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

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

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

Rich, it is so great to be back with you this week. How are you doing? What's going on down under, so to speak?

Well, it's all good down here, Niels. It's great to be back on the podcast, but the weather's trying to make up its mind at the moment. It's still sort of hanging on to winter a bit, but we clearly are sort of in a transition into a new season on the way. So. Yeah, we'll just see how that one plays out. But all good down here in the southern state of Australia.

Yeah, well, speaking with you over the years, Rich, I've learned that winter where you are, it still means like 20° in very nice weather. So, I'm sure it's all good.

Yes, we do winter a bit down here.

Yeah. But we have actually a really interesting kind of… I wouldn't even… It's a topic, but it's much more than that we are going to go through today. But before we do that, I'm always curious to hear what's been on your radar, which I think might be surprising to some people listening to us.

Yes, Niels. So, my mind is currently focused on the publication of our new book. That's your and my book Niels, The Trend Following Manifesto, world premiere. But if you remember, Niels, it's been three long years in the making. So, what's on my radar today is what you and I discovered over those three years we wrote that book.

But if you remember, Niels, you and I, we definitely weren't convinced by this explanation, as we were both accustomed to disappointing performance periods over long periods of time. But we didn't want to simply answer one opinion with another opinion. So, we wanted to examine what had actually happened. And if you remember, that was what actually inspired our writing of this book in the first place.

I do remember, but I can't believe it's really been three years in the making, and also keeping it to ourselves, more or less, over that period of time. And yeah, I do remember. We both lived through kind of that “difficult period” for the strategy that we love so much. And so, I understand the concern from some people.

Now, how would you describe what we found, just kind of briefly when we started looking at the evidence?

Well, from that research we undertook, we concluded that the edge had not disappeared, but rather it had moved. So, if you remember, Niels, we examined a swath of futures markets, across asset classes, using about four decades of data, and found that the underlying tendency to trend was actually stronger in the most recent decade than in any of the three decades before that. It was actually strengthening while much of the industry was worrying about its disappearance. And we found that while the edge had not diluted, the edge had migrated from the short lookback to the longer term lookback.

No. Over here in Europe, migration is a hot topic. So, the fact that it also happens within trend following, maybe we shouldn't be that surprised. And of course, I think we both believed all along that what we were doing had not really lost its appeal, but that we also, as we've always said, it's not a strategy that necessarily performs in all environments, as we all know now. And of course, we've also talked a lot about, over the years, that slower models and speed, every time we brought that up, the certainly slower speeds had been favorable in the last few years.

But if we look at maybe the deeper argument of the book, and today's not, by the way, about the book. We will do, I'm sure, a special episode as we move closer, but just a little bit of a teaser for the audience, who, of course, I hope (I'm sure you do as well) that they will all buy a copy of this book. But what would you say is kind of the deeper argument that we make?

Well, I think the book argues that outliers, or these big trends, they're actually structural features of markets. They're not temporary anomalies waiting to be arbitraged away, as many might assume. So, this does fly in the face of detractors that say our edge may have diluted as trends. These trends, they're not ephemeral price patterns, they're structural features of how these markets behave.

So, markets contain, I mean, interacting participants, leverage feedback, forced activity, and changing flows of capital. Those ingredients, they create movements that feed upon themselves and travel much farther than anyone expects.

So, while the structure does persist, as we found in our book, what can change is when and where those movements appear and they've drifted out to the longer-term horizon. So that uncertainty actually helps preserve our opportunity. But it does also make trend following difficult to practice because the returns don't arrive when and where the investor necessarily wants them to.

But you know where there was an unexpected return for you and I, Rich, that was the fact that it took us so long to do. Because, actually, during that two or three years, I would say that the evidence actually became stronger. So that was a little bit of an unexpected bonus, you could say.

Anyways, let's cut it short for now. When should listeners expect the book? And don't be too optimistic here, Rich.

But once we've got a firm publication date, Niels, we'll announce it properly. And as you say, hopefully that's a time to have a special episode on your podcast where we can talk at length about the book and the details of it.

Absolutely so completely agree. The manuscript is finished, but of course the evidence will continue. And that actually leads us, in a very natural way, to today's conversation. Because if outliers are structural features of markets, we need to understand what kind of systems could keep producing them.

But before we get to that, let me just add that we are really super excited to share the book, not least because our belief has always been that every portfolio should have a meaningful allocation to trend following, sized of course, for the investor's goal and the ability to stay with it. Because we cannot choose the next market regime, but we can choose to own a process that adapts when it arrives. And that's what trend following offers. This disciplined way to live with uncertainty, keep losses survivable, stay available for the big moves that shape long term results.

And there is so much evidence. I've always felt it's kind of weird that we work with a strategy where there is perhaps more evidence, hard evidence, than any other strategy, yet it's probably the strategy that meets the most skepticism and doubt that it continues to work. So, hopefully our small contribution, even though it took us two or three years to produce, will help in some shape or form.

Now, before we move on to a little bit of a trend following update, there was one piece of news that hit my screen this week, actually probably this morning. I thought it was kind of interesting. It's just a simple stat and that is, in Japan, there are now more than 100,000 Japanese people at the age of 100 or more.

