227 Systematic Investor Series ft. Richard Brennan – January 21st, 2023

[00:00:00] Intro

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

[00:00:24] Niels

Welcome and welcome back to this week’s edition of the Systematic Investor Series with Richard Brennan and I, Niels Kaastrup Larsen, where each week, we take the pulse of the global markets through the lens of a rules-based investor.

If you’re new to the show, I hope this episode will trigger your curiosity enough to check out the back catalogue and listen to some of the past episodes you may have missed, like my conversation with Rob last week where he answered some really detailed questions about moving average crossover systems and a lot of other stuff.

Also, I would encourage you to listen to the midweek episode where Cem and I were joined by the Head of Commodities Research at Goldman Sachs, Jeff Currie, for what I think might be even a top five episode of 2023, even if we’re only three weeks into the new year. But it was really that fantastic so I hope you’ll enjoy that conversation.

Of course, not least, I hope you are enjoying our new mini-CTA series where Alan and I have managed to line up the real decision makers at most of the largest CTAs in the world for some very meaningful conversations on many of the most important topics. I don’t think this has ever been done before so I hope you enjoy and appreciate all the time and access we have been given to these true industry leaders. So head over and check it out once you’re done listening to me and Rich today.

Anyway, Rich, it is amazing to be back with you this week. It’s been way too long. How are you doing down under?

[00:01:48] Richard

Doing well, Niels. It has been a long time and I apologize for my lack of attendance but it’s been quite hectic over here. It’s quite warm and wet at the moment. It has been for quite a few months, but the holiday season was good. I managed to do a lot of things non-trend following and take my mind away from the roll.

But in that process, I managed to catch up with some of the things that I’d left out while I had my head fully engaged in trend following. So a bit of timber logging and a bit of reading, all of these different things. It was great, a great holiday. I don’t think I missed much with the markets, but it is good to be back. My head’s back in the trend following world and it’s good to be back.

[00:02:32] Niels

Yeah, well, it’s good to have you back. Actually, I think I’m going to take full advantage of you being back today because instead of doing a full market wrap, we are actually recording a day early and it has been a very busy week for me as well. So I haven’t really got my head around following what’s been going on.

Instead, I think I have persuaded you to give some thoughts to what we’ve heard so far from two of the very well-known, very established firms in our industry that are part of this new CTA mini-series, namely Man AHL and Transtrend. I know some of it you only just managed to listen to today. But I appreciate the fact that you’re up for it and sharing your thoughts about what you’ve heard so far.

Now, for those who may not have listened to this series, we are trying to get all of the guests to talk about the same overall topics, not necessarily the same questions, but the same overall topic so that we can get a global assessment of how these extraordinary managers think about some of the issues that we have often discussed on the podcast and that are so important. So feel free to start wherever you want. What are your thoughts? What are your takeaways?

[00:03:51] Richard

Well, look, it’s a very exciting project you’re doing and I’m loving it because it gives us a chance to look at the different perspectives and viewpoints of people that are being classified as diversified systematic trend followers. So at the moment, I’ve listened to the Man AHL podcast you did and also with Transtrend with Harold, and I really enjoyed them. I loved the consistent questions asked for both because it gave me a chance to sort of evaluate the similarities and differences between both camps.

Clearly, from my perspective, there was nothing that stood out as being controversial from what you and I have been saying for a very long time and the members on this podcast. So that gives a big tick in my box that I think our industry is focused on the same general principles, but there are definitely nuanced differences between say Transtrend and Man AHL and my approach and your approach.

There are sufficient nuanced differences to create diversification within trend following and that fascinates me. It is also very important to me because I think we all understand that we all interpret trends differently. I think that’s actually because the way we interact with the market is an application of how we apply our systems to that market and the relationship is always a unique one. My systems will be different to your systems, will be different to Harold’s systems, to Man AHL’s systems.

So, the way the systems have been designed is basically being discretionarily developed through philosophical logic and then they’re being tested through the crucible of systematic backtesting and ensuring that they’re not curve fit. Harold had some very interesting things to say about the situation that arises when we are too reliant on a backtest, which does sort of encourage this ability to overfit systems, which therefore, reduces its ability to respond to future differences than what has been experienced in the past.

Also, Man AHL confirmed. I was very interested in how they are very closely following an ensemble system approach in their processes. Harold is very much concerned with what drives these infant trends and how trends sort of emerge and how you can classify different collections of trends. So his approach is definitely different to my approach. Harold, I think, uses a more fundamental view of how markets behave and then he develops his philosophy about the trends that occur in the different classes of markets that he categorizes in those different forms of trend.

I look at simply the outliers of those markets, the anomalies of those markets. Man AHL, they’ve got a very heavy research division and they’ve got significant diversification of system types that are attacking different elements of trends. So all of these, there are similarities, but there are differences, enough to create diversification in trend following. We often have a laugh about, well, 100% trend following is only the way to go.

Well, there is clearly enough diversification in trend following alone to justify that statement. As you and I know, when we group programs together into a trend following index (like the TTU Trend Following Index), we see that that grouping of diversified systematic trend followers actually improves diversification, improves compound growth over the course of time. So, there are only benefits in bringing together this collective of different opinions and different nuances. That really is a powerful concept and a reason for why we diversify so widely.

[00:08:28] Niels

Yeah, beautifully said. You’re certainly highlighting a lot of the things that I also took away from these conversations. I actually also remember clearly, when you do a lot of conversations right after each other, sometimes it’s difficult to remember one from the other. But actually, Harold’s conversation, one of the things I remember that stood out to me was also how he really focused on how we should embrace being different.

In a sense that being part of an index (I think this came up during maybe CTA replication or whatever it was), the fact that we should try to be different even though we are operating within the same framework and with the same core principles. So yeah, lots of good stuff. I can only say, since I’ve done a few more conversations already, there are a lot of good stuff coming from different people, different perspectives.

We have short term managers coming up as well because they’re also part of the index of course and they have their views and their focus points. So as I said, it’s been a great learning journey as well, I think both for Alan and me, lots of fun. Now, before we jump into our topics today, I just want to make sure that the battleship is still floating.

[00:09:55] Richard

Yes. Battleship is still floating. It was a great year last year, as mentioned in the monthly report that we posted on the blog. My battleship suffered the same symptoms as the majority of trend followers in 2022. So in other words, a magnificent first six months or first half of the year and then effectively stagnation and just maybe a bit of drawdown and giveback in the latter part of the year.

