234 Systematic Investor Series ft. Rob Carver – March 11th

[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 Rob Carver and I, Niels Kaastrup-Larsen, where each week, we take the pulse of global markets through the lens of a rules-based investor. If you’re new to the show, I hope today’s episode will trigger your curiosity enough to check out the back catalog, listen to some of the past episodes that you may have missed, like my conversation with Alan last week where we discussed a couple of new research papers, of which one was about the footprint of CTAs in the markets as well as some of the key takeaways from our recent conversations with pretty much all of the largest and best CTAs in the world, which you can find in the Top Traders Unplugged series.

Also, if you did miss the Wednesday episode, I really would encourage you to listen to this midweek conversation with David Dorr, where we dive into a new topic that I personally find fascinating and which David called Galactic Macro, and where we tried to keep a very open mind to all of the recent releases of information from the Pentagon and US military about encounters that they have had with unidentified objects, but also then try to see if there could be made an investment case for this development, hence the new term Galactic Macro.

Anyways, I encourage you to listen to as many episodes as you can handle once you finish listening to Rob and me today. Anyway, Rob, good to be back with you. So much continues to happen between our conversations. How are you doing my friend?

[00:01:52] Rob

Someone actually messaged me and said, “Why is it whenever you go on the podcast, you always talk about the weather when Niels asks how it’s going?” I’m like, “Well, I’m English. That’s what we do.” In an awkward conversational situation, we just reach for the weather. So I can tell you it’s dark, it’s cold, it’s raining, and yeah, summer can’t come soon enough.

[00:02:12] Niels

All right, well, you could move to another country, Rob. Then you don’t have to talk about the weather, I guess. But anyways, that would be boring of course.

[00:02:18] Rob

That would be boring.

[00:02:19] Niels

Anyways, as I always say and I really do mean it, we have a wonderful lineup of questions and topics. But before we do that, normally, I would run through some kind of market wrap, but frankly, we are recording two days early this week and that’s because of my crazy travel schedule at the moment. So I don’t have anything detailed that I’m sure you haven’t already read in the news.

Instead, what I asked Rob to do was to maybe give his views and observations, what he’s taken away from the episodes that he had listened to so far in the series with the CTAs because even for myself where I would say I know most of the managers, there has actually been a few surprises along the way. So anyways, Rob, I’d love to hear your overall thoughts on what you’ve listened to so far.

[00:03:10] Rob

Yeah, it’s an amazing series, I have to say. It’s really good. I think one of the things I find most interesting is coming back to people who actually I know very well, like Russell Korgaonkar who I used to work with at AHL, Katy Kaminski, Marty Lueck from Aspect, just some random results of people I know quite well actually, but I haven’t heard their views for many years. It’s sort of interesting to see how people’s views develop and change over time.

So for example, if I just think of myself, I’d be more keen on fitting my trading system, whereas now, I try and avoid fitting completely. That’s just an example. So yeah, I would say even if you’ve heard the episodes that you’ve done with Katy in the past for example, come back and listen to them. Again, with Marty, he’s been on the podcast before, but come back and listen to them because it’s really interesting especially given the developments faced in the markets last year but also developments in things like AI and technology generally. There’s always new stuff to hear there.

[00:04:22] Niels

I think that’s also a little bit what, I wouldn’t say surprised me, but that I haven’t thought about too much, and that is just how we, as managers, evolve. Because when you’re in it day by day, you don’t really see the evolution. But as you rightly say, when you hear people speak a few years after you heard them last time, you can see that something has evolved in their views, in their systems, even the way they build their businesses. There’s a lot more solution providers than I thought there would be in our space now. So I agree.

Anything that stood out to you since we last spoke a few weeks ago in the world, in the markets? Anything that you recall from that period?

[00:05:05] Rob

Yeah. It’s been an interesting year so far actually. Just looking at my own performance, the first month of January was pretty flat. But actually, in February and now into March, I made a little bit of money. So I’m actually up about just under 4% for the year, which is all right. So we’ll see how that goes. Obviously, that’s not as spectacular as last year’s performance at the same time, but it’s still pretty good. Just looking at where that performance came from, it’s not an obvious kind of pattern.

My best market was actually the VIX. I did pretty well on the bubble as well, made some money in CAC, made some money in lean hogs, and then on the downside, I lost money in US 20 years. Obviously, interest rates have been… In history, the bond market was a boring place but that’s no longer the case. So I think yesterday or the day before, there were some pretty sizable moves in interest rates futures coming up on my market monitor report. So definitely not a dull place in the bond markets that’s for sure.

If I look at how I am positioned going forward, the risk remains still quite low, probably running at about somewhere between a third and a half of my typical target level. Actually, I do still have more risk in the bond market than anywhere else, followed by the equity market, and then the agriculturals. So it still looks like I’m making what you might call factor one, factor two there where factor one is equity market risk and factor is interest rate risk.

In case people aren’t familiar with that terminology, what you do is, just to get really geeky for a second, if you take the correlation matrix of all the assets you’re trading and you do an eigenvector decomposition, the first two eigenvectors that normally come out are something that’s basically equity market risk and something that’s interest rate risk, just in case people are wondering, “Why are we suddenly coming up with these terms factor one or factor two.”

I’m actually still short bonds and that’s my biggest position. I’m short US 20-year short bonds, and short VIX. Those are all, I would say, kind of pro-risk bets. So I’m betting that interest rates are going to rise and I’m betting that volatility in equities is going to fall. So it’s kind of a pro risk bet. On the long side, I’m long CAC, I’m long Euro stocks, I’m long Nikkei. So those are all, again, pro-risk bets. But I’ve got some other positions in there. I’m long soymeal. I’m short hogs, I’m short gas.

What am I doing with gold and silver? Let’s have a look. I’m short gold and silver as well. So there’s kind of a big factor one, pro-risk, interest rates falling bet there, but not a massive one and there’s also some other things in there so there is a bit of diversification.

[00:08:20] Niels

Sure. Well, we’ll come back to this, circle back to this once we get to your topics, so that’ll be interesting. Just to talk very, very briefly about March so far, obviously, from a trend following perspective, people have already, I’m sure, worked it out from your comments. It’s very similar to what we saw in February and where we left off in February. Those trends are extending to some degree, in particular, in interest rates still which has been good for our industry so far.

