🐢 Turtle Talk Podcast: Episode 007 – Fractals, Expectancy, and Survival

[Music] Welcome to Turtle Talk. Dive into the fascinating world of timeless trend following strategies and the pursuit of the elusive outliers made legendary by the Turtle Traders. [Music] Join your hosts Adam and Rich alongside our esteemed co-host Jerry Parker as we uncover the powerful principles of classic trend following. From counterintuitive practices to the mechanics behind capturing market outliers, each episode offers a masterclass in navigating the financial markets. Expect in-depth discussions, expert insights, and actionable advice to help you master the art and science of diversified systematic trend following Turtle Style. Tune in every month as we explore the dynamic and ever evolving world of trend following, equipping traders and enthusiasts with the tools to succeed. And now without further ado, let's hand it over to your hosts Adam, Rich, and Jerry. Welcome back to Turtle Talk. Episode 7. It's Friday the 26th of September 2025. And I'm Rich Brennan coming to you from Brisbane. With me as always is my fellow Aussie Turtle Adam Haverliffe calling in from beautiful Scotland Island in Sydney and of course the legend himself Jerry Parker. Jerry, before we dive in, I've got to ask, how did you enjoy Europe? I went a great time in France and Italy. Good food, good people, lots of nice people we ran into who were really still being nice to Americans. So that was a lot of fun. And so it was all good things must come to an end. My dogs and birds missed me and I'm glad to be home again. I'm going back to the hedge Nordic Roundtable in November again. So it's my second time in Sweden and meeting up with my roots again. And so pray for me when I get in front of all those powerful. I will. I will. Yes. Best of luck. And Tom and I managed to keep the weight off with all of that beautiful Parisian food. Well, not as much as I needed to, but I'm fasting now. So back to my Fadi Weight, Richardson. Back to the treadmill. Yes. Oh gosh, yes. Yes. And Jerry, you're calling in from Florida today? No, yeah. No, no, Tampa, Florida. Yes. Nice. Nice weather here. All right, Gents. Well, it's great to have you both here. And for today's pod, we're going to keep things tight and focused. No guests this time around. Just a three of us digging into the big themes. We'll kick off with the battle of the trend following indexes. Check out some standout markets and charts in what's moved the needle and then get into turtle tidbits where we cover fractals, expectancy, the cutback rule, loose pants and even a seeking alpha article on high-voltrend following. Plus, we've got some excellent audience questions in Shell Mail, which we'll go over. But let's start with the indexes. So Adam, over to you. Okay, thanks, Rich. Time for the battle of the trend following indexes. It's a monthly pulse check on how classic trend is performing. And over to the report on the AussieTurtles.com website. We had a nice little bump last month. And that is really, really nice to see. And notably, it's significantly higher than our fellow travelers being those other indices that we like to benchmark ourselves to. So a nice return of 4.4% for August. And if we have a look, I think we're traveling quite well for September as well. But obviously that will be finalized in the next couple of days and then report it on shortly. But we were talking last month about the recovery. And I know we were sort of hatching our eggs and all that. But we were discussing, was it a VU or a W? And obviously it was fingers crossed for investors and market participants for a V. And it looks like it is indeed shaping up as a nice V-shape recovery. And interestingly, I think our recovery here on the classic trend index is looking a lot nicer than the other indices. And we're going to dig in and have a think about why that is. But it is nice to see that recovery taking place. And then just visually here, you can just see the difference in the accumulated returns between the two different products. So the K-GAR remains far, far apart from the other indices. And then not only are we sort of demonstrating that standalone compounding K-GAR maximising performance. But on a risk adjusted basis, we're also showing a huge out performance. And looking here at the mar and the sharp, I mean, we're running numbers that are more than double, even triple what the other indices are demonstrating over that analysis period. So it's all looking pretty good. What are your thoughts, Rich? Yes. So we're going to be talking about a few of the topics today, which lead to the conclusion about how our process maximises K-GAR. So there's some good topics coming up, which we'll be discussing. But as we can see here, when we look at the VAMI chart, we see that the classic trend index had its high water mark in February 2025, while all the other trend following indexes had their high water mark in about April 2024. So we can see from this VAMI chart, when we look at the table of statistics, we can see that there are only 4.6% away from another high water mark. So the good thing about the classic trend index is this lifting power and a quick recovery. So with another good month or a couple of months, we might be back at that high water mark again. But in these topics, we'll be discussing today. We'll be going into the mechanics of why we have this fast recovery and how we address maximising compound annual growth. So it's going to be a good one today. So are there any thoughts from you about this month's battle of the trend indexes? Well, it's good that you're still looking really good for the classic and it just continues to show the power of