So, whilst my trajectory initially was a record that was unheard of in my historical record of my performance, my expectations were watered down by the end of the year. Still a very good year. Still a very good year, still a record year for me. But I haven’t had a very long track record like you guys, so I was very impressed with it. Yeah, so the battleship is doing well, still sort of stagnating at the moment. It’s not a bad start to the year so I can’t complain. But my waters are feeling good for 2023. Let’s see how it goes.

[00:10:57] Niels

I love the optimism as always, of course. Now, I actually think it has been a little bit of a soft start, at least the funds that I follow closely. It’s been a little bit of a soft start to this year, but perfectly normal after, as you said, a pretty strong year last year. Our trend barometer at the moment stands at 45. That’s kind of neutral.

But when I look at the numbers, as of Wednesday, this week, they are down for the month, somewhere between 70 basis points for the BTOP 50 Index and a 2% down for the month of the SocGen Trend and the SocGen CTA, somewhere in between. Short-Term Index is pretty flat so far. But again, this is a long-term game and therefore we shouldn’t put much into how we’re doing just in the first few weeks of 2023.

Now, what we should put a lot more into, of course, is answering questions from our loyal community. We go up north today, we go to Estonia. We go to Markus. Markus writes, “Greetings from Estonia. Great podcast. I’ve been a loyal listener for several years. I’ve heard Richard talk about his backtesting method, how every year he adds the new data of the past year to run his backtest and adjust strategies accordingly. I was wondering if you could ask him, on the show, what his biggest takeaways have been after adding 2022 data and running the optimization if he’s done that already. Thanks so much. From Markus.”

Absolutely. I have now put that to Rich.

[00:12:39] Richard

Yes. Hi, Markus. Good question. Yes, I’m currently in the throes of adding an additional years’ worth of data. So 2022’s data is now going into my dataset, and I am revisiting my portfolio. Just to provide a bit of context, my portfolio is compiled using a workflow process and what this workflow process does is it takes as much data history as it possibly can, across multi markets, to then generate what I regard as the optimal asymmetric portfolio before compounding.

Once that is achieved, I then run that series of systems, that have emanated from that workflow process, over that data set for the period of a year. Now, when that year is up, we’ve got a years’ worth of data that now can be directed back into the workflow process. What that additional years’ worth of data is doing is providing an extension of the historical dataset with new happenings that have occurred during that year that might not have been observed in the historical record.

So the way I’m viewing this is a process of expanding robustness. Each year, as new data comes into my systems, I’ve got more data history, not only for the year for a single market, but a year for all of the markets in my portfolio. So, let’s say that I have 60 markets in my portfolio, a single year gives me an additional 60 years of alternative market data.

And because I’m applying these simple systems across markets, that’s effectively giving me a significantly increased sample size, which is always assessing the portfolio in terms of robustness; how it survives, how it copes with risk with a new years’ worth of data. Now I’ve got additional data that may have been representative in the past or it might have been a new novel regime. So, now I know that my systems can navigate more regimes.

Each year that this occurs, the ability of my portfolio demonstrates two things. It demonstrates robustness in that it can survive the entire 30, 31, 32, 33, 34, 35, etc., years of history. Plus, it’s also an adaptive process in that the additional years’ worth of data is continually being reinjected into my workflow process and a new fresh batch of algorithms is being developed.

Once this process (I’m currently undertaking now for 2022) is finalized, I will have a complete sequence of new strategies which I then inject into my portfolio. The strategies I’ve currently got running in my portfolio, they get turned off but they’re still allowed to complete their trades.

In other words, those systems that were valid in 2022, that are still active, still have trades, they’re allowed to play out those trades because they might be going for another one or two years if they’re profitable. But as soon as they exit, no new trades are entered into on that prior batch of strategies. But this injection of new strategies, which meet the robustness and adaptability condition, then take over for the next year.

So this rolling process, year by year, is a way that I feel that I’m sharpening my portfolio. Now you might say, “Well, Richard, what are the major takeaways you get from this process that you do?”, which is part of the question you asked. I don’t actually get any takeaways. But what I do is this process, this process I use, is what I regard as sharpening the portfolio, continually keeping it sharp, continually keeping it up to date.

So I don’t, for instance, examine what were the differences between the strategies this year versus the strategies generated last year because it is just too complex to do that. I’ve got thousands of strategies operating across 60 markets. I might have six strategies per market across 60 markets. Well, that’s 360 systems operating in a particular year. There’s an overhang of systems operating from the prior year, from the prior year before that, etc. So there might be 400, 500 systems all operating.

To try and work out what is the significance of this to me is too complex. But what the process does (process, is important) is it’s sharpening the portfolio based on the data that comes into the stream of the workflow process.

[00:17:53] Niels

Excellent. Now, of course, I love when you keep saying sharpening because it could very well be interpreted as you’re trying to get the highest Sharpe ratio, but I’m sure that’s not what you mean.

[00:18:06] Richard

Exactly.

[00:18:08] Niels

All right. Okay, great question to start the day, so to speak. Then we have another question in from Tim and then we’re going to jump to your quite interesting topic that we’re going to talk about in a couple of minutes.

Tim is writing, “Thanks for the interesting and informative podcasts. As someone who is involved in the systematic trading industry, I particularly enjoy the Systematic Investor Series. For your discussion with Richard Brennan, I would like to ask the following question. Having listened to several different podcasts on trend following including Top Traders Unplugged, I’m a little puzzled with what to me seems to be an inconsistency in the selection process of instruments to be added to a trend following portfolio.

Let me explain. Richard stated a few weeks ago in one of the sessions that it was essential to invest as broadly as possible in order to increase the likelihood of capturing as many outliers as possible. An example he made went along the lines of that investing in 4 out of 12 categories could miss out on the outliers if the outliers only occurred in two categories of assets. Professor Brennan has also done plenty of backtesting on real data and I believe also on synthetic data to identify what instruments are worth adding to the portfolio – with some AUM permitting and testing producing constructive results, an instrument can be added to the mix. So far, so good.

On the other hand, however, Richard has stated on many occasions that we don’t know which instruments will become the next outlier and therefore, we should ensure to have a wide range of instruments as possible to have the future outliers in the portfolio. To me, this sounds inconsistent as the backtesting is done on historical data and if the conclusion is drawn from that as to whether to trade an instrument or not, it must be deemed to have some information on potential future performance, which Professor Brennan then denies by being adamant that we don’t know where the next outlier will come from. I’m not claiming that the backtesting isn’t useful, nor that I have any idea where the next big profits from trends will emerge. But the combination of trusting backtests while then saying they are not really useful in identifying outliers isn’t completely logical to me. I would very much appreciate you and professor.”