The BTOP 50 Index, and this would be as of probably Tuesday, maybe Tuesday or Wednesday this week, I’m not sure, but it was up 1.81%, up 3.14% for the year already. SocGen CTA Index up more than 2% and up 3.44% already this year. The Trend Index up 2.5% so far this month, up almost 3% for the year. The Short-Term Traders Index up 72 basis points for the month and up 37 basis points for the year.

What’s interesting about this, to some degree, just because we talked about the term crisis alpha already, is that of course CTAs are doing perfectly well and there’s absolutely no crisis in the market. Last year, we made money during a crisis. This year, we are making money without a crisis. So yeah, again, I just think it shows the wonderful adaptability of this strategy. Equities are up slightly for the month so far and bonds are down slightly.

Anyways, we have quite a few questions that came in. We’ll do our very best to address as many of them but also as much of them because it’s actually not just one question people are sending in, it’s usually two or three now, and that’s fine. That’s great. We appreciate it. But it also means we sometimes have to be a little bit clinical in terms of how we approach them. Anyways, one that came in that I thought maybe you can quickly deal with is from Gene.

Gene writes in, “Found the testing discussion with Richard very interesting, but the line between overfitting and underfitting is still a bit unclear to me. If backtesting on decades of data shows that trend following, with a specific market such as gold, makes more money with a longer lookback, but another market such as gasoline makes more money with a shorter look back, would it be overfitting to use a different lookback for each market?”

Actually, I think I’m just going to stop it here because the rest are just more or less the same, because it’s a principle question, which is a good one, by the way. Because I think there are different ways which we’ve actually in the series, and sometimes I forget which conversation it is, I think it’s the one that’s coming up with Winton soon where we address this point. So stay tuned to that, Gene, as well. But let’s hear what Rob has to say
about this.

[00:11:19] Rob

You’ve got Mr. Harding coming on then, do you, Niels?

[00:11:21] Niels

No. We have Simon Judes who’s the Co-Chief Investment Officer?

[00:11:25] Rob

Yeah, cool. Interesting. I guess there’s two ways of thinking about this. One is the ideological view, and one is the statistician’s view. One ideological view is that you actually do a minimum amount of fitting as possible and then you treat all markets the same.

So for example, in my fitting, I treat all markets the same and the exception I make is trading costs. So if an instrument is particularly expensive to trade, I won’t use shorter lookbacks on that. But that’s not because I think they are necessarily worse before costs, it means I’m basically assuming that the pre-cost performance is the same, and therefore, after costs, I know they’re going to be terrible. So I just don’t use them.

That’s one way of doing it. To be honest, that’s a perfectly valid approach to them and it certainly makes your life a lot easier because you have to do a lot less work. Now, I would describe myself as an ideological plus. I am prepared to move away from that ideological position if there is strong enough statistical evidence for it, and also ideally, some kind of economic justification.

Now, the economic justification as to why a particular instrument should show faster trends than another one is difficult to be honest, and it’s also difficult to say why those patterns should persist over time. That’s because you might say, “Oh, well, everyone knows that FX traders are more gun happy than equity traders. Therefore, there should be shorter trends in the FX space.” That’s not a very compelling argument anyway. There’s not really any evidence as to why that should be the case. Even if it was the case, why should that particular pattern persist, for example?

Having said that, if there is extremely strong, compelling statistical evidence as to why one trend is better than another, then I would be prepared to at least look at it. Not everyone would. A lot of people would say, “You know what? Just don’t do any fitting.” But that’s fine, as I’ve said, but I’m prepared to at least look at it. Now, the problem is (and this is a general statement about fitting generally) you always need a lot more data than you think. You may think multiple decades of data is enough, and actually, that is enough to make certain kinds of decisions and not others.

Without getting too technical, it’s to do with these statistical properties of the distributions of returns that you’re either looking at or comparing against each other. So if you have, for example, two trend following rules for a particular instrument, the chances that one is going to be significantly better than the other to such an extent even with multiple decades of data are actually quite small. It’s quite difficult to find a statistically significant difference between performance of fast lookback and performance of slow lookback across different instruments.

Generally speaking, the amount you see of it is exactly what you’d expect by luck. Most of you are probably familiar with the concept of a p value. If you set a p value of 0.05, that means that you’re looking for less than a 5% chance. The finding you’re looking for is just pure luck. But if you’ve got 100 instruments in your portfolio, you’re going to find five where you see that distinction just by fluke anyway. So if you see like 50 and they’re all highly significant, much more significant 0.05, then there may be something going on there. But chances are you’re going to see five or six and they’re going to be significant 0.03, 0.04.

Well, that’s just what you’d expect by luck anyway. Generally speaking, whenever I’ve done this exercise (with a couple of exceptions), that’s pretty much what I’ve seen. So I’d say on balance, don’t do this either. Ideologically it’s unlikely the evidence is there. Having said that, there are cases where it does look like there is evidence for that, for example, in equities in particular, faster trend following strategies don’t seem to work so well, and if not worse, it’s like 1990, for example. You can maybe pull out one or two isolated cases like that.

So for example, with the carry system again, it seems to work well in one set of asset classes and not another set of asset classes. So if you’re going to go down this route, then I wouldn’t do it in individual markets, I might consider it on asset classes. This is not for the faint-hearted. You really need to have a really good solid understanding of statistical evidence and testing. As I said, the fact that there’s no economic justification as to why this should be the case would also concern me.

[00:11:44] Niels

Yeah, I think that’s a great answer, and definitely, I think we come across both camps in terms of how they select their parameters depending on markets and sectors. The other thing we did pick up for sure in the conversations we just had was this point that you also made, Rob, and that is, some markets are simply either not as liquid or more expensive in general to trade and then people slow down the lookback or the speed of the trend. That’s exactly what you pointed out. So hopefully, that was helpful, Gene, and everyone else dealing with this particular question.

The next one is from James. “I’m a big fan of your podcast. I have two questions for you: trailing slash ATR stops used in moving average systems. What type of stops are commonly used in moving average systems? I noted instances trailing ATR stops have triggered, but the buy condition is still met above certain moving averages. If we do use these kinds of stops, what are the common rules for reentry?”

Okay, so that’s question number one. Maybe we can deal with that first.