Wow.

I thought that's kind of a…

It's got to be diet. It's got to be diet, I'm sure of it.

Maybe, yeah. So, anyways… Well that just means that they need a strategy for the long run. So, it could be trend following, actually.

Okay, now we need a short update on trend following. As usual, we record today, one day early because of my travels. And so, it's Wednesday, today, 16th of September. And so, the data of course is through yesterday, although I think the trend indices are probably from Monday night, as far as I can tell.

Now, the trend barometer, yesterday, finished at 52. That's a pretty strong reading and I think it's very well in line with the performance we’ve seen so far this month. And actually, although we're only halfway through the week, it's been pretty encouraging so far, at least from where I sit. And what we see, of course the main drivers are really kind of energy, fixed income, and most other sectors, including equities, struggling a bit this month, but still the energies and the fixed income markets are overcoming all of that.

In terms of news, not that we pay much attention to it in our trading, but still, what is kind of interesting and it kind of drives some of these fixed income trends is, of course, that the Fed… There's a 95% probability, according to Reuters, that the Fed will hike later today. The Bank of England is also meeting this week. They're not expecting to hike, but maybe the Bank of Japan will be, also, later this week.

So, lots of interesting things going on with US 10-year yields sitting at 5%. Over here in Europe, 3.55% on the 10-year Bund yield. And of course, what's really driving a lot of this is this surge we've seen, again, in energies. I noticed that Brent now is at US$108. That's interesting. But also, I saw a little piece of news this morning from Reuters, again, and I don't really fully understand exactly what this is, but it says, European physical cargoes for Brent is above US$130 per barrel. And I think that relates back to some of the things that we spoke with Adam Rozencwajg about, a few weeks ago on the show, where he's saying, when you lose a billion barrels in the energy world, it's going to, at some point, start to hurt. And maybe that's exactly what we're starting to see now. So, who knows? This could go on for a while longer.

Any observations, just generally, on your side, Rich, about what's going on?

It's an interesting period. The energy is certainly rattling up again. The Fed, no doubt they will put up interest rates. I think also in December, they probably will put it up again. So, it's a challenging period of time for the economies, but I certainly think it's going to be beneficial for trend following. So, let's cross our fingers Niels and see how it all plays out.

Maybe it's a good time to publish a book about trend following.

Okay, all right, enough.

BTOP50 this month. September up 2.73%, up 14.48% so far this year. SocGen CTA Index up 3.13% so far this month, up 14.59% so far this year, very strong. The Trend Index up 3.23% in September, up 14.67% so far this year. And the Short-Term Traders index up 36 basis points in September, up 6.15% this year. And, from memory, I think actually yesterday was a pretty decent day as well in our world, so hopefully more to come of that.

In the traditional world not so strong. MSCI World down 1.55% so far this month, up 11.65% for the year. The EAFE, meaning ex US and Canada, down 2.41% so far this month up 11.49%. Then we have the US Aggregate Bond Index also down 1.13% so far, and for the year down 1.07%, so, no offset, no risk mitigation there. The S&P 500 Total Return is down 1.23% in September and it's up 11.74% so far this year. Just want to make sure we get it right.

We have a question before we dive into today's topic and it is from John. Understanding we are typically outlier hunters, and we should let winners run as fast they go, I've been wondering if there is some quant measure where taking some profits earlier on more trades, which are up say 200%, outweighs the give back on those trades relative to the few which go up 3x4x etc., getting past the narrative. Is this testable? How much are we paying with profits given back to remain exposed to a fewer which go up more? Is there something in this idea? Consider that those profits also are not sitting there, they are closed trade equity, to reinvest, with limited downside, and new upside, increasing compounding.

Well, it's a very sharp question from John, but it does challenge my philosophy as an outlier hunter and I'll just explain. So, to answer John's question, yes we can, of course, test this historically, but the answer depends on what you're trying to achieve.

So, taking partial profits does convert open profit into closed equity that can therefore be redeployed. And it can also reduce drawdowns and produce a smoother return path. But we are not creating that smoothness for free. That's my argument.

So, we are reducing our exposure to those small number of exceptional trades that might drive much of the long-term result. We're exchanging some positive skew for a smoother equity curve. So, that's not wrong, but it's just a different strategy. It moves us away… well, me away from a pure outlier hunter hunting objective and towards more actively managed exposure. So, we could, of course, compare both approaches using the same entries, initial risk capital, reinvestment rules, all of these things to then examine CAGR, drawdown, MAR ratio, skewness, etc., historically. But the problem is we have very little statistical confidence in this history because the outliers we are depending on are so infrequent, and few, and far between.

So, these outliers are rare. And one extraordinary trade can change which approach actually looks superior. Simply one of our trades could do that. Now a backtest will therefore tell us what worked through the historical sequence, but it can't tell us reliably how many exceptional trends the future will contain or how large they'll become.

I've certainly done research back to over 125 years of research, and I've seen some absolute doozies in the past, and we haven't seen them in the 40 year track record we've experienced in our research, in our current research in our book. But I tend to view things fractally and therefore I think that, as the sample size increases we inevitably get these bigger and bigger events, either positive or adverse.