“The professor,” it’s very confusing when they call you professor, Rich, but I will try to read what is being said here. “I would very much appreciate your and the professor’s articulate response to this. Many thanks from Tim.”

[00:20:47] Richard

Look, okay, let’s start with the professor thing. All right. I’ve got a laugh. I’m aware of this growing stigma, this professor being attached to my name.

[00:21:00] Niels

I think it’s meant as a compliment.

[00:21:02] Richard

That’s what I’m hoping. But look, there is another side though. I am aware that, in my very nature, I sound as though I’m preaching. This is a comment that lots of people make when I’m talking to them. I don’t wish to come across that way. It’s just the way I come across. The reason for that is that I view this world as a very complex world, and my view of my explanations of this world [is that I] have to have a consistent narrative to explain it in a way that demonstrates its complexity.

It’s not a simple world. One thing affects another thing, affects another thing, affects another thing. So, when I’m talking, like when I’m talking with you, Niels, it might come across as a preach or a spiel or an academic spiel or a professor’s spiel, but it’s only because I’m excited. The way I talk, in this narrative, is that I’m wanting to get to a punch line, but I need to discuss lots of different facets before I can throw the screaming punch at the end, which hopefully challenges people. That’s what I’m after.

I want to challenge people. I want to challenge logical thought, because that’s the way for me. That’s how I get excited in learning. And I think that’s why, so, I apologize if I’m preaching. The professor tag, I don’t know what I’ll do with it. I suppose I just have to live with it. But on to the question. So it’s a very good question, great question. It does require some explanation.

What I do with my backtest process, the aim of my process is to develop an outcome at the portfolio level that has the most asymmetric profile, beneficial profile, before I consider any form of compounding treatment. The aim of my backtest is to get a big exponentially rising equity curve before compounding treatment. What I want to do there is I want a symmetry in that equity curve. I want non-linearity being represented by the upside and I want linear outcomes being represented by the downside.

When I say non-linearity, I want profits to be many orders of magnitude greater than the losses incurred in getting those profits. So the way that asymmetry is constructed is that I am cutting off, at the trade level, the ability for my trades to have an excursion towards the left tail. I’m cutting losses short, but I’m giving maximum opportunity for my profits to go to the right tail. I am focusing on right tail performance with my trade outcomes. That achieves this asymmetry, and if I get this asymmetry before compounding, I believe that is better than a straight line before compounding.

If we can imagine a portfolio result that is uncompounded, I am not looking for a straight line. That to me is looking for Sharpe. So in relation to a straight line, before compounding, a linear ascending equity curve, that’s nice for people that want smooth returns and consistent performance across that line. But I’m after something more than that. I’m after maximum asymmetry. I want this exponentially rising portfolio equity curve before compounding.

So therefore, I’m wanting what I call step-ups in my equity curve that the step-ups exceed the drawdowns of the equity curve. And I’m trying to keep the lower bound of my equity curve linear and the upper bound of my equity curve nonlinear. So how do I do that? This comes to the question.

I do separate the information I extract from a backtest into two areas. One is I am not extracting information about outliers from my backtest. All I want to do in my backtest is to apply systems to historical data that can simply demonstrate that when there are outliers in that historical data, they are capturing that outlier. What I’m also trying to do at the same time, with the upside, is restricting the ability of my systems to be on all the time. I only want them to be participating when prices become material in nature.

So, that information from the backtest is something that will not be replicated in the future because I am targeting outliers. Outliers are unpredictable anomalies in the future. So therefore, so long as my systems can catch what’s been provided historically, that is not used as a basis to forecast expected future returns. I have got no idea what the future is going to hold. I’m not looking at the upside of the historical backtest to do that.

Now, the downside. How do I get this linearity that I’m looking for in this asymmetry? This is the information I’m deriving from the backtest. Because this linearity is looking for the boundary that exists between normal market behavior and outlier behavior.

I don’t want those negative left tails in my trade distribution of returns. I’m cutting losses short. How short? Now, what you’ve got to do there is you’ve got to define a boundary. How short do you cut your losses? Now, because I’m targeting outliers, I want my losses to only ever be linear losses. So the way I’m using history is sort of like the reverse of how someone who uses a backtest to predict the future does. I’m using it as a process of exclusion.

I’m using the history contained in a backtest to identify what is normal market behavior versus what is abnormal market behavior. When I’m dealing with adversity, I’m therefore putting a boundary to say I will accept normal market behavior because I’ve got to give my models freedom to move within normal market conditions but their efforts are being directed towards capturing these outliers.

Now, we know that it is not sufficient just to have a market with outliers without a system that’s capable of extracting the edge from those outliers. So the interaction between how your markets interact with that market is very important in the correlations created between the system’s interaction with the market. It’s very important that your systems are configured to be able to exploit that opportunity.

Because for instance, if I had systems that cut losses too short, I would continually be getting whipsawed, which would be compromising my ability to focus on the outliers. So I need to know what normal market behavior is to reduce my losses to linear outcomes because I need freedom of movement to allow them to accept that level of volatility when it’s targeting those outliers so it doesn’t get kicked out of the trade all the time.

So I’m always looking for this boundary. If you could imagine, the outcome of this process does one very important thing. If my models were turned on all the time and I was looking for any type of market price behavior with my models, my outcomes of my process would be representative of the entire market distribution of returns. So I’d have a big right tail, a big left tail, and a bell curve distribution between that. The trade distribution would be mirroring the market distribution of returns. I don’t want that.

I want to exclude my participation in normal market activity, and by exclusion, the trade outcomes naturally are converging towards the tail behavior of the market distribution of returns, and this is through a process of exclusion, not inclusion. People using backtests for the purposes of achieving expected returns for the future are using the historical market data as a basis to say that history has some property that will continue into the future. I’m not doing it that way.

I’m saying that history has properties that identify the boundary between normality and the boundary between anomalies. I want that information from my backtest to help me with my models. I hope that’s clear.

So in other words, what this outcome does of my process is I don’t derive information from my backtests that helps me generate expected returns. I’ve got no idea what those expected returns are going to be because it’s contingent on outliers and I don’t know when they’re going to occur. But I do want information out of the backtests that manages my risk based on what history has presented. That’s as best as I can go.

So, because I’m defining my stops, trailing stops, and small bet size by what history is presented, I’m saying that’s the best I can do in extracting information from that historical dataset. I’m managing risk so that the risk is going to be linear in nature, and leaving myself open for nonlinear outsized wins. That creates this asymmetry in this process. I hope that explains it.