[00:17:21] Rob

Yeah. I don’t actually use a system like this, but I have written a book that describes a system like this. What I do in that book is say let’s suppose you get stopped out, say it’s a long position, you go into the long position on the moving average, you get stopped out, but the moving average is still long. Which can happen. That’s one of the reasons why I don’t like stops because you end up with this inconsistency and you have to make sure you calibrate the size of your stop to the length of your moving average, otherwise, you get weird things going on.

So what I say in my book is (and this is what I would do if I was trading that kind of system is basically to say), don’t do anything until the moving average has actually changed sign. So when the moving average indicates you should go short, at that point, go short, and obviously then place a new stop. So that’s how I would do it. I don’t really see any other way of doing it, to be honest. It makes no sense to go long in a moving average, get out a stop, and then go back in again straight away. Mechanically, you’re just doing two trades. What’s the point of that?

Some people might make the system more complicated and add another indicator to indicate when they go back in again but that’s probably back to overfitting to be honest, from my perspective. Niels, I don’t know if you’ve got more familiarity with these kinds of systems than I do.

[00:18:36] Niels

No. Funny enough, I’ve never actually used moving average systems in my career so I don’t really have anything to add to your thoughts there, so that’s fine. I’m going to skip over the next question, James, just because we have more questions that we want to get to. The next one is from Chris. “I was curious, when you run a backtest how far back you go. I know that the default is as far as you can, but what about scenarios where some contracts started in 1980s and others in the 90s, and some of them in the noughts. Say 1/3 of your initial capital allocation started trading in each decade, do you just run one-third risk in the 80s, two-thirds in the 90s and full risk in the noughts?” Any thoughts? You are the master of backtest, Rob.

[00:19:29] Rob

Okay. What I do is (it may sound a bit complicated), but basically, within my system, I have essentially a continuous process that’s adjusting for the number of markets that I’m trading and also the weights to those markets and also the amount of leverage I need to apply to hit my risk target. Let’s take a simple example. Suppose you start off with your data set in say 1970 and you’ve got one market you’re trading, let’s say it’s corn futures. I think my data for them goes back to then.

So then you’d have 100% of your capital allocated to corn and you’d be applying leverage of 1 to hit your risk target because you’d be running corn at, say, a 20% risk target on your system overall to get to 20% risk target. The ratio of those two numbers is 1. That’s simple. A few years later, suppose that another market arises like, I don’t know, cable, GBP, USD, FX. Let’s say for the sake of argument, you keep things really simple and you allocate half your capital to cable and half to corn, clearly more complex ways of doing it but let’s just assume that’s what you do for the time being.

Now, you’ve got two markets in your portfolio. If you actually look at the amount of risk that generates, it’s probably safe to assume that those things have got a correlation of about zero. That means that the risk you’ll actually be getting (assuming that you can forecast risk perfectly, which you can’t but you can get pretty close), would be 20% divided by the square root of 2, which is about 14%, something like that. So what you need to do is basically apply additional leverage to your positions to get back up to the risk target you want to hit.

This actually relates to something that we’ll talk about at the end of the episode as well, because people are interested in hearing that. So the advantage of this approach is I can keep stuffing more and more markets in, over time, and potentially even taking markets away. But generally speaking, in a backtest, you’re just adding markings. I’ve got a process for allocating weights to those and then basically this number (and I call it the IDM, the instrument diversification multiplier), that will then automatically adjust to give me the appropriate amount of risk that I need.

That’s obviously not really an answer to the question, but I guess to try and answer it, framing it in the terms of the question, essentially, I’m trading all of my capital in the 1970s, I’m trading all of my capital in the 1980s and so on and so forth. But what I’m doing in each of those decades then is allocating that capital between whatever instruments I have to trade with at the time and then leveraging my portfolio appropriately to hit my risk target given the correlation of those subsystems.

[00:22:22] Niels

Now, then Carl wrote a small novel when he sent in his questions, but I have shared them with you because I don’t think necessarily we’re going to get to all of them. So first, let me just read what Carl wrote. “I continue to enjoy your interviews and enjoy the broadening of perspectives that you’re sharing. I know it’s a lot of work and I want to assure you it’s greatly appreciated.” So of course, we certainly appreciate that feedback.

The first couple of questions are about what happens when a brokerage firm, a futures brokerage firm, an FCM, goes belly up. Of course, this is actually somewhat relevant for the whole industry and why we actually like trading futures. But then actually, Carl wrote to me that you’ve written (I think he referred to this) about this in one of your books. It doesn’t mean that we shouldn’t talk about it today. But if we are, we’re going to have to make it brief.

So I don’t know if you want to comment on these particular questions. Otherwise, I’ll jump to another question. But do you have a short answer that you want to share on that particular point?

[00:23:28] Rob

The short answer is actually that I’m not a lawyer or an expert on regulation and compliance. So anything I say will almost certainly be wrong and you shouldn’t rely on it. But there have been examples of futures brokers going bust and MF Global was a famous example. That was particularly problematic for me because I was working AHL and Man Group used to own MF Global and they were still a big prime broker for us.

Potentially, we actually managed to manage the risk very well. I think in the end, we got away with an extremely small loss but that could have been very nasty for us indeed. I think as a general rule, it’s definitely much safer to trade futures with an onshore exchange backed by a big clearing house through a good, well capitalized, well-regulated broker than it is to trade, I don’t know, just to take a random example, crypto through an offshore exchange run by some people starting with an F, just as an example.

But that doesn’t mean to say that there’s no way you can ever lose or be exposed to counterparty risk with a futures exchange and a futures broker. Of course, there’s still potentially things that could go wrong. It’s just that there are many, many more layers of protection there. But it may be the case that yeah, ultimately, you will get your money back, but not necessarily straightaway. But yeah, it’s a question that really needs addressing by somebody with a much more detailed knowledge than I do.

[00:24:54] Niels

I completely agree. But I do have one little follow up which you may know because I think that’s more of a factual question. If something happens to the broker you’re using and you have open positions in futures, do you know how they get transferred to another brokerage firm? Because your counterpart is essentially the exchange. So is there a mechanical way that they say, well, now the contract is not held with XYZ broker, it’s now with ABC broker because it’s still at the exchange. Do you happen to know that?