We get bigger drawdowns in the future. We also get bigger outliers, positive beneficial outliers in the future. And this is because as your sample size increases, you explore more scope over what I call the fractal tree of these markets. You get the big branches and the big… the entire tree, as opposed to the small twigs, and these things which we're seeing maybe in the last 10 years to 20 years or whatever. So, my worry is always what you call the type 2 error, missing those beauties, those outliers.

If we do what John suggests, we certainly can compromise that objective. So, as an outlier hunter, I'd say no, John, even though we can test this historically, it doesn't give me confidence because the statistical confidence I have with the number of outliers in my historical sample is always too little. The future's going to be different. I don't want to miss out on that potentially enormous trade that gives me the entire couple of years of whipsaws, pays for them, and some more. So, that that's how I'd respond to that one.

All right, well, let's move on to today's topic. And I know we should start with kind of an Introduction to the Santa Fe Institute. And I'll be completely frank, I had heard the name, didn't know much about it. So, I thought maybe before we dive into your side of the topics here, Rich, maybe I could just provide a little bit of background about the Santa Fe Institute and why its work matters to trend following.

And your chosen subject today is how this feedback look can produce trends, outliers, certain transitions, and why a trend follower responds to those movements without claiming to know their destination. So, we start our class today, Professor Brennan, and we're all strapped in.

All right, Niels, so look, I'm going to start with this El Farol bar. El Farol. I think that's how you pronounce it. So, the El Farol bar was a popular bar in Santa Fe, particularly on Thursday evenings when it featured Irish music. So, Brian Arthur, he was the economist who led the Santa Fe Institute's research into economies as complex adaptive systems. He turned the difficulty of deciding whether to attend that bar into a wonderful thought experiment. And I'll lead you through it. So here goes with that thought experiment.

So, imagine 100 people deciding whether to go, on Thursday night, to that bar when the Irish music was being held. If fewer than 60 attended, the evening was enjoyable. But if more than 60 arrived, it became too crowded and those who stayed home made the better choice. So, everyone can see the history of attendance. The problem, however, is in deciding what that history means for this upcoming Thursday. So, I'll run through it.

For instance, suppose all 100 people predicted a quiet evening this coming Thursday. They all therefore go and the bar becomes crowded. So therefore, the prediction defeats itself. Or suppose everybody predicted a crowd for Thursday, they stayed home, the bar is quiet and the prediction is wrong again.

So, how clever your choice depends entirely on what everyone chooses - not what you choose, what everyone chooses. So, even being contrarian can't solve the problem if everybody becomes contrarian together.

So, there lies part of the problem. We believe that we are using our models to forecast an outcome by ourselves. But where, in fact, it's everyone's models which are actually deciding the outcome. So, that's the first connection that Brian Arthur created that links to the markets. A market outcome is created, at least in part, by every participant acting on their expectations of that outcome.

The future price is not waiting to be discovered. Like the answer to a puzzle, it emerges from all the participant decisions, not just your own. So, the path is rewritten at every step as each outcome from the aggregate of participants influences what the participants therefore do next.

I mean, people might sit listening to us today and say, well, hang on, I mean, people make forecasts about the markets every day. So, what does this bar, El Farol, really tell us that's kind of different?

Well, Brian Arthur gave each simulated person several simple forecasting rules. So, one might expect this week to resemble last week. That was one rule. Another might use an average to determine another rule. Each person selected whichever of their rules had recently worked the best. So, the importance of this was the diversity of rules. So different participants interpreted the same attendance history in different ways.

This is how often we treat a market. Lots of people applying their different models to the same history. So, the actual attendance, therefore, became new evidence. So, once they made the forecast and then the actual attendance was observed, that became new evidence for assessing which rule to use the next time. So, if their prior rule was unsuccessful, that adapted to a new rule that was successful and vice versa.

So, when too many rules, however, predicted a quiet bar, too many people went to El Farol, and those rules therefore failed. Rules predicting a crowd became more attractive. And if too many people stayed home, the quiet bar rules regained their appeal.

But what he found, on average, was that the bar approximated 60 attendances, on average, over the long term. This was an average over the long term. But the popular rules underneath that average were continually shifting from going to the bar to staying away from the bar. They're continually shifting even though the overall result was an average of 60.

So, no rule became permanently superior because its use of that rule helped change the environment in which it operated. So, what Arthur's experiment concluded was that the market is not a fixed outcome with one correct answer. The outcomes change partly because people keep trying to solve it, and there's this reflexivity in the market.

Yeah, okay. All right. So that really what makes this or this bar, El Farol, a complex adaptive system, I imagine.

Exactly, Niels. The key is the reflexivity in it. So, it's complex because the result comes from many interacting decisions rather than one controlling hand. So, it's adaptive because people learn, they change their rules and they respond to what others have done. So, nobody plans the weekly attendance pattern of El Farol. Yet a recognizable pattern emerges from their collective action.

And markets work in a very similar way. Participants pursue different objectives, they use different information, and operate under different constraints. But their combined actions produce price behavior that no individual intended or controlled. So, a trend is one possible outcome. A rising price that attracts attention, triggers systematic rules and affects decisions about capital and credit. So, each response changes the environment facing the next participant and the movement can, therefore, begin to sort of feed itself, if you know what I mean.