[00:31:15] Niels

Yeah, wonderful. Very detailed, Rich. I’m sure that Tim appreciates that level of detail.

Now, we’ve got a really interesting topic actually today, I will say. We as humans, we like to predict outcomes, there’s no doubt about that. We have a predictive brain and the methods we deploy to trade markets tend to be predictive in nature. But some of us feel that opportunities actually lie in our inability to predict a future with precision.

So today, we’re going to focus on uncertainty rather than certainty. Not much is known about uncertainty. We tend to avoid looking at it and exclude it from our statistical treatment. But there are few traders out there that seek to exploit it. So now we, under your expert guidance, Rich, will be exploring why these financial markets carry with them inherent uncertainty. So maybe you could start by setting the stage a little bit for what we’re going to talk about.

[00:32:26] Richard

Sure, Niels. This topic of today’s podcast, I’m trying to bring us all to focus on this term called uncertainty. I’d like to start with a great little quote that I found from Zero Hedge, which was talking about Goldman Sachs, who wrote, “Goldman were channeling their inner Kipling,” and the topic of the discussion was called “If” which is Rudyard Kipling’s poem, “If”. Goldman says, “Unknown unknowns are, by definition, impossible to predict. But if we do not experience another black swan event in 2023, and/or if the US achieves an economic soft landing, and/or if the Fed stops raising rates while inflation recedes, and if China’s bumpy reopening transitions back to stable growth, and Europe successfully transitions its economy further away from a dependence on Russian gas, and the US consumer remains more resilient than many expect, then the S&P may just be able to eke out a flat performance amid still elevated volatility.” I had to laugh. And then they said, “Happy New Year, ” and I had to laugh.

But this fundamentally comes to this notion of uncertainty. All of these things that Goldman was saying were conditional events. Conditional events are very hard to deal with in the land of probability. Something that is conditional on priors, a prior conditional event, a prior conditional event before that, the probability landscape is continuously changing at each of those sequential events. That creates this probabilistic minefield for those that are seeking certainty.

What I want to do in this discussion, Niels, is focus on uncertainty. We hardly ever focus on it because it is statistically undefined what uncertainty is, yet it is such an important aspect of trading and there are some traders out there that actually respond to uncertain regimes, trend followers being one of them. So what this is saying is that there is something in certainty and there is something in uncertainty.

Most investors are focused on certainty because of their predictive brain. They like this comfort of dealing with certainty in terms of the expectation that something that happens from the past is going to continue with fidelity into the future. But the reality is that when I sit here peering into the future, all I see is uncertainty. Now, we often talk about backtests, we often talk about statistical treatment, etc. It doesn’t help me get rid of this feeling of uncertainty when I’m looking into the future. All I see is black.

All I see is the possibility of uncertainty from all of these possible events that are here now or new events that arise later. The way they interact together creates this indeterministic nightmare in my vision of the future. There is no certainty in that future. I think everyone feels that way. I think everyone feels that the only reason that we can predict what happens tomorrow is that what’s happened over the last few years has happened with repetitive frequency.

So our assumption is that that repetitive frequency will continue tomorrow. But you see, when we’re dealing with uncertainty, we’re dealing with real systems. Now, conventional statistical treatment doesn’t deal with real systems. They deal with idealized systems. But we’ll get on to that. But the reason I want to focus this episode on uncertainty is to shatter the illusions that people have regarding their belief in certainty. I want to challenge that notion in this episode.

[00:36:47] Niels

Yeah, I think that’s a great way to phrase it and give some context to it. So why don’t we just jump in? You classify your process as hunting for outliers. I think we’ve kind of touched on this over the years. But would you say that’s a standard treatment amongst all trend followers? How do you see that?

[00:37:10] Richard

Well, yes. I believe that there is enough opportunity in our divergent landscape to accommodate a fair few differing philosophical views about what constitutes trend. So I see trend followers as being this collegiate group of people with generally similar principles but there is a lot of diversity in those philosophical approaches.

So, I like to view myself as sitting apart from the rest of the trend followers in my passion for hunting for outliers. I don’t think all trend followers are doing that. I think they are to a degree, but they might also be focusing on other edges that exist with different definitions of trend. As we’ve discussed, Niels, I view trending markets as being found in lots of areas of the distribution of market return.

There are trending price series in what I call convergent market regimes, which we trade a trending price series and that’s why it’s called counter trend in mean reverting series. So some people focus on that as part of their trend definition and they might also capture some outliers. There are also trends, which none of us like in the trend following world, and they are the random trends that simply can appear like trends but they have no enduring potential in them, no correlation in them. So, with counter trend or with what I call convergent trends, they’ve got what I call negative serial correlation.

The trends I like to focus on are a specific class of trend with positive serial correlation that are found in the tail regions of the distribution of returns. So, they’re the ones I’m hunting. I’m specifically hunting them.

So why do I hunt them? I hunt them because I’ve only got finite capital. And in my interpretation, if for instance I believe (which I do philosophically), that the greatest change, the greatest opportunity lies in uncertainty and in these tail regions. Why would I invest part of my capital towards other forms of trend if I have this significant belief that the greatest opportunity lies in outliers or in the tail regions?

With my limited capital, I want to invest all of my limited capital in that belief. I don’t want to spread it, which might dilute my philosophy, I want to be focused with my philosophy towards the tail region. So, there are a lot of things I’ve got to deal with because of the issues associated with this concentrated philosophy on outliers.

One is that we know outliers don’t happen all the time, We know that they often cluster together. So it’s then up to me, if I’m so passionate about these outliers, to develop novel ways I can distribute those outliers more evenly throughout my portfolio time series. There are ways I can do that through the models I use to attack that. So, I recognize the shortcomings of my approach, but I’m always looking at ways to amend that so I can dedicate my efforts towards tails. Yeah, I hope that explains it.

[00:40:35] Niels

Absolutely, but I think just for the benefit of people who are new to the show or maybe just people who want to know a little bit more, if you don’t mind, let’s dig into a little bit more about what you mean by hunting the tails of the distribution. Perhaps you can elaborate a little bit on that.

[00:40:53] Richard

Okay, so this is where uncertainty comes in. An outlier is, by definition, something that is an anomaly, something that cannot be statistically defined. So if it can’t be statistically defined, how do we use statistics to capture it? Well, that’s a bit of a paradox here. We don’t use statistics to capture an outlier.