[00:25:28] Rob

My assumption would be that the broker obviously has an account with the exchange with a particular number on it, and that’s going to have in it the commingled customer positions. My working assumption would be that whoever takes over that account, which on day one would be the administrator, they then would essentially control that account and it just depends on what they want to do. If they want to continue running the operation as a going concern, then most likely, they wouldn’t do anything. They’d just let the customers carry on trading.

If for some reason they needed to wind things up, then they’d probably start closing those positions. I guess in an FTX type situation, you’re probably not going to want to start closing positions because you want to reduce the amount of exposure you’ve got to everybody as fast as possible. But, and this is speculation on my part, I think if a futures broker went into administration, the most likely outcome would be that the administrator would try and run it as a going concern, and as a customer, you’d see no difference probably.

[00:26:30] Niels

Right. Okay. But just to stress, Carl, whatever we’ve just said now is probably wrong so should not rely on anything we just said to be frank, but we’re just trying to help here. Anyways, then he moves on. Carl moves on to talking about a book that was just released by our friend, Andreas Clenow, who released the second edition of Following the Trend. Following the Trend number one, I certainly have recommended many times for people who wanted to understand what trend following is. I’ve got a few bones to pick with Andreas on the second edition, but we’ll see if we get to that.

Anyways, Carl writes, “In reading Andreas Clenow’s newly released second edition of Following the Trend, an impressive revision and well worth it, by the way, he shares an example of a $10 million account that holds $135 million notional open exposure, $123 million net on pages 99 and 101. When I saw that, my jaw dropped and my chin hit the table. My surprise is not based on experience, but I was surprised nonetheless. It should be noted that he limits risk by using a contract ATR calculation and limits the ATR per position to 0.002 of equity, two tenths of 1%.” So 20 basis points. I guess that’s easier to understand.

“Then he takes every position signaled by his long-short formula in his 80-plus contract investment universe. I was surprised that there is no overall risk budget such as, for example, limiting the total ATR risk to say 20% of equity and ensuring total equity never falls below 25% of total notional.” So, a four part question, “What are your thoughts on having a small position risk of 0.02,” so 20 basis points, “and then taking all signals across the investment universe without overreaching risk budget?” Then he asked, “Is this what Jerry does with his 200-plus contract universe?”

[00:28:36] Rob

This is pretty standard, right? I just explained the way you construct a multiple instrument position. Essentially, you set a risk target. Let’s say it’s 25% a year. Actually, let’s use the same number as Andreas. What was his number? He’s got total ATR as 20% of equity.

[00:28:52] Niels

He’s using 20 basis points.

[00:28:55] Rob

Yeah. Okay, well, let’s say it using my units, which would be 25% annualized standard deviation target. You target your individual positions to that. That’s kind of like your risk budget per position if you like. Those terms aren’t exactly the same. Then you apply and multiply to get to your overall risk given that all of these positions are mostly uncorrelated. But yeah, there’s no mechanical way of saying, what’s the risk of that overall portfolio?

One of the reasons why we do that, in our world, is that we are basically not like an equity long-short manager, for example, who will typically say, “I have a 10% standard deviation target and I want to hit that every single day because I want those consistent returns.” What we’re doing instead is saying, actually, there is not much going on in the markets, like now, so we’re not going to have huge positions everywhere, we’re just going to have a certain amount of risk.

But you know what? It sounds like a year ago when things went absolutely crazy. We want to have much bigger risk on, so we want to allow our risk to swing. Mine swings probably by a factor of 5 or 6, potentially, between a really quiet month and a really insane month. Having said that, I do actually have on my own system some, I call it an exogenous risk overlay. One of those things will look at the whole portfolio and look at the portfolio risk. If the risk is more than twice my average, at that point, it will step in and start to reduce my positions. But that will only happen once every couple of years, maybe. It’s a safety barrier. It’s not part of a system, which is why I used the term exogenous.

Because we’ve got lots of small positions, lots of positions, lots of risks, lots of diversified risk, but each of those positions we want to allow to do their own thing, effectively, and then the portfolio risk that comes out at the end is just what it is. It’s not something that we target in the long run. We expect to get a certain amount of vol in the long run, but on a day to day or month by month basis, there can be big differences in risk.

[00:28:55] Niels

Carl goes on to basically try and summarize this and what he’s saying is that if you have 80 positions, and he’s referring this to the way Andreas does it in the book, if you have 80 positions, you’re using 20 basis points. That’s equal to 16% risk. But then when you calculate it as a leverage factor, it’s $135 million on a $10 million account, or $123 million he calculated on net on a $10 million account. So that’s still 12 or 13 times leverage and he’s concerned about that. He’s asked me also, can a low ATR Eurodollar future conceivably lose 50% overnight, etc.? So there’s a lot of questions in there.

[00:31:56] Rob

The thing is about leverage, which obviously is equal to the gross exposure divided by your capital. It’s the simplest risk measure but also the crudest and the worst and it’s the most simplistic. In fact, I don’t know if it’s still the case, but I think it was the Swiss financial regulators who used to limit the amount of gross leverage a particular fund could have before they were registered in Switzerland. This may have changed. This is quite some time ago now. They used to ask us, “What’s your leverage?” We were like, “Well, this meaningless number but here, have it.” We have this meaningless number.

So why is it a meaningless number? Well, imagine if you’re targeting 25% vol. If you were trading just crypto, your leverage factor is going to be about 0.5 because crypto vol is roughly 50%. So to target that 50%, you’re only going to need half of your capital. Your leverage is actually less than 1. Only half your fund is going to be exposed to the thing you’re trading. If on the other hand, you’re trading Eurodollar contracts which have a very low risk, then your leverage is going to be significantly higher.

Obviously, a black swan event in crypto will most wipe out half of your capital if you’re structured like that. A black swan event in Eurodollar, let’s just say for the sake of argument that you are leveraged five times, then a 20% drop would wipe out all of your capital. So that’s what he’s kind of inferring. If the black swan is really bad, then it’s going to destroy your capital and the size of black swan required will be smaller than the amount of leverage you have. That’s just maths.

But then we need to say well, how likely is it? He says, how likely is it that a Eurodollar contract could go down by 50%? How likely is it that crypto could go down again 50%? Well, clearly, at least empirically, crypto going down by 50% does happen, and for some of the dodgier coins, it’s probably happened within minutes. It’s never happened for Bitcoin or Ethereum, but there are crypto coins where that’s happened. Has a major interest rate future ever gone down by 50% in less than a day? No. Could it happen? It’s extremely unlikely, but not impossible.