Yeah. And I don't know about you, and I don't know if this is true, but I seem to remember that George Soros was kind of the first person who wrote about reflexivity. And I probably didn't understand it the first time I came across it, but I mean, obviously reflexivity comes into this story as well. So, take us a little bit deeper on that.

Yes. So, reflexivity describes the loop between beliefs, actions, and outcomes. So, people form beliefs about the system and then they act on them. Their actions therefore alter the system, and the changed system becomes the evidence from which the next round of beliefs is formed.

So, George Soros, he made this idea famous in financial markets. So, here's a good example. Consider a bank run. So, a bank may be fiscally sound on Monday, but a rumor spreads and depositors withdraw their money to protect themselves. Those withdrawals drain the bank's cash, and by Friday their collective actions may have helped create the very collapse they feared.

So, these similar processes operate also in the market. So, a rising price improves reported performance and often strengthens things like collateral. So, that attracts capital and credit, producing more buying and another rise. So, the future's not sitting there fully formed, waiting to be uncovered, participants are helping write it step by step, if that makes sense.

Yeah, so I wrote down a few steps here just to maybe to summarize some of this and correct me if I'm wrong. So, you have participants who form expectations, their expectations produce actions. Those actions change price and incentives. The new conditions alter participants expectations and the process can stabilize, reverse, or reinforce itself.

Yes. And you can see this feedback occurring. That's exactly right. And there's this continual loop that's occurring. It's this dynamic process.

And that feedback can result in trends, it can result in mean reversion. The collective patterns associated with that create all of the features that we see in the markets. And that's all endogenous news. That's all from the participants, all interacting in the market. There is no sort of news events driving this. So, that was an important point.

So that is important. And actually, I remember, and I'm honestly, I'm not entirely sure I fully appreciated the value of that. But we had Jean-Philippe Bouchaud on the show a few years ago, after he had released a paper about endogenous and exogenous impact on markets where a lot of people I think intuitively believe that oh, it must be a big news event that really drives trends, and so on, and so forth. But actually, the evidence is, as you say, that's not the case.

But when we have the setup, as you just described, obviously you're kind of a self-proclaimed outlier hunter. I just called myself a trend follower. I'm not maybe… I’m not entirely sure if there is .... I'm trying as much as I can. So, let's call it… Let's be generous. How do we, whether you're outlier hunter or trend follower, how do we respond to this future that is being written but hasn't been yet?

So, I believe it's by responding rather than predicting. So, from my perspective, an outlier hunter accepts that many trading attempts will be small because a few exceptional moves may drive much of the long-term result. Now, we can't know which market will produce the next outlier or how far that outlier is going to travel. But it depends partly on responses that have not yet occurred. So, we participate broadly, keep each initial exposure small, and accept failed attempts as part of that search.

But when a trend develops, we use a trailing exit that allows it to continue until the market shows us that the movement has changed. So, that's why I don't use personally fixed profit targets at all. The participants have not yet made the decisions that will shape the later stages of the path. So, it's not a future that's already written. This iteration step, this occurs across the path and we need a responsive strategy as opposed to a predictive strategy to manage that path, if you know what I mean.

Yeah, I think a lot of people over many decades were kind of subscribing to the Chicago School theory about markets being efficient. I remember my first conversation recording, many years ago with Andrew Lowe, who came up with the adaptive market hypothesis as far as I recall. And I remember him saying that yeah, markets can be efficient, just not all the time. But again, as I said, I think the accepted school of thought for many, many years was this efficient market hypothesis. So why does this kind of simple problem story, about a bar, challenge that kind of established economic thinking? If I can put it that way.

So, for example, imagine a fish stall in a food market. If the price is too high, some fish remain unsold and the seller therefore lowers the price. If price is too low, the fish sell quickly and competing buyers, they push those prices up higher. Supply and demand push back against that imbalance until the quantity offered roughly matches the quantity buyers want. That's sort of this equilibrium notion.

So, that's useful in many situations. But the problem begins when equilibrium is treated as the natural starting point for every market, particularly one in which participants differ; particularly in one where the participants are what we call reflexive and they learn and react to each other while a system's operating. So, that El Farol bar exposed that difficulty.

There can be no single forecast that everyone can successfully follow because, if everyone adopts it, their collective actions change the outcome and make the forecast wrong. So, Brian Arthur and the Santa Fe Institute, they didn't begin with the question of equilibrium, they began with a different question. Instead of assuming that the market had failed, found this correct balance, they asked what happens if different participants interact, learn and adapt, and we observe what emerges. And Arthur described the economy less as a machine and more as an ecology in which the strategies compete, lose effectiveness, and they evolve.

So, that was a big challenge made to the traditional orthodoxy of this equilibrium based system. So, saying, let's look at the market a different way, where reflexivity has something to say in this assessment of what the system is.

Yeah, and also, I mean, I still always have struggled with this idea of equilibrium in the sense that we could just look at the markets and see that they're not always in equilibrium, and we can see that the prices continue to change, which has always been my argument for, that maybe it's not a bad idea to have a career in trend following because I really don't think it's going to go out of fashion, despite all the narrative we've heard over the years of trend following being dead many times. But anyways, back to this story. So, essentially, were they claiming that the conventional economics was just dead wrong or simply wrong?