What we do is we leave our systems with open profit potential to allow it to capture any opportunity that could become an outlier without applying constraints to that assessment. Any constraints I apply through statistical reasoning in my world of outliers, there is no statistical validity to that because these outliers are these anomalous, infrequent events.

I was listening to a surfer, one of these longboard surfers the other day, who specializes in the big surf, the big waves. And when he was talking, I just saw the trend follower or that tail hunter in him, because he was saying, “Oh, I don’t focus on those little waves.” He said, “I spend a lifetime focusing on the big waves, the big set, the magnificent wave, because each one of them is an exotic, unique experience. None of them are the same. Each of these outliers are what gives me the experience of a lifetime. I want these lifetime experiences. I don’t just want any old surf. I want these things that thrill me and things that create an addictive passion to go out the next day and wait and wait and wait with dedicated patience, not bothering with the small stuff, worrying about the big stuff, wanting that big stuff.” That is a tail hunter.

Statistics have their uses, but in our world of outliers, they have very limited use. In the world of convergence, they have a lot of use. But in my world of outliers, I can’t rely on statistics, what I rely on is uncertainty. So let’s look at what statistics tries to achieve. The aim of statistics is to draw a sample from a population of data to examine whether the properties of that sample are correlated with the properties of the population.

So, we could say that some statistical properties are more likely to be representative of that population as opposed to other properties drawn from the sample itself, which mostly may be uncorrelated. So what we’re looking for is correlated properties between the sample to the population. And if our sampling finds that there are properties of a sample that are correlated with that population, we have this inherent assumption that that sample is sufficiently valid to use as a basis for projecting into a future based on that assumption.

With the correlated broader properties we use to extrapolate this broader future, I call that “an illusion of certainty,” because it is only if these correlated properties hold into that future that certainty is maintained. If you can imagine what I’m saying is that when people apply simple statistics, they assume these idealized models. They might say, “I have 10 billiard balls here. If I select one, what is the probability of that selection process? Ten percent.”

It’s using an idealized simple model where we’ve got a population of 10 billiard balls and then we’re working with if I take samples of those billiard balls, what is my sample estimate, is it correlated with that broader population? That is great in mathematical treatment but it’s not great when applying that mathematical tool to the real world. Because the population assumption, for instance, is only static at an instant in time.

As soon as that population progresses into the future, it changes its probability, its overall probability statistics. So the sample, or the inference you obtained from that sample, at that frozen moment in time no longer holds. The correlations change over time. Samples change over time. Populations change over time. This is because, unfortunately, mathematics, whilst being an incredibly powerful subject, can be used to construct fictitious worlds.

So, what I’m talking about here is that, for instance, the idea of a perfect circle. Does a perfect circle exist in the real world? The idea of pi, the symbol pi and its decimal points that go out to infinity, is that a realistic assumption in this real world? Do we get this continuous repetition of pi to infinity or is this a theoretical concept that has an application in a theoretical context, but not an application in a real context?

This is the problem I’ve got. It’s not the maths that’s the problem, it’s the application of the maths that’s a problem to the real world. Mathematics is a tool, just like any other tool. Mathematics is a tool. You’ve got to use the right math to interpret the right system. Now, that’s why I’ve got a passion for physics, because I prefer dealing with systems, something that’s tangible, fungible things that I can model, things that I can touch, things that I can feel.

I don’t like dealing with abstract mathematical symbols like perfect circles, perfect spheres that might not exist, alternate universes that might not exist, all of these things that mathematics, in its power, can predict. It can predict this entire fictional world out there called an alternate universe in mathematical terms that in reality has no practical application.

Physics, however, brings us down to earth. It brings us down to looking at the physical system. We of course, use maths, that are tools specifically configured to assess that system, but it does it in a much more realistic way. So if anyone asked me what do I prefer, maths or physics? I’m a physics boy through and through. Maths can get complex. I tend to avoid the complexity and especially when it starts getting esoteric, I start just closing my eyes because I just don’t see its application.

So really, I suppose where I’m getting to is that when we deal with statistical properties, typically, we’re dealing with Gaussian statistical treatment, simple statistical treatment for complex situations. Now, you can’t apply simple statistical treatment to complex situations. If you do, you get a boundary of uncertainty created between the difference from what your simple systems are telling you or the simple analytics is telling you to what the actual real system is.

There is a discrepancy between the conclusions of your statistical sampling versus what the reality is. So I might be able to say, with certainty, if I have 10 billiard balls and I take so many billiard balls at a fixed instant in time, I might say with certainty what is the sample distribution within that population, because that’s an idealized model.

But in the real world where populations are frozen in time, and as soon as that next time lapse, they alter in their nature, this is where statistics starts getting very murky in its precision. The difference between the reality and the statistical inference is uncertainty.

[00:49:39] Niels

Now, of course, only an Aussie like you can use an example with a surfer to describe these things. But unlike a surfer who has to get up on the board and therefore might actually only want to concentrate on the big waves, we have systems, we have computers, so, maybe to phrase it in another way, why avoid all the smaller trends that are found elsewhere in the distribution and only focus on outliers?

[00:50:11] Richard

That’s a great question, Niels. Look, I appreciate that there are others that do that, and for good reasons. They’re getting diversification of opportunity by focusing on other trends. But in my world, in my landscape, I think that uncertainty is the biggest change of all. When I look at a complex adaptive system, the financial markets are an example of complex adaptive systems.
Let’s take a planet as a complex adaptive system. So, when you look at the history of our planet, you notice that there are periods of stable regime punctuated by major transitions. So, major transitions occurred in the early Cambrian when bacteria emerged, when algae emerged, when seashells emerged, when mollusks emerged, when lungs were created where the fish could leave the sea and occupy land, when sexual reproduction occurred which created a major transition event.

All of these punctuated moments of major change were the biggest change of all in that sequence of events. There were periods of stable regime where these regimes were inherently predictable. But when you got these transition events occurring, these transition events were epoch moments. They were major transitions. It might have been a meteor hitting the Earth that created such a big impact on the stability of the regime. It changed its very nature in perpetuity.

So what we find is that when we deal with change, we see that there is a lot of gradual change in fairly stable regimes. I’m not interested in that. That stuff to me is just small perturbations or fluctuations. So periods of stability, ho hum, pretty boring to me. However, what I’m looking for is this transition event which disrupts the stability, it totally changes the nature of the system. Transition events are things that cause outliers.

The way to view a transition event, in terms of the way a system is configured, is a transition event is a symptom arising from the reconfiguration of a system in its entirety. So, given a major transition event, what we find is that the period before that transition event was a particular regime, the period after that transition event is a totally different regime. It’s not the prior regime, it’s not a repetition of a regime. It’s a new regime.