[00:34:17] Niels

Can I interrupt you here, Rob, because I think this is interesting. To some extent, this touches on this ongoing debate, static position sizing, dynamic position sizing. Just to be sure that Carl understands here that if you have extremely low market volatility for a while, such as we did have in short-term interest rate contracts for many years actually, if you don’t take this into account, you can end up with very large positions notionally.

The markets don’t have to move 50% for you to be blown out, because if you have huge positions, and if there’s one contract that literally can move dramatically overnight, that is short term interest rates, because it’s like the Fed is either going to be 50 basis points or 250 basis points, you have no control and once the market believes that it’s one or the other, it will go to that price pretty quickly. So you have to be incredibly cautious is just what I wanted to say. I think this is also why those who do advocate for using static position sizing, they don’t use the total account size as their AUM. They use closed equity.

So they have a bit of a buffer, to some extent, because they’re basing their calculations on a lower amount than your total account size, etc. So I just wanted to put that in there, Carl, so you fully understand that not everything you read, even if it comes from Andreas, is something you can just use without thinking about it and actually understanding the risks.

[00:36:09] Rob

Yeah. That figure of, what was it, 13 times leverage or something like that, that would not be atypical for a fund that has interest rate futures to things like 2-year bond futures, and so on and so forth. Now, actually, I think there’s probably six pages in my new book about this specific problem, so I’ll just throw that in there. But very briefly, there are a few things you can do if you are concerned about leverage.

I just actually looked at my own leverage. It’s currently running at 1.75 across my whole portfolio. There’s a few things you can do. Specifically for Eurodollar futures, and actually, this applies also to things like VIX, don’t trade the front contract, trade further out. Because generally speaking, the fat tail behavior if you like is less likely, because, as you say, if the Fed is going to change interest rates next week, then the front Eurodollar could move a lot.

[00:37:05] Niels

By the way, you make a good point because I’ve never come across a CTA that trades the front contract in short-term interest rates. It’s usually nine months or beyond and there’s a good reason for that.

[00:37:15] Rob

I actually target 3. I go out as far as I can once they’re liquid, which for me is about three years. A big CTA would need to be closer because they need a bigger position size. The other thing you can do, in a more general sense, is actually if you are worried about this, if you exclude instruments that have relatively low volatility from your portfolio, your leverage will come down. Obviously, there’s a cost to that if you like because you’re losing the diversification benefit. Because my system has this fancy dynamic optimization, I can, to an extent, overcome that.

So what that will mean is, for example, if I’m not trading the US 2-year because its risk is too low, and that would result in me getting too much leverage, I can still trade the 5-year and my system will automatically put risk into the 5-year. I actually have a report that I run that tells me, and it actually says on the report, markets that are too safe to trade. Because my leverage is quite conservative, my annualized standard deviation there is about 5%.

Actually, I don’t trade for example, the bubble, the German 5 year, I don’t trade the 3-year BTP, I don’t trade Eurodollar, I don’t trade your Euribor or the Fed funds or JGBs or Korean 3 years or Shatz or US 2 years or US 3 years or US 5 years or US 2 and 5-year swaps. So there’s actually quite a lot of fixed income that I don’t actually put positions in although I do use them as part of my forecasting calculations for my fancy model.

So if you are generally really, really concerned about this leverage issue, and I am, actually, which is why I put in this thing where I remove markets with very low risk, that is a way of doing it. Just be aware that you lose pretty much the whole of the front end of fixed income, which for me personally, I’m happy doing it because an ex fixed income trader, I’m very much aware of how, as we said at the start of the podcast, the boring asset of bonds, when it goes, it goes.

[00:39:14] Niels

Yeah. Absolutely. I hope, Carl, that those answers were helpful. I know you had a few other comments but I think we’ve covered most of it. But I do want to ask you a question, Rob, that just springs to mind since you are the backtesting king as we just established. If you were doing a backtest and you wanted to apply more or less a universal commission rate, slippage and commission rate for your backtest to keep it simple.

We like Andreas, we like his work, but I did notice that in his book, he mentioned something about that I do this backtest and I’m going to apply, say I think it was $2.5 commission because things are becoming cheaper so I can, and so on and so forth. Then he applies this and creates this backtest. But I’m thinking, well hang on, you’re not really taking into account the fact that 30 years ago, it didn’t cost $2.5. So I thought that was not quite how I would do it. So I just wanted to hear your view as to how you take into account the fact that probably both slippage and commissions change over time.

[00:40:25] Rob

Actually, I do use exact precise commission figures for the present day and Andreas probably does in his own work for his own firm. It’s only obviously in a book that you want to make things simple for people, and that’s completely fine. I think that’s a perfectly fine approach as long as you explain what you’re doing. Because there are some futures that have less for a retail trader where you can pay less than $2, and there are others where you pay more than $2.

[00:40:50] Niels

Yeah, but also historically, it would certainly have been more expensive. .

[00:40:53] Rob

The historical question is obviously what you’re asking me. What I actually do is I basically take those present-day figures and I adjust them for volatility in the past. I do it with dollar volatility. If you go back far enough to when the S&P was at 400 rather than about 4000, I would basically assume that the dollar cost was one-tenth of what it is now, assuming the volatility is the same. This actually is clearly a simplification as well. What I’m basically doing is assuming that the risk adjusted cost in the past is the same as it is now.

Now, that is a simplification. It will indeed lead to almost certainly me underestimating what costs were historically. That’s fine, I think, if you know you’re doing it. Because the way I see it (and this is a misunderstood kind of thing) a backtest serves different functions. It serves as a test of what happened in the past. For that, you’d need to know what the commission was to trade S&P futures in 1982. If you’ve got that information obviously, then use it. Use that instead. But you’d have to be like an AHL that’s been trading since 1987 or a DUNN that’s been trading even longer and to have been diligently storing that information every day in your database to have that information.

But most retail traders don’t have access to that level. It would be amazing to know what the slippage was on the 2nd of January 1982 and how much commission I’d have paid as a retail trader. I don’t have that so I have to make an assumption. By doing that assumption, I’m basically thinking about the other purpose of a backtest, which is telling me is this a good thing to do in the future. So actually, I don’t necessarily care what the trading costs were in 1982 because I’m not trading in 1982, I’m trading now.