No, they're a bit more sort of open than that. But they believed equilibrium models, they're useful when the important relationships are reasonably stable. But they ask what balance would look like once supply and demand have adjusted, the forces pushing in different directions are settled? That's what equilibrium asks, what happens when that outcome is settled in a point of equilibrium? But Santa Fe asks, what happens while the pieces themselves keep changing?

So, imagine people arriving at a construction site with different plans. So, each can see what the others are building and they revise their own plan. Every decision changes the site confronting the other builders. So, there's no fixed blueprint telling us what the finished structure must become. And that's what equilibrium models assumed. But Santa Fe adopted this other pathway.

So, the two, the non-equilibrium model and the equilibrium model, these two approaches, they began in different places. One starts with the conditions needed for balance. That's the equilibrium model. The other starts with interacting participants and asks what patterns emerge from them. So those patterns can appear stable, like with Santa Fe, where it's settled at 60 on average over long-term. But the rules underneath them were flipping from go and stay away, go and stay away. But these rules can also change or suddenly give way as these participants adapt and learn.

Much to our displeasure and probably many of our listeners, trends don't continue forever. I wish they did, but they don't. And we find these periods of what we would define as consolidation, and so on, and so forth. So, does that mean that equilibrium still has a place somewhere in the narrative or in the whatever we call it, define it?

Yes, it does. It finds its place in that term ‘mean reversion’, which I'll get to. But markets certainly contain forces that resist movement. A high price can attract supply and reduce demand. High profits attract competitors which reduce margins. But economists call this negative feedback. So, in plain language, the system is pushing back against the change. That's what the equilibrium models are assuming. So, mean reversion fits comfortably into that framework.

But Brian Arthur showed that markets can also contain forces that reinforce change. A rising market may attract buyers whose buying pushes the price higher and attract still more capital. So, this is positive feedback in action. And this change begins to reinforce itself.

So, real markets, they contain both forces, negative feedback and positive feedback. But they're open systems. Capital flows in and out, participants, technology, regulations, instruments even, change. So, a market can settle temporarily into an apparent or a quasi-equilibrium without arriving at a permanent resting point.

So, the distinction matters to trend following. So, if every movement were quickly pulled back towards equilibrium, that would be a mean reverter's paradise. Persistent trends, however, would be very difficult to explain, but reinforcement gives us a rational foundation for understanding how some movements travel much further than anticipated.

So, what Brian Arthur talked about, and he did this for many years and was openly flogged by the economics community, he had this term called increasing returns. This was in opposition to the term diminishing returns (which a lot of us have heard of), and we've heard of diminishing returns, particularly in these equilibrium based models. But he came up with this term called ‘increasing returns’. So, most people are familiar, however, with the diminishing returns.

And a diminishing return is, imagine adding workers to the same fixed patch of land. Initially, each worker may add substantially to production, but eventually the land becomes crowded and each new worker contributes less. The fixed constraint begins to dominate and pushes the system towards a limit. That's this diminishing returns picture.

But Brian Arthur studied situations in which the opposite can occur. And for many years he was totally disregarded. Yet that explains so much of the tech race, the technological improvements, Amazon's success, Google's success. All of this is explained by this concept called increasing returns.

So, with increasing returns, expansion makes further expansion more attractive, and early advantage changes the environment in ways that reinforce that advantage. And I'll give you a really good example. Do you remember, Niels, the competition between VHS and Betamax tapes? I remember this clearly.

I do as well. I'll tell you a story about it when you're finished.

All right, so Betamax, if you remember, it was widely regarded as having advantages in picture quality. But VHS secured far broader adoption. So as more households bought VHS machines, video stores stocked more VHS tapes, and the larger selection made VHS more attractive to the next household. More users attracted more content, and more content attracted more users. This is increasing returns in action. Can you see how that's working out? Tell me a story about Betamax and VHS.

So, this is back in the late ‘70s, where maybe it was the pinnacle of my football career. And we were going on football exchange. I mean, we were not very old, but we were going on a football exchange to the US of all places, a fantastic opportunity. So, we go there and the coach of this team that we were kind of exchanging places with for the summer (we would go to them and they would come to us in Denmark), he happened to be the marketing manager of Sony in the US. And when the team came to Denmark, he happened to have to stay with my family. So, we got to know him pretty well – a super interesting person. He was also the press secretary, by the way, of Robert Kennedy when he was assassinated. So, some interesting stuff there. Anyways, that's the sidestep.

Anyways, he came to our house, and he started talking about this Betamax stuff. And lo and behold, I think when it came out, he organized that we could buy one of these machines. People don't understand that this was so new technology. It was like crazy. You could record a football game and see the same game even. Who would want to do it? But still, you could do that. And my father and I really into football, and I think the World Cup was coming on and all of that stuff.

Anyways, we got super excited. We got this Betamax machine and then, as you rightly say, we started to realize that there were not really that many movies we could rent. They were all on VHS. And we're kind of stuck with this machine that could only record, really.

You can see Sound of Music.

