217 Systematic Investor Series ft. Rob Carver – November 6th, 2022

[00:00:00] Rob

You’re about to join Niels Kaastrup-Larsen on a raw and honest journey into the world of systematic investing, and learn about the most dependable and consistent, yet often overlooked investment strategy. Welcome to the Systematic Investor Series.

[00:00:23] Niels

Welcome, and welcome back to this week’s edition of the Systematic Investors Series with Rob Carver and I, Niels Kaastrup-Larsen, where each week we take the pulse of the global markets through the lens of a rules based investor. If you’re new to the show, I hope that today’s episode will trigger your curiosity enough to check out the back catalog and listen to past episodes that you may have missed. Like my conversation with Alan last week where we discussed how to approach asset allocation when adding trend following to your portfolio, how to find the optimal allocation, and why trend following is often the most consistent part of the portfolio, a topic that we may return to today. 

Also, if you missed our Wednesday episode, I really would encourage you to listen to the conversation with William White, economist, policy advisor, and policymaker who worked at the OECD, the BIS, the Bank of Canada, and who is perhaps best known for warning the central banks about their failed policies after the Great Financial Crisis. This was a really important conversation on monetary policy and how it can negatively impact the economy. 

I would really encourage you to go and check out these episodes, if you missed them this week, but of course not until you’re finished listening to Rob and I today. Rob, always a pleasure to be back with you. How are things with you? How’s the UK government? Is this still there, or…? 

[00:01:38] Rob

Last time I checked, but I haven’t looked at the internet for about 35 minutes, the Prime Minister could be different now. You never know. 

It feels like we’ve gone from summer to winter in about a week here. We’ve gone from way above average temperatures to wet and cold, sort of normal weather for this time of year. The viewers can’t see this, but Niels can. I’m all bundled up in my ski jacket and keeping nice and warm and toasty in this seasonably cold weather. 

[00:02:09] Niels

You are indeed. Actually (we say it a little bit like jokingly), but I think this year it really has mattered, the fact that we had a very mild October. It helped a lot of people save a lot of money, which is, of course, great. Let’s hope it doesn’t get too cold. Although we obviously have no idea if that’s going to happen. By the way, just to set my curiosity a little bit at ease, you know that thing that they did in the UK, whether Liz Truss would last longer than a bowl of lettuce or something like that. Did the lettuce win? 

[00:02:47] Rob

The lettuce won. 

[00:02:49] Niels

There we are. Now I can relax and enjoy the rest of my weekend. 

[00:02:55] Rob

I’m glad, I’m glad, Neil. I’m very glad. 

[00:02:57] Niels

Okay, of course it was another eventful week. Let’s put it that way. Chairman Powell made it pretty clear, in his Wednesday press conference, that the ultimate level of interest rates will be higher than previously expected, and for longer, too. The future’s markets now see Fed Funds staying above 5% for all of 2023. That’s a big change. As recently as in September, the market expected Fed Funds to U-turn lower, before it got to even 4%. Now it’s reasonable to think it’s going over 5% and staying there. Maybe, it was also the week where we learned that the stock market might be fine with that, although, to think that the market can also shake off disappointing take earnings, as many analysts expect the ‘tech wreck’ of 2022 to be transitory (much like inflation was supposed to be) may prove costly. 

Let’s go back to the beginning. It all started with the Fed’s well publicized leak hinting that the Central Bank would raise the Fed Funds rate by 75 basis points this week, but that another hike of equal magnitude, in the December meeting, was not as certain proved, at least, to be partially correct. The committee did raise by 75 basis points and with it offered a new sentence to the statement, “In determining the pace of future increases in the target range, the committee will take into account the cumulative tightening of monetary policy.”

Perhaps this was a written acknowledgement that the committee realizes that they have already tightened aggressively, and importantly, policy change works with the lag. However, Chairman Powell’s tone 30 minutes later, at the post meeting press conference, was decidedly hawkish. Bond traders immediately took rates higher.

May, 2023 Fed Funds futures rallied to 4.805% on the day of the leak, but have since reversed and are closing out the week at 5.12%. Similarly, the two year note, which traded down to a recent low at 4.3%, reversed violently and finished the week at roughly 4.71%, which is the high of the year.

Perhaps Powell has been handed an advanced copy of the Friday morning’s employment report as he was headed down the hall to the press conference because it showed no sign of a moderating economy. The report showed 233,000 jobs were added to the economy, outpacing the 200,000 that was expected.

Hourly earnings also rose last month, which is certainly a worry for the Fed, especially since the most overlooked Jolts survey of unfilled job openings unexpectedly jumped by 1 million rising to 10.7 million unfilled jobs. Granted, rising interest rates are crushing the auto and the home sectors, but the robust demand for workers and the rise in wages seems to be more than offsetting that effect.

Rob, it’s been another eventful period since we last spoke. What’s been catching your sort of attention in the world of finance? 

[00:06:11] Rob

Yeah, I’ve got this report that tells me (I talked about it before), that classifies market movements over different periods based upon risk adjusted terms, given how big they are. Interestingly, over the last seven days, the three markets that lost the most money, they’re all related to US technology stocks. I guess that tells us a little story there. 

Yeah, it’s been about six weeks since I have last been on. PNL wise I had a pretty good time after I was on, actually, at the back end of September hit a high watermark, then lost a bit of money, made a bit of money, then going sideways. I’m up pretty much at the level I was back in May. I’ve gone sideways but a lot of up and down since then. It’s very much still a very clear picture that the first few months of this year were actually incredible, and then I’ve just flatlined ever since. Which, I think, comparing to a long-only portfolio, is the sort of performance that you definitely take, I think, absolutely. If I look at my year here, year to date, just to filter out the noise really then it’s still a pretty similar picture in the sense that obviously I’m up about 26% I think.

And my biggest gains are still in markets where I gained a lot in the early part of the year. Obviously that’s mainly in energies, but also with a bit in FX and also with a bit in bonds. I’ve actually done well being short bonds, which is unusual. If you look at the back test, that doesn’t happen too often.

Turning from the rear view mirror to the front windscreen, if you like, my current risk is still lowish. It’s not gotten really that high since the beginning of the year. My biggest bets are in bonds and in energies and I’m short, obviously still short bonds, still long energies.