What this transition event does is a magnificent reshuffling of the entire system. It’s reshuffling the relationships of the system. It’s reshuffling all elements or aspects. It’s changing the state of that system. With that altered state, we get major opportunities arising, we get the extinction of species that were reliant on the prior stable regime. We get a complete wipe out of competition that was prior to that transition event, and we get almost like a new opportunity arising which can be exploited under far less competition.

We get new species suddenly exploding, things that we’ve never seen before, mammals coming from dinosaurs, all of these things. So, these transition events are uncertain events. They’re things that the regime itself could not predict. They came out of nowhere. I remember that we had this discussion with Jean-Philippe Bouchaud, where he was talking about endogenous versus exogenous events. So most traders are concerned with news events, information getting into the market, important news events.

His understanding of the microstructural interpretation of price moves said that only 10% was attributed to news events, 90% was attributed to these endogenous events; things that we couldn’t statistically define at the time. Maybe in hindsight, we had the ability to say what they were, but they arose almost out of nowhere. There was no ability to predict these things arising. They’re off the statistical landscape. They suddenly popped up out of nowhere. Where did they come from? Well, it’s this concept called entropy, Niels.

What actually happens with uncertainty? The origins of uncertainty lie in this principle called entropy (I’ll get into that later, probably won’t get into it right now). I suppose what I’m saying is that I believe the major changes are the transition events that create significant opportunity. Why would I invest my time bothering with small changes when I’m looking for these major transition events?

Because if I look at the most successful investors on this planet, Warren Buffett, Jerry, Dunn, yourself, Niels, all of these people, what do they attribute their success to? It is these handful of events that are outliers. It’s not all of the different machinations. They’re the things that keep you afloat, keep you alive. It’s these rare opportunities that you can list on your fingers. It’s the 5% trade, 5% of total trades that are the windfalls.

These outliers have the ability to be the major bonanza for those that can exploit them properly or they can be the risk of ruin events for those that fail to take heed of uncertainty and the outlier. I focus on these outliers, because to me, that is where I get the biggest bang for buck.

[00:56:20] Niels

Yeah, indeed. It makes very logical sense. There’s no doubt. I think we can all think of things within the last two or three years, perhaps, that came completely out of the blue. And I don’t mean inflation and what Christine Lagarde thought about that coming out of the blue. I mean things like COVID that we didn’t have or people didn’t really think about until it was there. So I completely agree. Now, in terms of outliers, and maybe we have to kind of limit the details of all of this because we got a few points we want to cover. But in terms of outliers, would you say that they are generally features of liquid markets?

[00:56:58] Richard

Yes, I actually think that they’re almost a universal quality of all liquid markets. Now, what I’m meaning here is that traditionally, statistical treatment has been dealt with what I call equilibrium models, models with the assumption of equilibrium. So Gaussian statistical treatment assumes linear outcomes, it’s assuming distributions are operating about an equilibrium. Now, these real-world financial markets are anything but in equilibrium, they are open nested systems.

The financial markets are connected to the producers, are connected to the economy, are connected to the central banks, are connected to other markets. One person’s holding in a particular class of derivative instruments might be influenced by the bank account he’s holding or the assets he’s holding on another instrument. They’re interconnected, they’re causally connected. It is a minefield out there. It’s an open connected system. Because it is anything but a simple system operating around equilibrium, I view it like all complex adaptive systems, you’ve got to take that term seriously, “complex,” because they are complex; adaptive, because they are non-stationary, they adapt.

Complex adaptive systems inherently operate what I call at the boundary of chaos. In other words, they’re just holding on to respectability while peering into the world of chaos. They’re operating at the border because where they’re operating, this gives the greatest opportunity for a complex adaptive system to become more efficient over time. For instance, what it will do if its state is bound by a stable regime (there is not much change that occurs there), it needs these transition events, stepping into chaos, temporarily, to create a complete restructure of that system to make it more efficient.

So, a complex adaptive system, over the course of time, starts self-regulating itself, it starts becoming much more efficient. Any complex adaptive system, you find that it becomes more and more efficient. We get nonlinear relationships occurring.

Let’s take a city for example. A city is a complex adaptive system. When it first starts out, you might have isolated houses occurring. Then we get a grocery store coming into that township, a dependency starts existing between all of those houses and the grocery store. It’s not a one-to-one relationship. It’s not a linear relationship. It’s a many-to-one relationship, a nonlinear relationship being established between the participants in those models. Then we get a police station coming in, then we get roads coming in, then we get electricity coming in.

Now, this system is continually being reconfigured with efficiency in mind. What is the most efficient way, with limited resources, to produce the biggest outcome for the city that can be achieved? So, there is this magnificent reshaping of that system continuously over time. It’s continuously getting more efficient, but at the same time, it’s becoming more fragile because of its dependencies building up in that system.

So we might think that there is a huge city out there called New York, but it has amazing fragility because it is entirely dependent on electricity. Turn that electricity off, and that city stops operating as a city, the transport system. Everything breaks down in the city in a click of an instant because of this dependent relationship developed between these one-to-many relationships and nonlinear dependencies that creates this fragility. We might think the city itself is stronger, is bigger, but it’s actually more fragile.

So, what happens with all liquid markets, Niels, fundamentally, they are seeking to become more and more efficient over time. But at the same time, they’re becoming more and more fragile over time. As their efficiency builds, so does their fragility. Uncertainty is an entropic condition; I’ve mentioned this before. I don’t want to have to get into the physics of this. But all I’m saying is that in any complex adaptive system, there is an entropic direction, just like the direction of our universe. It’s always looking into the future, moving into the future.

There is this entropic ratchet that is moving us into the future from a condition of low entropy to high entropy. And what that arrow does, that arrow from low entropy to high entropy, it actually gives the motion to all processes that occur on this planet. Every single process applies this entropic directional arrow towards the future, and this is the principle that from low entropy, we move to high entropy.

So this time related directional dependency means that in the future, it’s always more disordered than the past. There is more uncertainty in the future than there was in the past and it’s because order is progressively being broken down as we are moving slowly towards ultimately an equilibrium condition where all that movement stops, everything just occurs in equilibrium.

This whole universe is perched at the boundary of chaos, slowly, progressively unwinding, with an entropic direction towards this ultimate demise when it hits equilibrium. That’s when everything stops. So all of those statisticians that are using simple models, they’re effectively saying the world has stopped. We’re saying the world is alive, we’re focused at the boundary of chaos. That’s the only thing that’s keeping us alive.