Yes, exactly. So, we kind of stuck with this machine that could only really record stuff, and we would have to watch the same football games over and over.

So yes, I am very familiar with this battle, and I chose the wrong horse to back or how should I put it? But not that we really had a choice to do it, but anyways, fun memories that you brought out.

This concept of increasing returns. I'm a bit blown away because Brian Arthur, in this book that I'll mention at the end of our podcast, Niels, a fantastic book that talks about the characters in the Santa Fe Institute. But Brian Arthur, he struggled for years to get this concept across called increasing returns. The economists of the day refused to believe that it existed. Yet it is so predominant in the tech stocks of today. And it's the reasons why Windows got such superior holding in the markets despite a lot of people challenging it, saying it was a more inferior product compared to others, but it's this concept of increasing returns that really does …

Today would call it network effect, Rich, is this what people call the network effect today?

Yes, it is. It's definitely network effect. But take for instance the success of Amazon. It's a path dependent effect. Because if Amazon didn't have its bookstores and didn't have its reach, there's questions to whether Amazon would be as big as it is today. When you look at the sequential things, or the things that have shaped the innovation, the process and the path of Amazon, you can see how there are certain things that accelerate their process.

AI wouldn't have come around today if there wasn't an Internet. So, internet was a precursor. AI is using a lot of those Internet resources. It's using its search functions. So, each thing is contingent on another thing. It's a path dependent process. And increasing returns is where adoption helps create the winner. So, in the VHS vs. Betamax, once VHS had a large lead, households owned tapes, stores carried inventories, producers served its larger installed base. So, a small early advantage created that environment that continued to favor the same technology.

And now we could maybe say that another thing might impact Amazon's returns or earnings going forward. And that is when we publish our book on Amazon, that effect will have massive impact on their future earnings… massive impact.

Okay. All right, well, let's try and put this into kind of our world, and maybe we should try and understand kind of the impact on financial markets, and trend following, of course, specifically.

Yeah, so this concept of this reinforcement process, it shows how movement begins to feed itself. So, a rising price improves the performance of existing holders that may attract fresh capital to those holders of that capital. It strengthens collateral, it encourages further buying. And, in the same way, falling prices can trigger stops, margin calls, and selling. So, it occurs in both directions, this amplification effect or this increasing return effect, it's not just a long only effect, it's long or short.

But this is amplification over the path it takes. So, this doesn't tell us which movement is going to become an outlier or how far it's going to travel. But it certainly explains why a market need not return immediately to equilibrium. So once participants respond to the movement itself, the process can extend far beyond its original cause.

Okay, all right. So, we've talked about two really important concepts. One is the reflexivity. And we've now also introduced increasing returns. I mean, are these the same we're talking about, just with different names?

They're related, but they describe different parts of the process. So, reflexivity is the loop between beliefs, actions, and outcomes. Increasing returns is a reinforcing process in which success or adoption changes the environment so that further success becomes more likely. But they often work together.

But more importantly, both reflexivity and increasing returns, it's about path dependence. So, it's saying that how we arrived at this point in time can affect what happens next. The eventual outcome is not contained in the starting conditions alone. It's actually created step by step along the path, if you know what I mean. So, this is why, when I am on an outlier, I have no predictive stance about how long it's going to go. There might be some reinforcing element in the very next iteration that's going to occur to keep that trend going, and going, and going, if you know what I mean.

Okay, all right, so maybe more technical in terms of a question would be, so what does movement that kind of feeds itself change when you make kind of actual trading decisions?

So, it tells me not to assume that a movement must end simply because it's traveled a long way. So, a movement that can reinforce itself will not necessarily do so. There's no guarantee there. So, forces extending it always compete with forces resisting it. There's always this sort of resistance and amplification effect going on. So, we just don't know what's going to happen. That therefore shapes a process.

So, I therefore spread many small positions across many markets because I don't know where the outlier will appear. I accept frequent failed attempts because most movements will not become exceptional. But when one persists, I use a trailing exit rather than deciding in advance that a particular profit is enough. So, this is why I believe trend following is a response, it's not a prophecy. So, the entry recognizes that movement has developed. The initial stop recognizes that it may fail. The trailing exit keeps us involved while it continues, and that's it.

Okay, but I mean, the interesting thing is that then the guys at the Santa Fe Institute took it a bit further, so to speak. They started applying or rebuilding this El Farol problem with real… maybe not with real money, but with money, and prices, and so on, and so forth. So, in that process, what happened to their, let's call it, artificial stock market?

Yeah, this was interesting. So, they created a simulated market populated by investors using different forecasting rules. So, those forecasts influenced whether each investor wanted to buy or sell. But their combined orders changed the price. And the new price then became evidence for judging which rules had worked well. So, when investors adapted slowly, the market behaved relatively calmly.

But when they searched more actively for better rules, calm periods were interrupted by clusters of turbulence. And (this is the important thing) price based strategies emerged. Trend following strategies started to become preferred. So, learning did not guide everyone smoothly towards one correct answer. But under some conditions, a search for better rules helped create the instability everyone was trying to understand.

