a lot of research and thinking about this. And I do think, certainly in the age of AI, that, you know, this is… I mean backtesting and overfitting was an endemic problem before, but now we’ve got AI, we can do it automatically. It’s great. So yeah, the wisdom article is full of wisdom, as you would hope. And the basic premise is one I kind of agree with myself, which is that, you know, it’s fine to use AI essentially to generate ideas, to generate strategy ideas.

That’s absolutely fine. I think AI is quite good as a sort of a creativity assistant tool, you know, so I think it’s quite good at, say, drawing silly pictures, for example. That’s what I mostly use it for, to be honest. Occasionally I use it to write little bits of code.

I’ve never used it to write any more than about 10 lines of code because I just don’t trust it enough. And I spend more time checking and rewriting it and I certainly never trusted it to do a backtest of a trading strategy.

But these guys like, well, it’s fine to use it to generate ideas, essentially. But the important thing is that once that idea has been generated, that it’s then tested in a robust fashion away from the AI. So basically, you need to have a process and it’s more about… People think that fitting and portfolio optimization is about maths. It is to a degree, but actually the most important thing is the process that you follow.

One of the main points about this book, when I get around to finishing, or starting it, will be about the process rather than about the maths. There are lots of books full of maths about portfolio optimization and fitting that you can feel free to read. But it’s about the process that you follow, once you have an idea. And it doesn’t really matter whether that idea has been generated by an AI model or by you by looking at a novel source of data, by looking at this, say, this Quantica paper and think, well, that looks interesting, or whether it’s been generated by more of a kind of old fashioned data mining exercise where, you know, you’ve gone and sat and looked at charts and said, well, it looks like… Well, you’ve got this wick of this candle engulfing this Ichimoku Cloud near that Fibonacci level. That looks like a bicycle to me. It doesn’t matter where it comes from.

But ultimately you then have an idea which you need to be able to robustly test, bearing in mind that idea will almost certainly have been conceptualized using future information that you wouldn’t have had at the start of your backtest. So that is the key point. Your backtest process needs to be able to handle that in a robust and sensible manner.

And the end output of that backtest, it will essentially not be a yes or no. Going back to the very first question, near the start of the podcast, when you decide to add or delete strategies, actually it’s not really a yes or no question as to whether an idea works or survives in your overall portfolio of ideas or strategies. It’s going to be a percentage. It’s going to be a probability that it’s good and that will translate then into a portfolio weight.

So, if an idea is outstanding, it may end up with 10% or 15% of your portfolio because I’m sort of assuming you’ve got a fairly decent one already. And if it’s rubbish, it may be 0.01% but, you know, it’s still in there because like cocoa, the chance it could work… You know, there’s never a 100% chance that something doesn’t work. So, that’s, you know, if people want to start using AI for backtesting, then that’s really the safest way of doing it.

Otherwise, yeah, you’re basically back to the bad old world of overfitting, except that now you’ve got a nice tool to do it for you that’s using up lots of water and you know, tokens and generally being a very wasteful resource.

[01:00:59] Niels

So, the interesting thing will also be, because we do know that some managers already use AI, right? And some, obviously you should be completely transparent about it. But it will be interesting to see how do you, as a due diligence person, actually evaluate that?

I mean for some things is probably pretty easy to understand how the AI is used, but there will be, I’m sure, things where, well, you don’t really know what goes on behind the scene of the AI. And so how do you evaluate that?

[01:01:37] Rob

Yeah, I mean, it’s an excellent question, Niels, but actually, I’ve never been on that side of a due diligence. I’ve always been on the other side, you know, back when I was managing money. I’d have had people come to me and say, right, you know, we’ve got this allocator, this sovereign wealth fund, this pension fund, who wants to invest in us. Here is the 50 page questionnaire we have, you know, that we have to fill in at the RFI, you know, and we’re stuck on question seven, can you help, sort of thing. That’s been my exposure to that process.

But, you know, I probably think I could do a reasonable job of designing a questionnaire for a CTA investment because having been on that side, you know, poacher terms, gamekeeper, I know what questions to ask. And the questions I would ask now would be exactly the same questions I’d have asked 20 years ago before AI became this big thing.