My biggest short, I think it’s unchanged from last time, which is US 20 Years. My biggest long, again maybe the same, is crude, although actually interestingly (and this is unusual), I’m actually short gas. So, I’m long crude, short gas, make of that what you will. Obviously gas prices are a lot more volatile and more driven by seasonality than crude because gas is very hard and expensive to store. It’s very much driven by what’s going on right now. As we were saying earlier, we’ve had a (certainly Europe) an unseasonably warm time. I think gas spot prices have actually gone negative, and certainly in the UK, a few times.

I trade Henry Hub Gas in the US. I guess the US is probably not that dissimilar a story, so, that’s interesting. So yeah, it has been a pretty quiet period and just a question of sitting on some trades that haven’t really done very much. And just, as we do, waiting for the next big move to come and while it doesn’t come, just preserving capital and that’s what the system seems to be doing pretty effectively. 

[00:09:23] Niels

No complaints. No complaints. Yeah , it is quite interesting actually. First of all, a couple of observations.

I think for a  lot of trend followers, certainly if I look at the early numbers that came out at the end of this week for the month of October, there were relatively small pluses and minuses for the industry. Actually, when you look at the underlying market moves that happened during the month of October, there were some pretty sizable moves on all fronts – with equities up 8% or 9%, you had coffee down 20%. You had some pretty sizable moves. I think that actually speaks quite well for how the industry is handling this period. Also, by the way, just speaking of gas, something that I just listened to the other day because there’s been a lot of talk about why gas prices are down and, as you alluded to, it’s been mild, but also we are hearing all the time that yeah, European gas storage is full.

Great. I’m not so sure people are aware that actually the gas storage only makes up about 20% of the gas that Europe uses during a whole winter. Yeah, it’s great that it’s full, but if we can’t get gas on an ongoing basis (enough), then it’s not going to keep everyone warm. 

[00:10:45] Rob

Yeah, interestingly, actually on that subject there was a big controversy because the UK closed this rough field gas storage facility a couple of years ago. And obviously that would’ve been a big help in smoothing out the volatility in energy prices last winter and coming into spring. That has now reopened with, I think, 20% of its original capacity. Yeah, it does seem insane that was closed because it was a classic case of making a short-term saving for the long term cost. OK, it’s costing a few million pounds a year to keep going, but you know what the hell, the potential benefits could be huge if energy prices spiked, and of course they did. That was a very shortsighted decision that looks like it has been reversed now.

[00:11:33] Niels

Another shortsighted decision. Let’s put it that way. 

Anyways, in terms of the week as a whole, it was actually also a pretty wild ride on Wall Street, so to speak. I do think that actually, trend follows, made small gains during the week. Gains still from the short exposure to fixed income, I would imagine, was pretty good, and also long exposure to energies, just as you said, Rob, I think that’s helped. I think that could have been a little bit more challenging. Currencies, although the Mexican peso/British pound probably could have outweighed losses in other currency pairs.

If I look at the soft markets, from a performance point of view, they look to be pretty soft so to speak. In the metals, I think probably trend followers would’ve lost a little bit of there, and grains maybe a more of a mixed picture. All in all, I suspect the industry made a little bit of money this month, although yesterday actually I think was a down day for sure. 

My own trend barometer finished the week at 44, so that is neutral. If we just look at the industry numbers BTOP50 (and this is as of Thursday, so without Friday’s down day) as of Thursday, up 34 basis points for November, up 20.36% for the year. SocGen CTA Index up 63 basis points for November, up 27.5% for the year. SocGen Trend up 38 basis points, up 36.36% for the year, and the Short Term Traders Index down 23 basis points, up 12.31% for the year. Comparing that with MSCI down 1.6% so far in November, down 22.4% for the year. The World Government Bond Index is off to another rocky start, down 68 basis points in November, and the S&P 500 down 2.62% so far this month, down almost 21% for the year, despite a very strong October, of course. 

All right, Rob, you brought along some great topics that we want to dive into, but we have quite a few questions as usual when you’re on. Before we can go into any of those, I have to ask you, how’s the book coming?

[00:13:41] Rob

The book is basically done as far as my bit is concerned. I’ve just finished the proofreading of the final print copy that then has to go to the printers, which is going to take, apparently is going to take a while, I guess plan and supply chain issues or something I’m not exactly sitting back and relaxing because I have some work to do in terms of I put a together a website, for people who are going to buy the book, that has spreadsheets and Python code. So, I’m building that side of things, but the actual content is done. The feedback I’ve had from the people who’ve been reading it for me and checking copy and stuff has been pretty positive. So, I’m very much looking forward to that coming out. Maybe we’ll have a longer chat about it when it comes out. 

[00:14:27] Niels

Oh, I have a feeling we will, maybe not just one, but it’ll be fine for sure.

Okay. All right, let’s get into it. The first question came from Simon. The question is simply, “What is the shortest term that Rob has looked at in trend following/momentum trading? There are obviously a lot of breakout traders who can look for sharpe breaks and hold for quite a short period of time.

I appreciate the liquidity constraints for very large funds, but for smaller managers I imagine there would be some interesting advantages there. For example, the higher trade count with corresponding smoother equity curve and larger sample size for back tests. Perhaps this is a style that the intra day future traders are more focused on, given there might be too few opportunities on an end of day basis. 

For stock traders, there’s obviously quite a large universe to be looking for these kinds of opportunities in. I don’t suppose anyone has looked at that. There is no shortage of interviews with discretionary stock traders looking for these kinds of opportunities. However, I’d really love to know what’s being done on the systematic/quantitative side of this space. Thanks so much for any insights and keep up the good work. Love the podcast, and wishing you all well from Australia.” 

[00:15:54] Rob

Interestingly, this is actually something I talk about in my new book. There’s this stylized fact that trend following works really well for time periods of between about a month and about a year, very roughly. If you go beyond that, so if you go to multiple years, you’re looking at some value short mean reversion strategy performing best, like the classic equity value market neutral strategy. If you go shorter than that, then you start to see the profitability of trend following break down.

In particular, if you look at the momentum crossovers that form the core of my system, the faster ones, of those, aren’t as profitable as the ones that have a holding period of at least a month, for example. I have this theory that different effects play out in different time periods.