[01:03:28] Niels

It sounds like what you’re saying is that maybe it’s not really the math that we should be applying, but maybe other things like physics, and if so, what other things? What’s the incorrect math we may be using and how can we correct that situation?

[01:03:49] Richard

Okay, here’s a bold statement, Niels. Now, what I’m saying is a majority of the industry are using the incorrect maths. You and I know our problems with Sharpe. What that’s trying to do is it’s trying to create straight lines in a nonlinear world. In other words, it’s trying to say that the straight line equity curve with no volatility is the best outcome we can get. Now, what that’s doing is it’s saying let’s apply straight lines to a nonlinear world. How do you do that? How do you do that?

To me, I scratch my head. If you do not accept volatility in your equity curve and you’re going for straight lines, to me, that is a symptom of what I call warehousing risk. There is some trick you’re using to make this straight linear opportunity in a non-linear world. As soon as I see that straight line, I think long term capital management, some trickery is going on, because that’s not how this universe or these financial systems operate.

So, I’ve got real problems with the standard statistical treatment. It’s the wrong tools to use for the complex adaptive systems we use. In fact, I think it’s very, very difficult to apply mathematical tools until we know how these systems operate. So our first thing we’ve got to do, before we blindly apply any mathematical tool to something, is first understand how it operates. When we look at complex adaptive systems, we see the incredible complexity, hence the word complex adaptive system.

We see it in things such as collective behavior, that’s a subject all in its own right; game theory, a subject in its own right; nonlinear dynamics, a subject all in its own right; systems theory, a subject all in its own right. Pattern recognition, evolution and adaption, networks, all of these things are these immense topics which have an influence in us understanding how complex systems behave. It’s not a mathematical tool we need to understand these systems.

Firstly, we’ve got to understand their Machiavellian nature. What creates this complexity, this non-stationarity, these moving goalposts? For God’s sake, don’t apply statistical tools because you’re assuming simple things, but which are incredibly complex, evolving all the time. The thing is that what we’ve got to understand is this. We’ve got to understand that physics has demonstrated, over the course of hundreds of years, that we often slip into the rut of being biased with a particular approach simply because we refuse to take alternative interpretations of how something works.

We see this change from this view of things existing as objects, or building blocks, or billiard balls. So science initially was concerned with, let’s slowly in a reductionary logic investigate this system and look at its component parts, like the breaking down of a machine into its assembly construction pieces, into its fundamental constituents. They’re assuming they’ll find this structure at the fundamental level of objects. That’s what they’re assuming.

They’re assuming that when they got down to the atom, when they got down to the electron, when they got down to the quark, they’d find ultimately that there were some fundamental objects. Now, what we’ve found is that every time we’ve conducted this reductionary exercise in this real world, it’s not objects we come to, it’s processes we come to. You and I, we’re sitting here, we’re talking to each other. And someone says, “That’s Niels,” and I say, “No, that’s a process.” Someone will say, “Well, what do you mean by that?” I’ll say, “Well, okay, this is an argument called the Ship of Theseus.”

The argument goes like this: If I have this ship built of wood and I replace all the components of that ship: its wood, its sails, its anchor, but still keep the form of the Ship of Theseus, if I replace every component of that, is it the same thing? People say, “Yes, it’s still the Ship of Theseus.” The thing is you’ve totally replaced all of its componentry. What we’re actually saying is it’s the process that is important.

So when I look at Niels, I see a brain, I see molecular structure, I see processes in Niels at the skeletal, the organs, the molecular level, continuously replacing themselves. Niels as an object, is changing over the course of his life from when he was born to when he is old. It’s a process that emerges and changes, grows, builds. It’s not just a collection of objects being rearranged.

We see this in the human body, we see this in the brain. When we look at how the brain works, it’s process driven. We see this when we look into a rock, we find that it’s comprised of atoms. When we look into the atoms, we see that it’s mostly space and that it’s just derived by processes between electromagnetic repulsion and attraction between electrons and protons.

We go into the protons, we find that the protons are built of quarks. When we go into the quarks, we find that the quarks are the way that forces are being constrained by vacuum space which is giving material mass to things that we objectify. It’s the processes that are ultimately important.

When we were at school in our early days, Niels, you might have remembered the early interpretation of Niels Bohr’s work, where we looked at the atomic structure as having a nucleus and these electrons rotating like orbits around the nucleus of the atom. The problem with that was that it was a totally wrong interpretation. We were looking at the world as a revolving billiard ball system, a bit like a solar system here today.

But the reality is there are no balls, there are these excitation fields called electrons which is a field orientated process. There are no particles, there are just interacting fields: electromagnetic fields, gravitational fields. Things that we assume as being objects like space and time turn out to be fields. In fact, in physics, space-time is the gravitational field. It undulates, it’s dynamic.

It’s not this arena that Newton assumed which was there was this arena called space, and there was this arena called time; these external contexts of which everything within could be defined. That was Newton’s view.

Einstein put that to bed saying, “Wrong, Newton. Liked your approach but it’s just a special case.” The best case we can say is that space-time interacts with all of the fields within it: so, electromagnetism, the nuclear weak, the nucleus strong fields, the space-time fields, this interacting fabric are just processes, events. So, there are no things in systems. But we like to think there are things so we treat them as statistical things. But there are no things, there are processes which have ephemeral nature. They have a duration. Every single process never lasts for eternity. It only has a finite duration, a shelf life, and that duration ends and that process is over.
This is unlike the world of billiard balls of Newton where things lasted and had permanence, statistics has permanence. It’s not like that, a moving goalpost, ephemeral nature, all of these things. Probably waxing lyrical here Niels but you know where I’m ongoing.

[01:12:13] Niels

I do, but I have to disappoint you. I actually don’t remember much from my school time, Rich, but I appreciate your faith in that. Let’s round this up a little bit in some ways, if I was going to simplify this, because it is a very complex system, not just adaptive system, but it’s very complex, what you’re talking about today.

In many ways, I think I’ve referred to it and I know that we’ve spoken about this for a while. When I think about how I think that we’ve seen the change in the environment in the last couple of years, I think about this where we have to go back and we have to imagine the unimaginable, that’s kind of how the markets will behave, meaning we have no idea what can happen.

But in this world, what we’re trying to do, we’re actually trying to manage risk through this completely unknown and uncertain world. So, I don’t know if you can briefly summarize how do we then fit that into it and then we’ll wrap up for today. We’ll probably come back to this in the future.