And they’re down to, again, they’re really down to the understanding the research process. So, tell me what happens through that process, from idea to live trading. What are the steps in that process? What are the checks and balances, you know, what is the robustness, what’s the peer review, and so on, you know, and so on, and so forth?

And the fact that, at the beginning of that process, it is typing into ChatGPT or, you know, other AI models that are available, you know, analyze this data, come up with a trading strategy that works. To be honest, that’s no different from, you know, 20 years ago that might have been a junior researcher sitting at their Bloomberg terminal or other terminals that are available, and looking at lines and saying, well, it looks like this predicts that. I think I’ll see if that works.

You know, there’s not really, from the point of actually coming up with the idea, almost certainly using forward information to actually putting it into your portfolio, that process shouldn’t really be fundamentally different. So, I wouldn’t necessarily have to create a whole custom load of sheets about AI unless it was being done wrong.

So, if I started not liking the answers to those questions, then I would be circling back and saying, well, okay, you need to tell me things like for example, this AI model, is it an external or an internal model? And if it’s an internal model, which you trained yourself, did you train it on, say, data before a certain date and therefore have a pure out of sample period? You know, you’d have to start asking questions like that.

But that would only be the case if, as I said, the process was effectively flawed. Because, ideally, you wouldn’t be trusting the AIs sort of evaluation of itself, which means you don’t need to worry too much about whether it does have, you know, kind of forward information plugged into its learning set. Now if it’s a public AI model, then all bets are off. You’ve no idea where the data is coming from, whether the data is real. It would be a lot more of a concern in that case.

[01:04:30] Niels

Yeah. And to that point I imagine that a lot of these AI trading strategies that are being offered not by CTAs but on the Internet at large, of course probably do make use of public AI. So, that is definitely a dangerous place to be spending your time and your money, would be my guess.

Anyways, I think we’ve come up against time limit today, even though there was one more paper that we kind of briefly touched on last week and we never really got around to it. Maybe it’s kind of a little bit the same. And it’s a paper from another group of our friends, namely the people over at Man.

I mentioned it last time, it’s called Don’t Look Down, because I thought it was kind of an interesting summary, reminder, as to how drawdowns show up in different types of assets, how frequently they show up, how deep they are. And I think some people might be surprised by some of the findings.

And I mean, unless you want to say something about it, I think people should just go and read it themselves. It’s obviously on the Man group’s website, under Insights, but anything…?

[01:05:47] Rob

Yeah, it’s an interesting paper. I mean one thing it gets into quite early, which most people don’t bother with, is, actually, what is a drawdown?

[01:05:55] Niels

Yeah.

[01:05:55] Rob

You know, because you kind of have a very pure definition, which would mean that, essentially, you’d be in a drawdown almost all the time, because unless you’re literally making a highwater mark that day, you know, then technically you’re in a drawdown. So, he does actually define drawdowns which is quite useful. And then yeah, he looks at them across a number of different asset classes, but also you know, risk, premia, so, as we mentioned before, things like trend value, momentum and quality.

So, he looks at the sort of distinctive characteristics of different asset classes in terms of, you know, how big their drawdowns are, how often they occur and things like that. You know, I’m not sure how sort of statistically significant these results are because, by definition, drawdowns are episodic. You know, they’re not happening all the time.

And if you apply a strict definition to them then in your dataset you’re probably not going to get more than, say, 10, 15 or maybe 20 of them in, say, a 50 year backtest. And that does mean I’m not sure you can draw too many firm conclusions about whether equity drawdowns are, you know, substantially different from say bond drawdowns on that basis.

But the more interesting part of the paper for me is the second half where he looks at the interaction of drawdowns. So, you know, what drawdowns tend to happen at the same time. And that’s quite a nice take because I think one thing that people do a lot is what I call the sort of crisis alpha test.