holding on to those outliers and letting the profits run and not trying to get out too quickly and manage the smoothness. So we're not going to have the smoothness. We're going to have some ups and downs and some volatility and maybe some even bigger givebacks, dry downs from profitable trades, but the profitable trades themselves will be larger. So all of these good characteristics that we keep talking about are manifesting themselves in this excellent outperformance that we've had. So let's just knock on wood and help we keep pin you. So yeah, great. Let's shift gears into the price action itself, time for what's moved the needle. So over to you, Rich. Great Adam. So this is what's moved the needle and this is where we look at standout markets and the price action driving momentum. So Jerry, you've got the JGB chart, the Japanese bonds, the copper chart and the LME copper chart lined up for us. So why don't you kick it off? Can you go into each of these charts and let us know what your thoughts are? Well, I've been having a tendency to focus on charts sometimes that just teach a lesson and I think the JGB is a really good example of a choppy market that with the loose pants and with the long term systems, sometimes the best you can do is just not lose too much money and do in frequent trading. And this last breakout of the JGB looks like January to the downside. It had a tremendous, nice profit built up and it gave it all back very quickly with a big spike, but I didn't get knocked out. So not only am I loose pants on my trading exit, my breakout exit or moving average exit, but also on my stop loss. I don't have my stop loss set too tight and too close. I don't really take small losses, I take optimal losses. They're sort of small, but not too small. And so I don't get chopped around. So I think that this chart is a good example of sometimes the best you can do is just hang in there and not get out unless you're really the system is really measuring whether or am I really compelled to get out here. So we had a nice profit. It turns into a loss. We still don't get out. The market goes right back down for still then at the original ATR and original sale price. And so I think this is another great example of what it's like sometimes in a really choppy market if you play your cards right and have a system that's to lose pants. I know we're going to talk about loose pants. but in this particular situation, the loose pants, not only does it help you stay in a long-term trend, like it bounced out, that the trade maximizes until a big outlier. Sometimes it helps with choppy markets where you just not do it a lot of trading. You're able to stay in there and make the best out of a not a great situation. Yeah, if I look at that chart, and I see if we made tighter pants here, it would be in and out all the time. It'd be massive overtrating for no age, so yeah, I can totally see where you're coming from. And the copper charts, I wanted to show the two copper charts because believe it or not, I'm the king of diversification, right? But I just started trading LME copper. I've been trading New York copper, and I really wasn't thinking too much about it. I trade Brent Crude and WTI Crude and Golden Silver and Swiss and Euro markets that are sort of correlated, but I mean, you can't get two markets, can't become more correlated than LME copper and New York copper. But because of the tariff situation a few weeks ago, the New York copper had a much different sell off, a violent sell off, and the LME copper sort of stayed in there and stayed remained flat to long. So I was listening to a podcast. Pretty sure it was Moritz and he mentioned, or he got me thinking about this copper somehow, forgot how it was. And then I just said, okay, I'm adding it back to my portfolio and we're trading two different coppers. And he came up with this great saying the other day, I was messaging with him, because same saying but different. So they are same saying, but they're different. And that's what we go for. Like no one cares more about diversification than classic trend followers. And trend followers in general. But for us, we don't really care that much about it. I trade more markets than almost anyone, but I don't care that much about correlation. And I'm just looking for the outlier, I'm hunting the outliers. And so can I get an outlier in LME copper and not New York copper? Sure. Can I get an outlier in one clue to not the other crew? Yes, of course. So that's what we're kind of doing. We're saying, yeah, it is almost the same market 99% of the time, but sometimes we'll get this outlier. And the charts do look different. And it is worth adding it into a portfolio. And for me, it's 400 markets. So even if it's an exact duplicate of New York copper, it's going to have almost no impact on me unless it's a big huge outlier trade. And that trade lending sugar, a little coffee, as well as New York versions as well. So but you know, oftentimes on podcast and stuff, we do hear CTAs talking about diversification and not trading markets and trying to find markets that are not correlated with each other and find diversified markets. But I would like to do that as well, but I'm not going to give up on markets that to produce an outlier, even though they're very similar to some other market. And we see it happen many, many times. So it's going to happen again. So Adam, over to you, you've got charts for a number of charts you want to go through. Can you step through them with us? Yeah, sure. So it's a bit of a metals special this month for me. And I know we spoke about platinum a couple of months ago and I was getting pretty excited because it had a sort of 40% rally. And