So, you have momentum between a month and a year, beyond a year you’ve got mean reversion. Below a month probably mean reversion works well, and actually, in the book, I do backtest a mean reversion system with a holding period of a few days. That does really well, although you have to be really super hot on execution because otherwise the transaction cost will absolutely kill you and you cannot execute it the same way that you can execute a trend following strategy, and you have to use limit orders effectively to do that. Now, it’s quite likely that if you go down an even tighter timeframe, holding periods of maybe a day or a few hours, it’s quite likely that trends will reassert themselves so you start to see trend following behavior again.

If we go right down to sub second level, then the kind if trading that market makers and high frequency traders do is very much mean reversion. They’re very much trying to buy at the bid and sell at the offer. It kind of makes sense then, in terms of my theory of these different patterns of price behavior at different time periods.

Now the question is this few hour, single or two day momentum period effect, is it something that exists and is it something we can capture? And there has actually been some work done by various people. The main people who’ve done quite a lot on this is the quant team at SocGen who’ve looked at this sort of intra day momentum effect.

They found that there is something there. They also have theories that it’s related to things like option dealers hedging, and hedging at specific times of the day, for example, potentially. Although, we are generally, on this podcast, certainly not that bothered about an explanation for an effect as long as it’s there in the data and it’s significant enough, that’s fine.

Now the issue of course is, as you trade faster and faster (as I’ve already alluded to) in this mean reversion system is transaction costs. You cannot use limit orders when you’re trading a fast trend following strategy. If you think about what you’re trying to do you basically think that over the next few hours the price is going to go up. You really want to buy now, at this point because if you wait half an hour, well, your expectation is that in the next half an hour, the price is going to be increasing. You’ve either got to set yourself up in a situation where you are missing a lot of trades because you’re not getting the execution done and then the thing just moves away from you. Or, alternatively, where you have to cross the spread every time and do that trade at the market level and then close out, again, in a few hours time. Transaction cost-wise, that will absolutely kill you. Even in futures (which is a relatively cheap asset class to trade) there’s probably only about maybe a dozen futures that are liquid enough to trade this kind of strategy and still survive the transaction costs.

It’s not even really to do with size. My account is not that big, but even for me the transaction costs when trading a single contract would just be too big to overcome.

In conclusion, I would say there is probably something there. It’s really difficult to capture that effect. And I’d say it’s a playground that you should only play in if your execution abilities are absolutely on point and your back testing is really super, super good so that you can be confident there really is something there that’s worth capturing. You need to be really careful about which markets you trade as well. And finally, as the question says, it’s probably not something that a $500 million fund even can do successfully. You need to be much smaller than that, so it has, probably, a very limited capacity. 

I generally would say for 99% of the people listening to this podcast, which includes me, don’t do it. So that’s the short answer. 

[00:20:44] Niels

Okay, so let me just try and make sure I fully understand everything you said here. When you talk about ‘timeframe’ you were talking about holding periods. I normally think about it as lookback periods. Can we translate some of that? The reason I say that is we actually do some analysis where we take a classical (and I don’t mean to say necessarily in a particular style), let’s just call it a trend following approach that I think is pretty universal. We then go back and we look at that over different lookback periods. We go all the way down to about 20 days and we go all the way up to about 300 days to see if there is some pattern as to which lookback period would be more consistent. We do this more for visualization purposes, not how we trade, I should say. 

What you see on this, if you go back (we go back to the year 2000), is that there is a clear gravitation between these annual observations. Essentially, we’re looking just to see how the return is, based on sharpe, distributed across all these lookback periods for each calendar year. Then, we identify which one was the best, just for illustration purposes. If you look at which one was the best, I would say (and this is a little bit from memory), but I would say probably, of the 22 year observations, I would say about 15 or so probably is around the 200 day mark. plus/minus say, 25, 30 days.

Then you have a couple which were quite short. For example, interesting enough, 2008, it was 50 days, but all of the time periods were profitable so it didn’t really matter whether you were trading 50 days or whether using 200 days. It did not make a difference. It’s just that 50 days was the perfect pick.

Then you have a few other ones that are probably 75 to a hundred days. Anyways, there seems to be a clear gravitation towards longer term timeframes in terms of lookback period. I think that’s what you are confirming as well, because 200 trading days is about one year.

[00:23:06] Rob

Yeah, I mean there’ll be a relationship between lookback period and average holding period for a given type of trading rule, or time series filter (if you want to get technical) for the moving average crossovers that I use. If I look at 16 and 64 days crossover with exponential weighting, that will trade once a month.

[00:23:35] Niels

Once a month. That’s a little bit of a surprised. 

[00:23:37] Rob

The lookback period of that is going to be somewhere between 16 and 64 days. So, it’s going to be about 50 business days, which is about two months. So, you’re looking at, for this kind of moving average stuff, a two month lookback corresponds roughly to one month holding period.

If I talk about the holding period of the mean reversion system we’re talking about, that’s got a holding period of about two, maybe three days. The lookback for that is about a week, roughly. If we’re going right down to this very short timeframe, to less than a business day of holding period, then the lookback is probably going to be a couple of days. very roughly. It’s well beyond the point where you are doing your analysis. It’s well beyond the point where I trade momentum as well, because it’s just too damn expensive, and because it stops working for time periods of less than about a month. It may start working again for time periods that are much shorter than that. The trading costs are very hard to overcome and I just stay away from it.

[00:24:47] Niels

Absolutely. The other thing I wanted to just comment on is that I think when we get down to the shorter-term timeframes, I’m not so sure I would call it trend following. The reason I say that is I don’t think you can actually make money if you had to wait for the trend to turn to get out. I actually think that the people who do short-term trading have other ways of getting out, such as time stops, price stops, price targets, et cetera, et cetera they’re just looking for the explosion of volatility in a certain direction and then they get out a day later, three days later, whatever. But it’s certainly not what I would call trend following. So, I just want to throw that out there. I think you’re nodding, so you’re probably in agreement with that. 

Okay, cool. Let’s move on to a question from Matthew. Matthew writes, “I have a couple of questions for next week if you think suitable to have when Rob is on (of course) to get into the trend following funds within tax wrappers in the UK. (And by the way, I will say just upfront, we don’t do any tax advising or any advising really, but I’ll let the question run.) 

To get into the trend following funds within tax wrappers in the UK, the only viable route seems to be the UCITS funds. However, the target volatility of these funds is usually half of compared to the programs available to larger investors. Do you have thoughts on how the allocation to trend following funds, in an overall portfolio, should shift because of that?” 