This model was definitely simplified, a very simple model, but its message was powerful. So complicated market behavior did not have to come entirely from outside news. It was also being created and amplified by interactions among these participants. This is the endogenous source that we've talked about so frequently on our podcast. Niels.

Yeah, by the way, something I maybe should know, but I don't. When was this actually created? Do you know?

Okay, all right. Anyways, back to the story, so to speak. So, as trend followers, why should we care that suddenly they discover that price-based rules appear inside kind of a computer model?

Because in that simplified model they found that trend followers emerged with successful models. So, they became trend followers in that simple model. They adopted price-based forecasting rules as opposed to the other forecasting rules they were all given at the commencement of the process.

They're given a whole swathe of rules, but they settled or preferred these price Based rules and they were trend following rules that they preferred in this simulated simple model. So, some started using these models because those rules survive the competition between strategies. So, it's showing that in these simple models how trend followers actually emerge as dominant competitors in that framework.

Now, maybe it comes from here, I'm sure you'll know, but a lot of people say that once you have a system, at some point that's going to be gamed (I think there are many words to describe it). I think Arthur did that as well, as far as I know, but what did he mean with that term?

So, he wrote a great article about all systems being gamed. And so, what he meant was that whenever a system contains rules, incentives or boundaries, people search for ways to benefit from them. So, gaming doesn't necessarily mean breaking the rules. It can mean following them in a way that earns a reward while defeating their purpose.

I'll give you an example. So, this is related to Goodhart's law. So, this is when a measure becomes a target, people begin managing the measure rather than the result it was intended to represent. I'll give you a good example. Imagine a company rewarding staff for every new customer account. Someone discovers that opening several accounts for the same customer earns several bonuses. So, the account total rises, but genuine customer growth doesn't. So, the measure stops describing what management wanted to know. That's a good example how systems get gamed.

So, markets are also filled with benchmarks, risk limits, margin rules, expected policy responses. Participants organize around them, they exploit them, and they protect themselves from their consequences. And through that adaption they change the system itself. So, they're continuously testing the boundary of those systems. And it's a bit like a dam wall. If you are continually testing the dam wall, you will inevitably find the fracture or the crack in that dam wall and the water's going to pass through. This is how these systems operate. The agents in it are always trying to compete in that system and get the benefit or the edge. This creates this symptom that all systems are game, which Arthur didn’t put so eloquently.

Is that also how what looks stable can kind of transition into something that is… or transition as we know it.

Exactly. So, for example, a long period of stability encourages people to assume it's going to continue. So, under that circumstance, leverage rises, hedges shrink, and positions accumulate around boundaries that have held for years. But then a disturbance exposes that dependence. Then stops trigger margin calls, force reductions, lenders withdraw and buyers step away. These responses move price further and activate the next layer of responses.

So, an isolated break becomes a transition into a different market environment. An outlier may, therefore, be more than just a large ordinary fluctuation. It can be the visible result of relationships changing as participants respond to one another. And a boundary can look strongest immediately before the behavior built around it helps produce that break.

Yeah, who was it that says this, stability begets instability? Is that a Minsky moment?

Is that a Minsky moment? It might be a Minsky moment. It certainly sounds like it should be.

All right. Okay, okay. Now, kind of going back to one of the things we talked about earlier, and that's a little bit about this, if you have something that survived and thrived for 40, 50 years, that should be evidence of something.

Yeah, it is evidence, but I don't think we should take it too seriously. So, it comes from environments with particular participants, incentives, costs, those sort of things. These conditions, however, that's the history. And over that period we had these particular incentives, constraints, costs, all of these things. But those conditions, they don't remain fixed. So, a successful pattern, it does attract capital. Other participants then copy the rule. They anticipate its trades or they search for its vulnerabilities.

So that El Farol bar problem that gives us a subtle point. Average attendance can appear stable while the rules underneath it are continuously changing. A pattern can persist while the reasons producing it evolve. So, even though it looks stable for a long period of time, underneath, these processes might be competing against each other, finding the boundaries, finding the weak spots, and then the calamity occurs.

So, that's non stationarity in practical terms. Relationships change over time. More history, it does help us judge whether an idea has survived different conditions, but it can't turn an evolving relationship into a permanent law, if you know what I mean. So, the back test, it provides evidence, but it's certainly not a promise for the future.

Okay. So how do you avoid using, say, non-stationarity as an excuse to dismiss things that you see, results that you get that you don't like?

Yeah. So, it's by being disciplined about what the evidence can and can't tell us. So, historical testing, certainly it reveals turnover, drawdowns, concentration, reliance on exceptional trades. But it also rejects a rule that repeatedly fails to produce a behavior it was designed to capture.

So, the restraint applies when we project results forward. So, an average annual return calculated from one realized history is evidence about what occurred. It's not a physical constant that the market must produce in the future. So, skepticism should remain a priority for us and we should use historical evidence with a degree of discipline. It should never become an excuse for ignoring evidence we don't like.

Now, not that we would say that it's a good thing to completely compare playing roulette with trend following, but there might be something we need to talk about in that sense. And I think it was you who told me about someone called Doyne Farmer, and they had a project that provides a contrast with markets. And this was different. I mean, I think a lot of people had heard about Ed Thorp's adventure into the casino and there may have been others as well, but talk a little bit about the difference.