So for the purposes of working out whether I should trade something now, what really matters is the current risk-adjusted cost, whether that’s too high or too low. The only caveat of caution, of course, is this. This relates back to something I said earlier about, it looks like, for example, fast equity trend following made money up till about 1990, and then it didn’t. What could be happening there was that my backtest costs obviously don’t know that equity trading costs have been coming down over time.

If I’d actually looked at the net profits using accurate historical trading costs, I probably would have said, “You know what? This has never worked.” It just looked like it worked pre-1990 because you were underestimating what the trading costs really were back then. If you actually had access to real trading costs, you’d have seen that the apparent extra money you were earning was actually eaten up entirely by costs, and then once the markets became cheaper and efficient in the 1990s, the effect apparently goes away when, in practice, it was never there but it’s just that people couldn’t exploit it because the costs were higher than you thought they were. So the one thing I’d say is to be really careful whether you’re using Andreas’ system of fixed commission or my slightly more complex system of consistent risk-adjusted costs. If you see something that’s made a lot of money in the past and hasn’t done recently, well, probably stay away from because that’s probably an illusion that you’re seeing and not a real thing you could have exploited and made profits from.

[00:44:18] Niels

All right. We just today, this morning, received a few more questions from Matthijs. Matthijs writes, “Thank you again to you and your team for producing this wonderful podcast. This is always such a joy to listen to on my commute to and from work and while I’m working out. I have three quite different questions that I’d love Rob to answer or actually to ask Rob given his wealth of knowledge in running his own portfolio and seeing a very large institutional manager run theirs. Depending on the direction the episode goes and how it fits, I would be grateful for you to answer them.”

All right, quick one. Operational matter: “What is in your experience,” that is your experience, Rob, “the optimal moment to roll a contract? How much liquidity do you ideally want in the current versus the forward contract? Equal liquidity in both, half liquidity in the forward, one quarter, etc.? What are your thoughts?”

[00:45:18] Rob

Again, this is something I’ve probably written 20, 25 pages on in my new book so I could go on at length about this.

[00:45:25] Niels

Well, I don’t want you to do that because we do have your topic.

[00:45:29] Rob

Exactly. Yeah. Just to answer the question really simply, basically, I will usually roll as early as I can for various reasons, which you’ll have to buy my book to find out. For example, if there’s 5% liquidity in a new contract but that 5% is thousands and thousands of contracts because it’s the S&P, I’m happy to roll down with just 5%. On the other hand, if it’s something that’s relatively illiquid and it only does 300 contracts a day anyway, then yeah, I’d probably wait till it was 75% of the way there.

So it’s not a fixed percentage for me. I like to roll as early as I can as long as there’s sufficient liquidity in the next contract and sufficient depends on what the contract is.

[00:46:14] Niels

That makes sense. Trading strategies: “What are your feelings about adding a defensive betting against volatility strategy to a multi-strat CTA to diversify existing trend following, carry and negative skew seeking strategies, that is a strategy that allocates more to a market if it has low volatility either compared to its own past or compared to other assets. Such a strategy is well-documented as working in equities and bonds. In my experience, it also works in commodities, but not in currencies, strangely. The reason it might work is because of limits to leverage slash leverage aversion. I’m not sure why that reason wouldn’t apply in currencies.” Any quick views on this.

[00:46:57] Rob

Basically, the reasons why it will work is because of leverage. So for example, if you just take bonds as an example, if you want a certain amount of return, you wouldn’t necessarily invest in a high Sharpe ratio asset if it’s got low volatility. So you wouldn’t necessarily invest in 2-year bonds, you’d want to invest in 10 year bonds instead. Therefore, it makes sense if you have access to leverage as we do as CTAs, as futures traders, it would make sense then to buy 2 years and short 10 years, for example.

I haven’t looked at it myself but there is literature at work in the financial markets. I like these kinds of risk premia/risk aversion type strategies where basically you can identify who’s on the other side of the trade. There’s a good reason why it should work. But yeah, as a strategy that’s going to have negative skew and high leverage by definition, it probably wouldn’t be a big part of my portfolio. But yeah, worth looking at. Definitely.

[00:46:58] Niels

Okay. Final question and then we’ll go on to your topics. This is about portfolio construction. “Why limit trend following to a small part of your portfolio, say, 20%, when it has a much higher historic risk-adjusted return and more positive skew than most other investments? Is it because it involves leverage and therefore can go to zero while a diversified stock slash bond portfolio will not do that?”

[00:48:17] Rob

There’s a kind of theoretical thing I assigned to this, which I think I discussed actually on my last appearance, maybe on the podcast or the time before. We were looking at the ultimate correlation to trend following. In theory, it was much higher than the 20%. Based on some assumptions, it could be anywhere between 50% and 80%. So theoretically, at least, you’d probably never have 100% in trend following. That’s the first thing to say.

Most people have less than 50% or 80%. I, myself have less than that. For me, personally, it’s down to a combination of tax reasons, psychological reasons. Even though I know that I get a higher return from putting a lot more of my money into trend following, I always have this kind of view that all the money in my futures account could disappear entirely. If that happened, I’d still want to be able to pay the bills. So that’s kind of the reason why I personally don’t do it.

But yeah, other people may have different reasons. I think a more pressing point is why the average institution that doesn’t have to worry about that sort of thing has probably nowhere near 20%. It should be least 20%.

[00:49:29] Niels

That is the question for sure.

Anyways, we are going to move on to some exciting topics, I have to say, and the first one comes from a new article that you wrote that was published in the FT. So congratulations on that. I’ll let you talk about what it’s about. But of course, one of the things that really stood out to me was it seems like you spent the day reading Zero Hedge? What’s that all about?

[00:50:04] Rob

Actually, this is kind of interesting because obviously last week, Alan talked about the footprint of trend followers, but I hadn’t actually gotten around to listening to that episode before I wrote this article. So this was completely independent. Alan was talking about how various sell-side researchers like to put out pieces saying things like, “Oh, yes, if the market does X, then trend followers will sell $50 billion of stocks and we’re all going to die,” and all this kind of stuff. It’s all very full of this. On a similar vein, a friend of mine, not me (I don’t actually have a subscription to Zero Hedge, a friend of mine sent me this Zero Hedge article along similar lines, and they’d taken some Goldman Sachs research. You and Alan were discussing last week how the sell-side research is kind of a bit dodgy. Now imagine that, but with Zero Hedge covering it. You can imagine what this piece is like. So yeah, the massive hyperbole saying, “Oh, yes. If the S&P reaches these key pivot levels, then CTAs will dump $220 billion on the market,” this kind of stuff.