then the moment I talk about it, of course, we get a correction and it pulls back $150. And you know, oh, am I going to get stopped out here and look like an idiot because I've just been talking about platinum and how good it was going. And luckily the pants were loose enough and the trade remains on. But since then, we've had a real explosion in the metals and silver in particular sort of just crept up over 34, 35. And then I guess when we were overseas, rich and not really looking at the screen too much, it's just absolutely skyrocketed. And I can see here this morning, it's opened over $45. And we're getting into uncharted territory here with silver, very close to that all time high of a circuit $50. And you know, the way I look at that is less overhead resistance, the potential for, you know, it's already an outlier, I guess, in some sense. But we've got a potential for it to move into that, you know, real wilds sort of territory. And good support across the commodities or the metals complex. So, you know, gold, silver, platinum, they're all up 40 to 50% for the year. And we've got 10%, 17%, 15% for the month, you know, which is obviously a big contributor to classic trend index performance for this month, I would assume. Certainly for my fund, we're having a great month and the metals are a really, really big part of it. And then moving off metals into cattle, I just wanted to highlight the one year chart there, which is sort of reflective of recent performance for us. And I guess Jerry would also have this, you know, contract on. But then zooming out to the five year chart. And you can just see there the things being absolutely a fantastic performer over five years. I don't think it's been, you know, like a cocoa-like contributor to performance. But certainly just contributing almost every year, delivering something to the program. And I was just really impressed by that five year chart. Now they're wonderful charts, Gents. And look, we're a bit spoiled for choice right at this moment for some of these massive trends we're riding. So, it could be shaping up for a good end of year. I'm keeping my fingers crossed on that one. But, look, that's it for the charts. So now let's move on to our next segment, where we can deep dive into some of the great topics this month. So over to you Adam. Okay, so time for turtle tidbits, where we tackle the bigger themes. We've got some heavyweight topics today. So Rich, you're leading us off with fractals. Do you want to discuss the topic and lead us through it, please? So I've been putting pen to paper over the last couple of weeks with my favorite topic, which is complex adaptive systems and fractals. And I just wanted to bring this topic up to describe a different way of viewing the markets as opposed to the traditional Gaussian way of viewing the markets. Because in most traditional models where they're focusing on things such as variance at risk, standard deviation, or the levels of diversification, how much is enough, they tend to be speaking in Gaussian terms where you find that with a greater sample size, with greater bets, things start converging towards a mean. But if the markets are fractal in nature, which I'll discuss shortly, you'll find that there's this property and fractal systems called scale invariance, which means that we never get this situation where we get a convergence to the mean. We might get a very slow degradation to the mean, but in some systems where they have these dominant heavy tails, you find that they actually never converge to the mean. Now, what that means is it totally therefore changes the rule book. When we look at these systems, any complex adaptive system, which has a finite lower bound, in other words, a zero bound, but it has an open-ended possibility and it's an open system, we find that inevitably these systems become complex adaptive systems where we get fractal development occurring within these systems. And this is for those systems that don't have a central coordinator. So there's nothing centrally coordinating all of the actions of the agents in the system. So in the systems such as natural systems, in our human bodies, in the universe itself, and in the financial markets, there is no centrally governing body that controls what all of the agents are doing. So you've got to look at things from a ground-up perspective from this lower finite bound where the smallest agent interactions occur. And then as we progressively put in more and more agents, we find that there are structures to start evolving in that system, which are fractal in nature. So to understand this, if you could imagine the most fundamental unit of a financial market is the tick. And that's where a buyer and a seller come together and they transact price and a tick is generated from that activity. And what we find is that when we start getting more and more agents coming in and more and more ticks occurring, there is two phenomena being created. There's what we call positive feedback being created where we get an amplification of price or an increasing direction of price either up or down. Or we get what we call negative feedback, which is a suppression of that movement. Back to, let's mean reversion. Or when it's trying to bring things back to an equilibrium state. Now what you find when you add more and more agents into the system, all with their different models or directing price in different ways with their models. This is putting a force into the market which is creating this positive and negative feedback cycle. What we find is that with any complex adaptive system that has a finite lower bound and an open-ended upper bound, we find that there is not a perfect