And then the second question, “For smaller investors who want to run their own portfolio, would it be more advisable to start with an equity momentum strategy first, as described by Andreas Clenow in Stocks on the Move, due to the minimum tradable positions generally being smaller for stocks compared to futures, even with spread betting some minimum positions are too large for small investors.” 

Then he writes, “I have four trend following UCITS funds at the moment. Two of them appear to be about 10% target volatility and the other two about 15%. 

[00:26:43] Rob

So, mathematically if you have a low volatility, then you should allocate more to it. That’s just basic portfolio optimization. The downside to that (and we do talk about allocations to trend following later in the episode), the downside of that is that you may potentially lower your return unless you’ve got access to leverage. We’re in a world where, if you’re just trading only UCITS funds, you don’t have that.

So what that means in practice is, let’s suppose that you did some optimization exercise and said, “I want to put say a quarter of my money into trend following, with a trend following fund running at say 25% annualized vol (which is what I run my own money at). Now, if you can’t get that 25 vol fund, instead you can only get access to a fund that’s running at 12.5%, then mathematically it’s very simple. You should increase your allocation from 25% to 50% and that’s basically the answer to the question. 

Now, whether you are comfortable doing that, of course, is entirely another question. It’s disappointing because it would be nicer, in many ways, if there were more UCITS funds available, and they did have high vol targets. It’s also more efficient from a fee perspective as well. 

That’s why a lot of institutional investors… We used to have a big investor when I was at AHL who used to say to us, “We want 30 vol. fund subclass because, that means we don’t have to give you such a big allocation to get the same risk level (because those guys understood the math behind that). And we also want a discount on the fees” (because they were big enough that they could do that. I won’t name them.) 

They were big enough that they could do that. As a retail investor, obviously, you don’t have that kind of negotiating power, sadly. So the short answer is you should put more in really, which is the answer to most questions on this podcast. Yeah, more trend following, more trend. 

[00:28:39] Niels

Yeah, just to quickly comment on the equity thing, I just want to say to people that there is a reason why the UCITS space can’t offer 25% vol. In the UCITS space there is a cap to the value at risk budget we are allowed to use in the funds. That normally means that once you get past 15% annualized vol, maybe 17%, you run into the risk of running into the cap on VAR on a regular basis. So, we have some real experience with that, which is also why we had to lower the leverage we use in our UCITS funds.

So there is, unfortunately, it’s regulation. I hate to say it, but often regulation actually makes it worse for investors even though they try to argue that they’re trying to make it safer for them, and so on and so forth.

[00:29:36] Rob

Yeah, this is the classic thing that regulation punishes the good guys.

Regulators want to punish the bad guys. If they didn’t have this, there would be some ‘nut job’ advertising a 50 vol insane fund that just did something completely crazy and was just designed to eat people’s money up. So, they have to regulate that. It’s quite hard to regulate and say, “You can’t do bad things.” People say, “How do you define bad things?” And the regulators are like, “Okay, there’s a VAR limit – that’s nice and simple. Here’s how you calculate the VAR. You can’t go above this limit.” It makes the regulator’s life easier and it makes the industry’s life easier as well, because you know, very clearly, what will or will not meet the regulations. It does, obviously, (all regulations, all rules) have side effects and this is one of them.

[00:30:30] Niels

The same goes for commodities, as far as I understand. The new UCITS funds, certainly in some domiciles, are not allowed to include commodities anymore. Which is great for those funds who are still out there, with commodities (I can think of one at least), but it is a big problem because commodities are the best diversifiers in a CTA portfolio, frankly. So why would you want to take them away from the investment public? It’s not quite easy to understand. Did you have a quick comment on the other stuff? 

[00:31:03] Rob

Yeah. Let’s move on to the thing. There are two parts to this second question, really. One is, is Andreas Clenow’s Stocks on the Move strategy a good one? Yes, I’ve read the book. It’s well founded. It’s basically a single stock momentum strategy that uses a filter based on the overall momentum in the market. I quite like that. 

Is single stock equities a good thing to trade for momentum? We’ve gone over this on the podcast many, many times. Should you do it instead of futures? Should you do it as well as futures? We’ve been around this course quite a few times. Of course, in theory, if you have enough capital you wouldn’t start with equities because you get less diversification within equities than you do across multiple asset classes that you can trade with futures.

Having said all that, if you don’t have enough money for futures, if you don’t want to go down the fund route, then yeah, it’s a reasonable thing to do. It’s a reasonable place to start. I think that’s a good idea, and Andreas’s strategy is a sound one, so why not? 

[00:32:13] Niels

All right, cool.

[00:27:20] Rob

Andreas, if you’re listening, I’ll have the $20 in the usual place. Thank you. 

[00:32:18] Niels

Exactly, I was just going to say that. There we are. 

Anyways, now we have a comment. I don’t think Mark, who sent in this comment, meant for us to necessarily have it as a big debate, but I do think that he makes some interesting observations that I would like to comment on, but more importantly, I’d like to hear your comments on it.

Mark wrote in, “Thanks for your podcasts. I enjoy them – very helpful. You and most of your guests are getting asset classes confused with trading strategies. I hear this especially when you discuss adding trend following to a traditional 60/40 model. Academia does a poor job separating the two.

Let’s assume there are five major asset classes: stocks (which can be private or public), real estate, bonds, currencies (which are just a zero duration bond denominated in other currencies), and commodities. Let’s assume there are three major investment strategies: trend following, mean reversion, and indexing. There are many sub strategies under each of those categories.

Therefore, there are 15 independent asset class/trading strategy combinations. The first step in portfolio design is to decide asset class percentages. The next step is to decide trading strategies. Those are very unique and separate decisions that both lead to different types of diversification. A 60/40 traditional allocation is usually 60% stocks, index strategy; and 40% bonds using index strategy.

A diversified CTA usually has all asset classes, but only uses trend following strategies. Trying to compare a single asset class index strategy with diversified asset class trend following strategies is changing two variables, the asset class and the trading strategy and can be confusing. First, you should be talking about adding more asset classes to a 60% stock and 40% bond strategy, adding commodities, real estate, and currencies to have more asset class diversification.

Then secondly, you should discuss adding trading strategies diversification using trend following to manage some or all of those separate asset classes versus other trading strategies.” 

Anyways, it was a comment. 

[00:29:18] Rob

Yeah, I’ve got some comments to make back. Of course he’s correct. Should there be more asset classes added to a 60/40 stock/bond strategy? Yeah, of course there should. I wrote a book about this, it was called Smart Portfolios. I don’t talk about it very much because it’s more about investing than trading, but absolutely.