But the thing was, in this roulette experiment, the roulette ball doesn't learn, unlike our reflexive markets. So, the roulette ball can't copy the methods. It can't anticipate the bet or change its behavior to remove the advantage. But this is where a financial market can. And remember, in our podcast we had with Dave Dredge, we talked about Richard Feynman's famous quote about imagine if electrons had feelings, how difficult physics would be. Well, that's what these markets are. We've got these reflexive participants that learn, adapt. They produce a system response and then they respond to that system's response, this reflexive behavior.

So, Farmer carried this measurement, modeling and testing from the roulette tables then into the markets. But the object of study had significantly changed. In markets, learning, adaption, and competition are part of the system itself, which we didn't find in the roulette table. So, it's certainly a different system. And that's what these complex adaptive systems are. They're menaces to try and predict.

Yeah, well, I mean speaking of that, maybe on a very related note, I would say, we've just heard about all the things that can't be forecasted, so to speak. But yet, when we develop our models, we obviously try to look at history, see what a backtest looks like. But if market adapts, what kind of backtest, if we're being honest here, what can that really establish then?

Well, really a backtest shows what specified rules would have done through a known historical sequence. So, it reveals things like turnover, drawdowns, failed trades, all of these things, exceptional winners. But what it can't establish is that the same environment will recur in the future. So, I personally use it to test whether the rules produce the behavior I want out of my models. I don't treat it as a map showing where the next outlier is going to appear or how large it will become.

Okay, well that kind of leads very nicely into perhaps finding out a little bit more about, as a trend follower, as an outlier hunter, how do we prepare for a future that we cannot map, so to speak?

So, the best thing we can do is reduce how much needs to be predicted correctly using our system. So, we search broadly. We keep each individual loss small. We use trailing exits. So, an exceptional movement has room to continue. The rules need to remain mechanical. They respond to changing prices without being continually rewritten by the trader.

So, research is where we reconsider the design. Daily execution is where we follow it. And complexity gives us a reason to be modest about prediction. It's not an excuse to abandon discipline.

When you've had a conversation lasting an hour and 10 minutes, it's a really unfair question maybe to ask. But if people have listened to us all this, and you would say what are the three things that they should take away from our conversation? What should that be in your opinion?

Here's a summary, Niels. So first, market participants are inside the system, so their beliefs and actions help create the outcomes they are trying to anticipate. The second is that some market movements actually feed themselves. So that helps explain these persistent trends and exceptional outliers, but it doesn't tell us where the next one's going to appear. And third, history is evidence, but it's not a promise for the future. So, our response is, therefore, to participate broadly, keep failed attempts manageable and leave room for the movement that travels much farther than anyone expected.

So, that's why the market is not a machine, in my opinion. And I'm much more like Arthur. I believe it's an evolving ecology of participants that respond to a world that their own actions are helping to create. And finally, if we all have concluded that El Farol bar is going to be quiet this Thursday, it might be wise to make other arrangements.

Well, for example, to listen to an extra episode of Top Traders Unplugged instead of going to the local bar. Now, I know you've written about this obviously on your ATS website. And there are, is it a couple of series or was it just the book that you wanted to…?

There’s one big series devoted to this topic. The book I'd like to plug here is a book by Mitchell Waldrop. It's called Complexity and it's called the Emerging Science at the Edge of Order and Chaos. So, it gives a wonderful background to the Santa Fe Institute and it brings people, such as Brian Arthur we talked about, John Holland is another significant person to understand, and Stuart Kauffman to life.

So, it's not only a story about science, it's the human story about them and the struggles they had in getting their eyes and ears across. And if this discussion we've had today, Niels, has intrigued any of our listeners, that's a great place to start.

Sure, absolutely. And thank you so much, Rich, for taking us into this kind of scientific way of trying to understand something that we so casually just describe as trend following. There's a lot more to it, as people will find out, of course, when the trend following manifesto gets published later this year, we hope. So, thanks for that. I mean, we may not have completely solved the seating problem at the local bar, but we've certainly found a way to better understand the crowd a little bit, which can be very useful in our world.

Now, before we wrap up today, let me, as usual, always encourage everyone listening to head over to their preferred podcast platform, give a shout out, leave a rating and review as appreciation for all the preparation that goes into making these conversations happen every week, especially to all of the wonderful co-hosts that we have on this series and the host on all the other series. So that will be fantastic.

Now, I usually say next week I'll be back, but actually next week I won't be back because I will be traveling. So, I think the next couple of weeks you will hear the pleasant voice of Alan Dunne with some inspiring guests from the usual lineup of amazing co-hosts. I have a feeling it might be Nick Baltas next week. So, that's going to be super exciting without a doubt. And that pretty much wraps up this edition.

Even if I'm not on the show next week, I can still take some questions that I can put to Alan, and the usual email is info@toptradersunplugged.com for you to send it to. So, from Rich and me, thanks ever so much for listening. We look forward to being back with you next week. Make sure you mark your Christmas calendar shopping list already with the Trend Following Manifesto. And until next time, take care of yourself and take care of each other.

Thanks for listening to Top Traders Unplugged.

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

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

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

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

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

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

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

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