Anyway, obviously, Alan and you spoke about this a lot last week, and there’s this excellent piece by Quantica as well. But what I did was just take one specific claim, this $220 billion claim, and basically just with some simple maths said, first of all, how likely is it that that will happen? Secondly, what percentage of volume will that be? So would it actually be a significant event? What I basically did was a single use case of the Quantica methodology effectively, if you like.

What I did was say, first of all, we have to start off with the size of the CTA industry, and you guys talked about that last week. I used $300 billion and that was kind of roughly the figure that Alan was using as well. We can argue about whether it should be half of that or a bit more than that. It seems reasonable. The other thing that I talked about briefly was just this idea generally of how you actually do this kind of modeling and where you can get the information from so you can obviously just build your own simple model and then you could potentially combine that with public information, things like trying to extract from the commitment of traders report, for example, what CTA positions might be.

Obviously, funds publish reports as well. If it’s a UCITS fund, then it’ll have some information in it. So pulling that information in to build your model of flow and position sizing. Then what I said was, okay, basically, what the article is saying is that you’d be going from long 100 roughly to short 100. That’s where the $200 billion sell off comes. By the way, one of the really naughty things about the Zero Hedge article was that they deliberately confounded the S&P 500 and the entire global equity market. The Goldman Sachs piece didn’t do that, but they kind of reframed it so it made it sound like it was $220 billion just of S&P. If you read the small print of the Goldman Sachs article, it’s only $44 billion of S&P.

Another reason I actually worked that out was I said, actually, what you can do is you can work out firstly, roughly what the amount of leverage would be. Imagine if the whole industry was one $300 billion fund and they were trading S&P, you can work out, roughly speaking, what sort of position they should have in the S&P. The first thing you need to know is how much leverage they’ll typically be applying for that position. That will come from two main sources.

Firstly, it will come from the difference between the fund’s target risk and the risk of the instrument you’re trading. So this is exactly about the conversation about leverage we just had, right? The S&P’s risk of 15%, 20% is actually pretty close to what the risk would be of a particular CTA. I think you guys were talking about a 15% average risk again last week, so I actually was really conservative or aggressive if you like and I said, actually, I’m going to assume that CTA risk is much higher at 25% and I’m going to assume that equity risk is actually relatively low, like 15%.

The other source of leverage is the thing we talked about earlier, this diversification multiplier. So if you were only trading S&P, you’d have a leverage between 1 and 1.5. If you’re trading a whole bunch of other things as well, again, I was really aggressive and said, let’s assume you’ve got a leverage that goes up to 4 without additional diversification. So I said if that’s true, and there was that $100 billion, that would mean that the CTA industry as a whole would have 8% of their capital allocated to the S&P.

Now, given that not everyone is Jerry and has hundreds of instruments or me with 150 going on 200, but almost no one’s going to be trading that concentrated of a position in the S&P. They’re just not going to be doing that. Then I said actually the real figure is $44 billion, which means that the S&P risk weight would be 2%, which actually sounds a lot more plausible. I can believe that 2% of risk assets across CTAs are in the S&P. That, to me, sounds about the right figure to be honest. Because if you say they’ve probably got 20%, 25% of their money in stocks, they’ve probably got maybe a quarter of it in US stocks and that’s probably split between Russell and NASDAQ S&P futures.

So yeah, it seems plausible. Then what I did was, and again, this is something that you guys said last week, was actually, if you’re vol scaling (and obviously, famously, not everyone does), but if you are vol scaling, it’s very unlikely that the 100 position on the long side would be the same as on the short side. Obviously, the market does sell off by I think it’s 2.5 standard deviations in the Goldman calculations. Do we really think that equity risk is going to stay exactly the same? It just seems extremely unlikely that we’ll have that kind of very gentle glide path downwards without that happening.

So anyway, I basically said, the $44 billion figure is probably about right in terms of position sizing, it’s probably a bit aggressive in terms of the selling. So more realistically, I think $30, $35 billion. But I said let’s just assume that the $44 billion is correct, that they’ve successfully managed to model the whole CTA industry despite making quite a few fundamental errors. What actually does that mean? This is over a month, so $44 billion sounds like a lot of money. If you gave me $44 billion, Niels, now I’d be very appreciative of that. I’ll just put that out there.

To be honest, 44 million would make me smile. As everyone knows, I think that none of us get paid for this gig. So I think I’m owed some back pay. Anyway, we’ll just leave that there. But, yeah, so I said, actually, $44 billion over a month of S&P futures trading comes out to 0.6%. Spookily, that’s exactly the same number that Quantica had last week. The point, of course, is that 0.6% of volume, it’s nothing. You’re just going to be lost in the wind. So yeah, that’s a nice exercise you could do whenever you see any one of these claims, is to sort of do this very simple maths and say well, actually, that’s not really significant.

[00:57:31] Niels

It’s kind of funny to me that first of all, you were talking about an article where Goldman Sachs was doing this exercise and spooking people. We obviously were talking about other reports, usually Nomura, Morgan Stanley, JP Morgan, I think were in that article as well. So all the big houses coming out, which of course, I think people know already, I completely disagree with this kind of information being presented like that, not least, because it’s kind of scare mongering, I think, to some extent, but also, it’s factually wrong.

But it’s funny, you never hear them talk about their own asset management departments which manage much more money than the CTAs do. So it’s really disheartening to see this kind of stuff, and then even more so that it’s just being taken by the journalists uncritically and published on the news. Speaking of which, I can’t really say that this is in the same category, but I had picked this up and you also picked this up, and that is that people love to make forecasts about which strategies will do best in a particular year as if they know, psychic or can tell the future.

So we couldn’t help, both of us, noticing that the French bank BNP has come out with its predictions for hedge fund strategies for 2023. I’m not going to go through all of them. All I can say is that they believe that multi strategies will be the winner this year and CTAs will be the losers, and they think that CTAs will probably end up losing 9%. Now, of course, they could be right. We don’t know. There’s quite a long way until the end of this year. But it is a bit silly, in my opinion, to make these predictions, especially about strategies that are so adaptable like CTAs, because what on earth would they base these assumptions on? So I don’t know if you have anything positive to say about this.