I think the reason we talk about 60/40 a lot in the podcast is because it’s a very simple way of describing what investors have in their portfolio, apart from trend following. I bet you there is no one listening to this podcast who, literally, has a 60/40 equity/bond allocation.

We use it as a shorthand term. I think most people understand that. Should we go on more about adding other things into that long-only portfolio? Maybe. I personally have lots of things in my long-only portfolio. It’s certainly not just 60/40.

But I don’t think that necessarily falls within the remit of the podcast. Similarly, about adding trading strategy and diversification, yeah, I’m perhaps not unique, but I’m usual, in terms of the people on the podcast, in that I do talk about the other things I have besides momentum. I have carry. I have some mean revision. I have various other things in my system. So, that’s absolutely fine. 

Just very quickly, returning to the first part of the comment which is about this idea of saying, “You guys are doing it the wrong way.” If he’s right, and you’ve got three trading strategies of five asset classes, which is 15 combinations. He’s saying, “You should do asset classes percentages first, and then trading strategy rather than trading strategy and then asset class. From a theoretical point of view, the best thing to do is actually to do that optimization simultaneously.

In other words, to actually allocate all 15 chunks of your portfolio simultaneously. That will give you, in theory, the best outcome. It’ll certainly give you the highest backtested return because you’ve got more degrees of freedom in your optimization. Now, I’ve done quite a lot of work in this in the past, and it turns out it doesn’t really matter if you do it in a hierarchical way. In other words, if you do asset classes and then trading strategies or trading strategies and asset classes, it doesn’t actually matter that much. It will reduce the backtested expected return a little bit because you’ve got fewer degrees of freedom.

Whether it’s better in the backtest to do trading strategies first, or asset classes first depends mainly on the correlation structure of those two things. In practice it doesn’t really make much difference. If you look at out of sample rather than in sample backtest returns, there’s no difference at all. It doesn’t really matter which way around you do that. 

So, doing it one way or the other, to me, comes down to where you are starting from. So, if you’re starting with someone who has a 60/40 portfolio, or long-only portfolio, that you can’t play with, and you’re adding trend following onto that, which is the kind of model that we talk about in the podcast It is a shorthand, as I said. Then fine, do it that way.

Take the long earning stuff as exogenous, add in the trading strategy stuff separately. If you’ve got complete control of all of those buckets, then by all means do it the other way round, or do them simultaneously. It’s not, in the grand scheme of things, going to make a huge amount of difference.

I quite like the comments and I kind of agree with them, but I also think that in the long run, certainly in terms of explaining things on the podcast, there’s a reason why we keep things simple. We don’t start with, say a 250,000 by 250,000 matrix of possible assets you can allocate to and say, “That’s the only way to do it.” We’d be here all day if we started like that.

[00:38:23] Niels

Yeah, I completely agree with that. Of course, I have not done any of these tests that you have done. I’m just speaking from my gut feel. My gut feel is that actually if you, as a starting point said, yeah, you’re going to have some bonds, you’re going to have some stocks, but on top of that, if we can just pick one thing to add to such a portfolio that will really make a difference, it probably is trend following; of course, a diversified trend following strategy, et cetera, et cetera. 

I think it’s perfectly fine. It’s a good comment, of course. I think it’s perfectly fine and very reasonable for people just to use those three “items” in a portfolio. Because as you said, it makes it simple to understand, it makes it simple to implement and, who knows if there would be a benefit to try and juggle five, six different asset classes and various trading strategies and what have you. It’s a good point. That’s also why I wanted to bring it up. 

All right, next one is up. Here’s a question from Toby. Toby writes, “You’ve covered trend following at length in your podcast, and rightly it is a robust strategy. I know Rob has spent some time working on mean reversion systems. Could you spend some time explaining your approach by building mean reversion systems, apart from carry, which has elements of mean reversion for global macro, meaning it could be applied to bond stocks, currencies, and commodities. Have you looked at cross-sectional signals within each macro sector?”

What are your thoughts, Rob? 

[00:39:48] Rob

Yeah, so I’ve got this stuff within my portfolio and actually talk about it in the new book as well. The basic idea would be, rather than looking at the outright momentum of say, US 10 Years, I look at how US 10 Years have done relative to the overall bond complex.

If it’s done relatively well, I’d be buying it. If it’s done relatively badly, I’d be selling it. That’s the basic idea. It’s not a massively profitable strategy. The reason for that is that most of the trends, within an asset class, are actually global across a particular asset class. I have, yet, another strategy that picks up on that and actually trades the whole bond sector, the whole commodity sector as a single market, effectively. By looking at the cross-sectional effects, you’re effectively removing that and you’re left with the residual, and the residual is quite weak. I’d say (I can’t remember the exact figures), I’d say roughly 75% of the performance of trend following seems to come at an asset class level and only about 20% comes within the asset class level.

So it’s a nice little thing to add to a system, but it doesn’t give you a huge amount of extra value. 

[00:40:57] Niels

All right, cool. Here’s a question from Zach. Zach writes, “In chapter 12, Speed And Size of Systematic Trading, it is noted to ignore holding costs of certain instruments when calculating the standardized cost in SR terms.” 

By the way, SR, Is that sharpe ratio when you write about this? 

[00:41:14] Rob

Yeah. 

[00:41:15] Niels

Okay, okay. 

“Holding costs for futures is noted as the required rolling from contract to contract. If somebody has to trade the mini or micro contracts to achieve a position, should one still ignore those rolling costs or should they be factored into the sharpe ratio calculation? Or, if Rob needs a question to promote or discuss his new book (funny, he should mention that) will Rob’s new book include a section of standardizing trading cost calculations when trading micro futures contracts?” 

[00:41:46] Rob

Yeah, I’ll be honest I really hate it when people ask me about stuff from my books, especially Systematic Trading, which I wrote in 2015, that’s now seven years ago. 

I can’t remember why I said to ignore holding costs in certain instruments.  Maybe it was because, in a lot of futures, they’re pretty minimal, to be honest. They don’t have a huge amount of an effect. Now strictly speaking, you should include rolling costs in your calculations. If it’s something that’s traded monthly, then that definitely should be true. Obviously the slower your trading (so if you’re trading in a quite a slow trend falling system), then holding costs will form a larger and larger proportion of your overall trading cost profile. They are important. 