So, the classic thing is you say, well, you know, this is your standards sort of… If you look at any deck from any CTA, you’ll see this slide in there. It’ll be, these are the periods when the S&P did badly, or when 60/40 did badly, or when there were recessions. And look how well our trend following portfolio performed in these episodes. Aren’t we great? And you know, sometimes they’re a bit naughty and they cherry pick a bit and they don’t put in there the ones where trend following didn’t do quite so well. Like for example, COVID.

Anyway, so that’s sort of a classic thing. This is a bit more of a rigorous way of doing it and it’s a bit more of a sort of a different take on it, if you like. Because, you know, they’re not just saying, well, when the S&P was in a drawdown, how well did trend following do? They’re taking a much more holistic view and saying, well, when these assets are in a drawdown, what other assets tend to be in a drawdown and not in a drawdown? So, it’s a different way, if you like, of identifying what could be a good diversifying asset. Although, as I said, there are caveats.

You know, there aren’t that many drawdowns in most data sets and therefore you need to be wary about drawing firm conclusions from those figures. But yeah, it’s an interesting paper. I think it’s quite interesting to contrast the Quantica papers and the Man papers.

The Man papers are a little bit friendlier to read, I should say a bit less technical, a few more pictures, a bit shorter. So, they’re more your kind of light reading over your coffee in the morning. Whereas, for the Quantica paper you normally want to have a couple more cups of coffee and get your head down before you read those. But both great papers.

[01:08:56] Niels

Yeah, although I will say I think Man also produces the long deep dives, but they produce a lot more. So, you’re right, some of them are a little bit shorter and that’s also great. Sometimes you need that. So, it’s kind of interesting enough. It’s kind of like drawdown diversification. How can I get stuff in my portfolio that does not drawdown at the same time? So yeah, good stuff.

Rob, this was perfect. Thanks so much for preparing yourself for all these topics and I’m sure people enjoyed kind of this buffet of different types of trend following related stuff. And if you did, please head over to your favorite podcast platform and leave a nice rating and review for Rob and of course for all the other co-hosts that we have every single week.

Next week I’ll be joined by another one, great one, Mark, and he also has usually some really interesting topics that he brings along, but it’ll also be your chance to send some questions for Mark. You can as usual send them to info@toptradersunplugged.com and I will do my best to bring them up.

That’s it for now. I’m off to celebrate my children graduating both with a bachelor and a master’s, different ones. So, a big family moment for me. So, I’m going to leave it at that, yeah, and say, from Rob and me, thanks so much for listening. We look forward to being back with you next week. And in the meantime, as usual, take care of yourself and take care of each other.

[01:10:31] Ending

Thanks for listening to Top Traders Unplugged.

If you feel you learned something of value from today’s episode, the best way to stay updated is to go on over to your favorite podcast platform and follow the show so that you’ll be sure to get all the new episodes as they’re released. We have some amazing guests lined up for you and to ensure our show continues to grow, please leave us an honest rating and review.

It only takes a minute and it’s the best way to show us you love the podcast. We’ll see you next time on Top Traders Unplugged.

This podcast expresses the views of its hosts and the guests appearing on the podcast as of the date of its recording, and such views are subject to change without notice. Top Traders Unplugged do not have any duty or obligation to update the information contained herein.

Furthermore, Top Traders Unplugged make no representation to its accuracy and it shall not be assumed that past investment performance is an indication of future results. Moreover, wherever there is a potential for profit, there is also the possibility of loss.

This content is made available for educational purposes only and should not be used for any other purpose.

The information contained in this podcast does not constitute and should not be construed as investment advice or an offer to sell or a solicitation to buy any securities or related financial instruments in any jurisdiction.

Certain information contained herein concerning economic trends and performance is based on or derived from information provided by independent third-party sources. Top Traders Unplugged may believe that the sources from which such information are obtained are reliable.

However, Top Traders Unplugged cannot guarantee the accuracy of such information and has not independently verified the accuracy or completeness of such information or the assumptions on which such information is based. This podcast, including the information contained herein, may not be reproduced, copied, republished or posted in whole or in part in any form, without the prior written consent of Top Traders Unplugged.