[00:59:40] Rob

To be fair to the French bank, it’s basically based, I believe, on a survey of investors, so it’s not that they personally have come up with these figures.

[00:59:49] Niels

Fair enough.

[00:59:52] Rob

A more interesting question is people’s beliefs about performance, and in particular, whether it is likely to continue or mean revert, because that’s kind of all this is saying. People are thinking CTAs did really well last year, well, that’s bound to mean revert. So we’ll say they’re going to do the worst this year. How did multi-strat do last year? Actually, I thought they did okay.

[01:00:16] Niels

I think they always do okay. It’s like this magic box where you put markets in and you fire those who don’t make money, and then you end up…

[01:00:24] Rob

Interestingly, they believe that multi-strat, which did well last year I think, will also do well this year. So who knows what’s going on there? But yeah.

[01:00:33] Niels

Yeah, I’m glad you made the correction, because I actually had not read the headline of what I’m looking at right now where it does say investors predict so I apologize to BNP. But anyways, it’s their survey and of course it’s based on other people’s opinion but that is a very important point to bring up.

Now, we only have a few more minutes left but we do have a few, maybe one article or so you want to talk about that you think is the most relevant one because you put out some new blog posts both about diversification, you did one on standard deviation and you also did something about our friends over at the Bank of Japan and the struggles they have at the moment.

So is there one of these three that you want to briefly talk about? Because people should go and read them on your blog post. That’s a given. But anything you want to say?

[01:01:27] Rob

We’ve touched on leverage actually, and I said earlier during that conversation that there was a concept that I will come back to. So let me let me do that, otherwise, I won’t come back to that. Diversification across instruments is one of the most powerful things in finance, and particularly in trend following, because actually, the expected performance of any individual instrument is relatively low and almost close to being not statistically significant, but the benefit you get from combining them is enormous. It’s really big.

I spoke earlier about this concept of an instrument diversification multiplier. You can think about that effectively, because what that’s telling you is your reduction in risk. But if you think about Sharpe ratios, if your risk is reducing and you’re then leveraging your portfolio to have the same risk, you’re basically multiplying all of your returns by that number. So it’s also a measure of expected improvement in return through diversification. Another way of thinking about this which some people find more intuitive, which is to think about the number of independent bets in your portfolio.

So if you have two instruments in your portfolio with zero correlation, then the diversification benefit will be the square root of 2 because these things go in square roots. But you’ve got two independent bets in your portfolio. Now, you may think, well, let’s just keep adding things, and so on and so forth. But actually, you’ll start to struggle to find things that have got zero correlation quite quickly, only things that are 0.05 correlation and 0.1, and so on and so forth. What that means is that big portfolios of assets with trading strategies typically will have a number of independent bets that’s less than the number of assets in the portfolio effectively.

I took a long-only portfolio and basically looked at the number of independent bets in that, and that came out at 23. That’s with the 100 assets. That’s actually quite a lot because the diversification benefit of that is going to be the square root of 23, which is just under 5, so you can multiply your returns by 5. Now, that’s kind of the expected benefit from diversification. But what I then did, in this blog post, was compare that with what you actually get basically, and in the long-only portfolio, what you actually got was pretty much that your Sharpe ratio doubled.

So adding an awful lot of instruments to a portfolio in theory looks great because the correlation is relatively low. But actually, when you actually say how much return to actually get from this? The answer is not as much as you might hope. The interesting thing then is if you take that to trend following, you actually find that the expected diversification benefit is pretty similar. So it comes out something around 20, 25 independent bets. But the good news is that the actual real performance is actually higher than that. So you actually end up increasing your Sharpe by between four and five times, which is incredible.

To get that sort of benefit from adding instruments to your portfolio is just incredible. So it’s clear the trend following, it’s a combination of the fact that this is all using correlation, which is a linear measure, we have nonlinear returns, we have outliers and so on that we’re trying to capture. The fact that when you’re adding one of those long-only assets, the correlation looks low but actually, in practice, there’s only a small number of factors driving everything so there’s not really as many independent things going on as you think about.

Going back to the first thing I said about factor one and factor two, they kind of explain most of your returns. So in some ways, if that was true, you only have two independent bets in your portfolio. In practice, there’s probably about four in long-only portfolio. But yeah ,in a trend following portfolio, the benefit is much, much higher. So it’s not news to say that you should add as many instruments to your portfolio as possible and get that diversification benefit, but I just thought it would be interesting to kind of reframe that in a different way and present some new evidence to try and get that point across.

[01:05:50] Niels

Yeah. I think generally speaking, that’s also the takeaway from the CTA series. Most managers tend to buy into that. But there have been a couple of managers who don’t. We’ve been around for a long time and perform quite well. So it is interesting that one thing is theory and then looking at the numbers, maybe not so much. But it is a topic that I’m sure we will continue to monitor.

[01:06:23] Rob

Of course, this also relates back to the quite recent conversation that you had about hedge fund replication as well. So the idea that certain ETFs can replicate this huge universe of futures with just a few that they’ve chosen.

[01:06:38] Niels

Which they’re not quite doing this year. Interestingly enough, there’s a big return dispersion between that replicator and the index it’s replicating, just to throw that in there. Anyways, that’s a completely different discussion.

Rob, this was wonderful, delightful as always, and so if you agree with me, then can I suggest maybe you go and leave a rating and review for the show on Amazon, Spotify, Apple, wherever you listened to it. It might only take you five minutes, but it really does mean a lot for the show and for more people discovering our conversations every week. Of course, you should also continue to send in your questions to us. info@totradersunplugged.com is the email.

Next week, I think I’m joined by Rich. For some reason, I’m a little bit in doubt now that I think about it. but I think it could be Rich next week, it could be someone else. I’ll have to be a little bit uncertain about that. But in any event, it will be a great one. So do send us your questions ahead of time. As always, follow us on Twitter. See if you can pick up some information on that as well.

From Rob and me, thanks ever so much for listening. We look forward to being back with you next week. In the meantime, as always, take care of yourself and take care of each other.

[01:07:58] 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.