The question of whether to trade, the micros, the minis, or the full size contracts is something that is in the new book as it happens. It’s certainly the case that, for a lot of the smaller contracts (not all of them some, it’s pretty competitive, but for some of them), because they’re not quite liquid enough, those interest rate contracts that Zach’s talking about, I don’t actually trade those because they don’t meet my liquidity requirements. I do stick to the bigger contracts.

Yeah ,trading costs is one of my bug-bears, that people don’t spend enough time thinking about it. It’s definitely worth doing a proper analysis of what the trading costs are on a given instrument before you decide to trade it. So, definitely along the right lines there. It’s good to hear from someone who’s such an avid reader of my stuff. 

[00:43:19] Niels

Yeah, absolutely, absolutely. 

All right, last question, then we move on to your very interesting topics. It’s a question from Anise (I think it’s pronounced.) “Hi there. Firstly, I love the show. Thank you for all the advice and insights you and your guests share. A quick background on myself. I’m an individual trader and I trade my own account. I trade cash equities and I have no experience with futures. I have started to dip my feet into longer-term trend following. My question relates to using cash equities versus futures for trend following.

I notice that most commodities/FX/ETFs have very low liquidity, making it harder to incorporate them into a system. Excluding them would mean I lose market diversification. Do you think investing time in learning futures is worth it or should I stick to the available liquid ETFs? Thanks very much.”

[00:44:11] Rob

It’s that question again, isn’t it? Equities or futures, or ETFs or futures? So, there are pros and cons to trading ETFs instead of futures. Liquidity is an issue potentially also it can be quite expensive in terms of holding costs, the AERs that get charged.

There can be issues with things like if they’re leveraged; they’re two or three times leveraged. There can be issues if you’re buying the short versions of them to proxy for a proper short sale. In commodities, specifically, there can be issues in the way that they’re actually hedging their exposure and they’re proxying the futures contract. It’s not so much of an issue now, but  if you go back a few years, there was a very large oil ETF that was holding the front contract and was getting absolutely burned every time they rolled that contract. Now, they’ve moved out a bit further into the curve. So, it’s slightly less of an issue, but there are a lot of things you need to think about. 

Is it worth learning how to trade futures? It depends on how much capital you have and how much time you will need to devote to it. If you’re trading a reasonably sized portfolio, then you should probably start thinking about it; and you’re reasonably experienced and you know your stuff because ultimately futures are a leveraged derivative.

They’re automatically a lot riskier than a cash instrument. That’s one reason why regulators (going back to regulators) wisely tell brokers to make sure that people know what they’re doing when they’re trading them, and why we have things like margin and so on, and so forth, to protect us.

So if you feel you fall into that category and you have a decent sized amount of capital, then from a purely theoretical perspective, you will always probably do better with futures because there are more things you can trade. You don’t have to worry about a lot of the issues of ETFs that I’ve discussed, and they’re generally cheaper as well.

[00:46:13] Niels

Yeah, no, I think that’s sound advice actually. Very good.

All right, first of all, thanks very much for all the questions that came in for Rob. That’s fantastic. That’s great. We appreciate it. Now we’re going to jump onto some of the topics that you brought along, Rob. The first one I think obviously it’s something we’ve been discussing a little bit, certainly when it comes to CTA replication, you called it hedge fund replications, so maybe it’s a little bit broader as a topic. But I do think it’s valuable to discuss. It’s something that there’s a lot of time on Twitter being spent on right now. Where are we going with the hedge fund replication from your point of view? 

[00:47:00] Rob:

Yeah it’s a topic I have found very interesting because I actually used to work with a guy who wrote one of the original academic papers on hedge fund replication. He’s a great guy and really good, but I was always very skeptical about the methodology that was used. I think you, on the podcast a couple of weeks ago, or three weeks ago (I can’t remember now), but you had somebody who was describing a particular…

[00:47:35] Niels

Andrew Beer.

[00:47:36] Rob

Yeah, Andrew was doing it. He got this very interesting way of doing replication. My understanding is that what he basically does is do some kind of regression of recent returns to determine which instruments he should be holding that will replicate the ideal hedge fund.  I say hedge fund because it is wide in the CTA universe. In fact, in theory, you could pump any returns of anything; an index, you could try and, for example, replicate the returns of, I don’t know, renaissance potentially by using this sort of methodology. I just thought it was worth talking about a few different ways of doing this. It’s a topic I think is really interesting.

I don’t think we’ve ever talked about it in much detail on the podcast before, so I thought I would just bring what little knowledge and experience I have of the topic to the table, as it seems to be a current one. The way that Andrew is doing it is one way of doing things.

You know, there are potentially problems with this. One is, obviously, a lag, so any kind of regression that you are doing is going to suffer from the fact that positions are always changing. The positions that a fund has on today will maybe be very different from positions they had on three months ago, or six months ago.

It comes back to this thing we talked about already, which is a question of how long should your lookback window be? If your lookback window is too short, you are going to have a real signal to noise problem and really struggle to work out what the correct set of positions are because you’ll just have a very noisy regression.

If your lookback window is too long, on the other hand, then you are going to struggle to keep up with the way that the underlying fund the funds are changing their returns. That’s potentially an issue. There are other ways of doing it though. One person pointed out to me that you can actually get access to the underlying positions of many funds, admittedly with a lag of course. If you can do that then you can just copy them. That’s a simple way of doing it. 

You could do more sophisticated things, which is actually try and, again, using a statistical technique to try and work out a model for how their positions are adjusting and then try and do what’s called a nowcast. So you’re trying to forecast what’s happening now. You try and forecast where the positions are now, which is information you might not get for a day, or a week, or a month, depending on the reporting requirements. That’s another way of doing it. 

The way I actually do my own trading (this dynamic optimization I talked about before) is like that, in a sense, because I have a series of positions that an imaginary large hedge fund is holding. Then, I say, “How can I closely match that?” So, in some ways, the technique is not that dissimilar. 

The third way of doing it, which I, personally, think is probably simpler than all this other stuff is actually just to build your own little trend following model. So, what you do is you say, “I’ve got a series of assets I can allocate to.” 

This goes back to the earlier question about allocations. You might have 25 or 30 or 50 liquid futures markets, and you might have half a dozen different systems and strategies that trade those. Then you start with that. What you do is say. “What’s the best combination of those assets that most closely matches the index that I’m trying to replicate?”