[01:13:24] Richard

Yeah. Good point, Niels. The way we should be addressing it, and I think trend followers do address it this way, and this is different to a convergent trader. A convergent trader, when they’re talking about uncertainty, they’re talking about risk, and when they’re talking about risk, they’re seeing risk as being the deviation from expected returns. So, the convergent trader is saying, “Hey, I’ve undertaken the backtest. I’ve projected this backtest into the future and I have this idea called what I expect in the future. Any deviation from that is what I’m referring to as risk.”

What I say and what a trend follower might say is, “Hold on, you’re treating risk far too narrowly.” That is the risk of variation about certainty. What about the risk of uncertainty, that which cannot be statistically measured? You have only given us the risks that can be statistically measured in the volatility about expected returns. What about that which cannot be statistically measured but exists? What about that uncertainty?

So a trend follower views risk as the risk of an adverse event. What does that mean? It means the risk of a deviation from expected returns plus the risk arising from uncertainty, which cannot be deduced from statistical means. In other words, cannot be deduced from historical projections.

Just finalizing this, I need to talk about entropy really quickly. Entropy is the level of disorder in a system. What we’ve found and physics has found is that it’s not energy driving the universe, it’s entropy driving the universe. What drives our planet is the low entropy of the sun, it’s not the energy of the sun. This is a very important concept, something difficult to understand but it is essential.

Before we understood what entropy was, before that, physics was blind as to what is creating the directional momentum of forward processes in this world. All of their equations they were using from Newton’s equation, even quantum equations, all of these equations were what we call time symmetrical. There was no preferred direction in those equations. In other words, if you had a system process going in one direction and you reversed that process in a camera, it could go in the other direction.

There was no preferred temporal direction. This was a real problem for physics. They found that at the elementary level, there was no direction. It was neutral, it could be either way. But we find that thermodynamics gave us a clue. The first clue was thermodynamics and the first clue actually came from the investigation of steam engines, the investigation of a process in this universe that is actually asymmetrical. In other words, it is a non-reversible process, and that is the flow of heat, the passage of heat from hot to cold.

Further investigation into that heat, in the world of thermodynamics, realized that heat is a property not of the individual particles themselves. For instance, I can’t say how hot an individual electron is, or how hot an individual proton is, or how hot an atom is. When heat is involved, it’s where you’re using a collection or a statistical ensemble of individual particles. This is really strange. So at the elementary level, there is no preferred time direction. It’s only when we get collections of objects working together, we get heat and we get a direction.

So what this direction does is it’s suggesting strongly that there is some temporal order to the way things flow in this universe. What they understand from their physics is that this temporal order is a condition arising from the fact that the universe goes from a low entropy condition to a high entropy condition. Now, what we mean by this is that if I get a deck of cards, 52 cards, and I order the cards so that the first 10 cards are all hearts, the next 10 cards are all spades, the next 10 cards are all clubs, and the next 10 cards are all diamonds in that order, or whatever, that is an ordered state.

It is what we call something with low entropy because it is highly ordered. If I shuffle the cards, that deck of cards, that state, that low order state, becomes less ordered. It has higher entropy. What do I mean by higher entropy? It means that a low ordered state has the inability to offer many different alternative variations. The lowest order state has only one variation, which is that order.

But as we progressively shuffle things, we find that the the number of alternative states gets more and more and more. But if we only consider one of those possibilities, the rest of those possibilities could be possible but we’re only considering one of those possibilities. So what we’re saying is that as things get more disordered, the state we measure of a system has a higher entropy.

In other words, there are more disordered states in that system than what we are measuring through our sampling process. That’s important. So, this is what drives entropy, and this is what drives everything from steam engines, to the universe to the financial markets. It’s a thermodynamic problem of entropy from low order to high order. It’s not energy. Energy is conserved in this universe. It changes in its form from kinetic to potential, etc. Everyone thinks it’s energy. It’s not. It’s entropy, entropy: the order to disorder.

The world is moving forward at a classical level in this direction of time going forward. But you can’t pick this up if you look at the entire detail of every single elementary particle because there is no temporal direction. It’s only when you look at ensembles of systems, when you get collections of things together, you can talk in terms of directions, you can talk in terms of time, you can talk in terms of these sorts of things.

So, that’s why when we are looking into this future ahead of us, all we see is uncertainty. What this is saying is that the future is always going to be more disordered than the past. What our historical assumptions are doing is looking to a lower entropy configuration which has got more certainty. But where we are now, into the future, it becomes less and less certain. It degrades in its degree of certainty and increases in the level of uncertainty.

Always be aware of that uncertainty. Don’t consider risk as just a variation attributed to expected returns. Think of risk being the uncertainty of things that have never happened before that could happen in the future that significantly increase the scope of what risk is about, and that’s what a trend follower does. I’ll close it off there.

[01:21:23] Niels

Yeah, no, that was actually a great way of closing it out. Also, in a sense, I think a lot of us, even though this was a very theoretical explanation you presented and now you can’t be surprised why people call you the professor, I’m sure, but I will say that I think we all probably sit back as we enter this new year feeling this is probably even more uncertain than what it felt like entering into 2022.

If we just go five years back, I think we can probably all agree that the last couple of years have been significantly different and very much more uncertain than how it felt for a long time. Of course, this is also I think, why we, on the podcast, are generally feeling very good about the approach we take to investing in markets, but actually also something I sense on the older people we’ve spoken to so far in the CTA mini-series that they are very optimistic about the opportunity set that trend followers will find going forward.

Now, it doesn’t mean we’re going to make money every month or every quarter or even every year. But for the next 10, 15 years, there should be plenty of opportunities in this uncertain world. So I appreciate all of this. Richard, I know that you put a lot of work in preparing, I should say, for these episodes and I truly, truly appreciate that as I’m sure all of our audience do.

If you do like this level of detail, why don’t you support what we do by going to iTunes or Spotify, Amazon, wherever you listen to your podcast, leave a rating and review. It’s one way to tell us whether we’re doing a good job. Also subscribe or follow us on the podcast. That helps it as well.

Next week, you can send in your questions, like Tim did this week. I forget the first question we had, who that was from. But send them in, Alan will be here and we’ll try and tackle those questions as well. You can do that to info@toptradersunplugged.com as usual.

From the professor and from me, thanks ever so much for listening. We look forward to being back with you next week. But actually, we will be back many times before we get to the next Systematic Investor Series because we have so much content being released in the coming weeks, pretty much four episodes every week. Take advantage of it and until next time, take care of yourself and take care of each other.

[01:24:06] 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.