That, to me, probably makes more sense, to be honest. We actually used to do this with our competitors when I was working for AHL. We actually had a Winton model. We had an Aspect model. We had a Campbell model.  We were trying, (and I don’t know if they still do this, obviously it was many years ago now), but we would try and work out what their allocations were to, say, fast versus slow trend following versus carry, what they had in bonds, equities, and so on and so forth.

That would give us some color as to when there were big differences in our performance, where that might be coming from. It could just be that they were better than us or we were worse than them, but it could also be just a really short term thing. It was a useful context even if we didn’t do anything with it in the end.

So yeah, that’s just to give people a view of the different ways of doing this and there are other ways of doing it as well. In fact the original work done by my ex-colleague was a very much more statistical technique, which basically tried to match the distribution of returns of some index or some strategy, some fund. You could literally pump anything into that. 

One of the criticisms of that was they chose an asset that had a really high sharpe ratio. If you have an asset that has a really high sharpe ratio, you can do things like pump it through a particular filter and turn it into an asset that has a slightly lowest sharpe ratio, but a very high positive skew, which of course would make it look like a trend following strategy. You can do stuff like that. It does rely on having an asset with a very high sharpe ratio to begin with, which is a bit of an in-sample fitting issue potentially. So, there we go. 

[00:53:11] Niels

Actually very interesting. First of all, super interesting that you are a little bit of a replicator. You can put that on your business card as well.  You and I actually spoke quite a lot about your dynamic portfolio optimization. It actually makes perfect sense. Now that you explain it, that it’s replicating similar to what these replicator funds are doing.

Also, of course, it’s very interesting to hear that even people like AHL, at least at that time, would spend any time trying to figure out what their competitors were doing. I know for sure we are not spending any time finding out what competitors are doing, but let me post a couple of questions to you from my conversations and interactions with Andrew.

As I’ve said before, I like what they do in the sense that I think it helps bring trend following into more portfolios. So, the success they’re having, great for them, probably great for the industry as long as they don’t screw it up (to be frank), because if they did, it’s not going to be great. So, the pressure is on Andrew on that part. 

However, Andrew is quick to point out that one of the things that was part of the reason why they ended up doing this was that they would often meet investors, talking about single manager risk and how could you pick a manager because next year he might be the worst, even though he was the best the year before and so on and so forth.

That of course is true, but I also think it would be, then, equally true to say, couldn’t there be a single replicator risk? What if their model goes wrong? They’re only using one (I’m simplifying it), but they’re only using one approach, and it could go wrong, frankly.

So I think we can’t say that these replicator products have less risk. They just have a different set of risks, in my opinion. Of course my own view would be to say, why don’t you just get three or four different trend followers, buy them from people with long experience, and I think you would be absolutely fine, and you don’t have to worry about single manager risk and all of that stuff. So, that’s probably where I’m still sitting. 

The other thing I thought of was that with these replicator funds (and certainly the way Andrew explained his approach on the podcast a few weeks ago), he looks to replicate about 20 combined return stream managers who combine trade hundreds of markets, but he’s doing it through 14 different markets. 

[00:55:51] Rob

Yeah. There’s a reason for that as well. 

[00:55:53] Niels

Yeah, feel free to comment on that. 

If I can just ask you the question. The question is, the counterintuitive nature of that, for me, is that here we sit every week and we preach diversification of models, diversification of markets, timeframes, et cetera. Then you stick it into one model, and you stick it into only 14 markets. It’s doing exactly the opposite of diversification, in my opinion. I can’t articulate, it’s above my pay grade to see specifically where those risks may lie, but just based on my experience, it sounds more risky somewhere in that process. That’s just how I see it.

[00:56:39] Rob

Yeah, the reason why there’s a small number of markets is this replication strategy. You can’t do that with a large number of markets because intuitively there’s not enough information in the series of returns you’ve got to tell you what position you should have in 50 or 100 or 150 markets. You can only trade a small number of markets. 

The other problem is that they have to be the same markets. One of the interesting things about my dynamic optimization strategy is that, in theory, I’ve got 150 markets I can trade. Actually, I think I’m up to 200 markets now, but maybe 50 aren’t liquid enough.

I probably have about the same number of positions on as a CTF. That may be even fewer now. I think I’ve got 10 positions on. Those 10 positions will change over time. So, today I’m short US 20 Year, that’s where my bond exposure’s coming from; tomorrow it could be mini JGBs, it could be something else. You just don’t know.

So, if I was to run my dynamic optimization strategy (and I’m not sure if I’ve done this explicit experiment, but it’ll be interesting) and I’m only going to pick 14 markets and stick to those, to do exactly the same as I’m doing now. But instead of allowing myself to take a position in any of 150 markets, I only allow myself to take a position in just 14, I know for a fact that the correlation, that the degree of how close I’m matching the index (the hypothetical large hedge I’m trying to match), I know that would go down considerably.

Now, the expected return in the backtest may or may not change. In fact, it may even go up cause I may have, by luck, just picked some markets that have done really well. In fact in my book, this is an explicit experiment that I run. I say if I just trade a lot fewer markets, I do much better than with the dynamic optimization. But that doesn’t mean the dynamic optimization is poor, it just means I happen to have picked, at random, markets that have done really well. One example would be, say Euro/dollar futures which have great backtested sharpe ratio, but also have a relatively small amount of capital required to trade them because they’re quite low vol.

My process for choosing markets always picks those, and that boosts the backtested sharpe ratio. Of course that doesn’t mean that I should definitely have Euro/dollars in my portfolio. 

Yeah, I feel a bit bad because Andrew’s not here to defend himself.

I’ll say one really good thing about this fund, it’s really cheap. I think it’s 85 basis points management fee, which is really competitive. That’s definitely a good thing. I’ll be honest with you, I don’t know what the minimum investment for this thing is, but if I had, say, 2000 pounds, and that was all I could put into trend following, this would be right up there in terms of what I would do with that money because that’s nowhere near enough to run my own strategy. That’s nowhere near enough to invest in multiple managers or anything like that. This is a product where it’s definitely a good niche,  (and I’m sure Andrew and myself wouldn’t say this), it’s definitely not a replacement for having a full-fledged fund. In expectation, I think it wouldn’t do as well because of this limitation we described.

[00:59:54] Niels

By the way you left one option out in terms of replication. That is, of course you can replicate the replicator because they have to publish their positions every day, so you’d just be one day out. So, it probably makes no difference. 

[01:00:06] Rob

Yeah, that’s true. 

[01:00:07] Niels

I know Andrew, funnily enough, is not a big fan of being replicated, even though he does like replicating other people. Anyways, I’m sure he’ll be back to defend that comment, but there we are. 

Now, we need to go to the next point because that is also something we’ve been discussing. Actually, I discussed it a little bit with Alan last week, and you’ve written a long blog post about it, maybe not specifically in the same vein but let’s just go through it. That is, of course, about how do I decide how much to allocate to trend following?

So I’m going to give you the floor on that. 

[01:00:44] Rob

Yeah, I like the way you described it as a long blog post. My blog posts are always long. This is actually one of the shorter ones. 

[01:00:51] Niels

Okay, fair enough. It looked long to me when I was trying to prepare for this conversation.

[01:00:55] Rob

You can look at this blog post in two ways. You can look at it as me answering this perennial question, “How much money should I have in trend following?” Or you can look at it as an exercise explaining to people some of the problems, and the right way to do portfolio optimizations in general. I’ll keep it in the context of trend following because that’s what people are interested in.

But really it was a way for me to illustrate some issues and techniques I use to deal with portfolio optimization. The basic idea behind this is that you say, “How much money do I want to put into (let’s keep it simple for the reasons we discussed way back at the start of the podcast), into say 60/40, and then some trend following strategy.

And I make it really, really simple. I literally consider just two assets: the US 10 Year and the S&P 500 – that’s my 60/40 portfolio, and then I trend follow those. I’m simplifying it down to the absolute minimum, really, you can get away with. An important thing to say is that, for example, this technique will probably, net-net, give you a lower answer for what you should put in trend following, because it’s not including, for example, all the diversification you get from all those other asset classes that we can have when we’re trading a full suite of futures as we’ve discussed at length already.

So what I do is I say what are the basic problems with the idea of doing this thing where we lookback a certain amount of history of returns, and then we decide, based on that, how much money we should put in trend following how much money we should put into 60/40. The first problem is, if you look at your data set (so my data set for these two assets goes back to 1982) , from ‘82 until quite recently, 60/40 did incredibly well. If you dig into that, both stocks and bonds have gone up. Why have they gone up? Actually it’s because we have this secular period when inflation was falling, interest rates were falling, and as a result, equities are being rerated as well.

So, the first thing I do is say I don’t think it’s reasonable to use this historic period without any sort of adjustment or consideration for the fact that we’ve had these secular trends. The first thing I do is I actually take the price series and I actually adjust them to account for these secular trends.

So what I do is say okay, let’s take equities, the PE ratio in September ‘82 and the S&P 500 was about 9, and now it’s about 20 (at least it was when I did this analysis, It may have changed a point or so since then). If you actually do the math, that works out to an average annual return of 2% a year in equities coming from this rerating effect.

And if I do a similar thing with US 10 Year, I get a lower number. I get 1.2% a year. But actually, because they’re less volatile, that’s a more significant component of the returns. What that does, for example, for US 10 Year is it reduces the average return that you see by about a third. So, you still do well until quite recently, but you don’t do as well as you would do without this adjustment process. That’s the first thing I do. 

The second thing I do is say that any of these optimizations are relying on this data that’s effectively uncertain. It has uncertainties within it, and the optimization itself isn’t very robust as well. I use some techniques that I’m not going to go into now because it would take way too long to explain. (People who are interested in the technical details can go on my blog and just have a look.) What I show is that although there is a theoretical optimum that you should have, the more interesting thing to think about is the degree of uncertainty in that.

And if I do that, what I actually find is that the absolute maximum you should have in a 60/40 portfolio is about 45%. This particular setup comes out saying trend following should be at least 55%. That’s obviously quite a big number. I then explain where that comes from and so on and so forth.

So, there’s technical stuff there, but perhaps the more interesting thing is that (and you would probably expect this to happen) the amount of money that goes into trend following, versus 60/40, increases when I apply the de-meaning. If I take away these secular trends that have flattered 60/40, but have also flattered trend following, particularly in bonds, also in equities (to a degree).

If you’ve got this secular trend, then that’s going to be a very slow trend for volume models, that’s going to give you some extra return. If the market’s not going to buy up as much, and probably in the future we shouldn’t expect this rerating to continue. It just can’t. 

It’s recently broken and it’s a nice way of saying trend following was good in the past if we don’t expect bonds to have this sort of drift effect behind them that they’ve had for the last 40 years, then you should put even more money into trend following, which is intuitive what you’d expect, but it’s a nice way of demonstrating it.

[01:06:19] Niels

I think Rich did a slightly different study. He just used the SocGen Trend Index on a 60/40 portfolio, et cetera, et cetera, and he came up with the optimal allocation to the trend index being 64%, interestingly enough.

So we reversed the 60/40 to 40/60, but where the 60 was trend and not bonds, but there we are. 

All right. Let’s move on because I know we have a hard stop today. 

[01:06:46] Rob

Last year we did a two part episode of Christmas and New Year where we had all the co-hosts together at once, and that’s going to happen again this year. Hopefully we’ll be able to get a date sorted. So, that’s one to look forward to. 

Niels, do you want to give us the lineup of the luminaries that are going to be on this episode? Of course. You and I, obviously. Who else? 

[01:06:08] Niels

Yeah, it’s the usual suspects. It’ll be you and I. We’ll have Cem, we’ll have Rich, we’ll have Mark, and so it’ll be super fun, and Alan of course. (I was just thinking that there was one missing here.) But I think more importantly, I think what we’re going to try and do is try and find themes, each of us, that we think defines the year of 2022 and discuss them.

That doesn’t mean that we won’t be open to some really good suggestions, right? So if anyone out there today thinks that there might be a great topic that encapsulates a theme of 2022, why don’t you send them in to the usual email address and we will prioritize the best ones and see if we can fit them in.

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

[01:08:11] Ending

Thanks for listening to the Systematic Investor podcast series. If you enjoy this series, go on over to iTunes and leave an honest rating and review, and be sure to listen to all the other episodes from Top Traders Unplugged.

If you have questions about systematic investing, send us an email with the word question in the subject line to info@toptradersunplugged.com and we’ll try to get it on the show. 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.