elimination between mean reversion and positive feedback. There's no perfect equilibrium created. What we find is that at the lowest level we find that mean reversion tends to dominate. As soon as price starts moving, high-frequency traders, liquidity providers etc. start imposing forces to try and bring it back to equilibrium. As we scale up in the fractal system, we find that the positive feedback creates structures or a bias in the price series. That bias feeds on itself. You get structures being built on structures being built on structures. What you find when you actually look at the price formation in these financial markets, you'll notice that they have a fractal structure. If I look at a one-minute chart, for instance, it's very hard to discriminate between a one-minute chart and a monthly chart. They look very similar, but they have a fractal structure. In other words, they have what we call a scale invariant structure. The structure persists at all scales. What I mean there is that in a Gaussian system, when you start zooming into a Gaussian system, you find that the structure starts dissipating down to zero as you get more magnified and more magnified as you zoom into the system. This is different to a fractal system where you find structure at all scales, from the macro scale to the meso scale to the macro scale. We can see that in our charts. When we look at the minute chart, we see a lot of mean reversion. We see some trends in that chart. As we scale up, we find that the positive feedback of the trends that exist in that structure, they start actually amplifying themselves. The way we understand this is that more and more participants start acting on that structure as you scale up. At the lowest level, you might get high frequency traders, intraday traders. As you scale up in the system, you start seeing trend followers coming in, you start seeing institutions coming in, you start seeing larger players coming in who are acting on the structure, which is this fractal structure, which is amplifying the structure. No matter what complex adaptive system we look at, we see this occurring. The good example is when we look at the formation of a human being from a zygote, which has no differentiated structure, to a human being itself, who has differentiated structure, we see that this is a complex adaptive system following these fractal rules. At all scales, we start seeing the fine-scale structure being amplified into organs, arms, feet, head, etc. We find that structure starts dominating as you scale up. The same as when we look at the universe, which they anticipate started from a big bang and there were very small deviations, quantum deviations, which created the structural filaments upon which amplification occurred, to as we scale up, create galaxies, planets, stars, all of these things. This positive feedback is creating this structural fabric, which amplifies itself. No matter how much mean reversion or noise we get in the system, we find that as we scale up these structures, the mean reversion and noise starts dissipating because they cancel each other out. They're not amplifying, but we find that positive structure starts compounding. This is this multiplicative process in the structural formation of a complex adaptive system, and we find that the positive feedback amplifies itself, trends get bigger. So if we imagine these financial markets where we have this structural, fractal structure, you'll recognize that it doesn't matter how big and scale we go, there is always larger fractal structure ahead because of this positive feedback and amplification, which means that when we look at things such as diversification, the Gaussian assumption is, oh, as you add more bets, things start smoothing towards an average or a median. But we find in these complex adaptive systems the larger scale we get, the more chance we have of actually aligning our systems with the dominant fractal structures of that system. So if you can imagine if I've got a tree as an example of a fractal system, a tree which has a large trunk, many branches, many leaves, many smaller tweaks. If I put a very small sample over that tree, it is highly likely that I'm only going to get a few leaves, a few small tweaks, but you'll find that the relationship in that sample has this fractal structure and you'll find that relative to the sample you're taking, you'll still get some dominant contributors and small scale contributors, though you might get some leaves, some larger tweaks in that small sample. They will, the scale invariant nature of fractal systems means that there will be outlies in that sample, maybe 5 to 10 percent in that sample. But as you scale up and you get bigger and bigger and bigger and you start increasing, in compacy and tire tree, you get access to some of the bigger fractal structures such as the trunks, the major branches. These are the things that are dominating financial returns. And this is why in complex adaptive systems, we get this Pareto principle, the 80/20 rule you might have heard about. This is because there are power laws in this fractal system and this means that we find that there's a small percentage because it's scale invariant, no matter what scale we go into, there is a consistent small percentage, 5 to 10 percent, which is dominating the overall returns of that sample. And that's because of the scale invariant, fractal base structure of the system. So I just wanted to bring that to the attention because when you start looking at things from complex adaptive systems and you start seeing that the major dominant drivers of what moves price is the interactions that occur between buyers and sellers, you can reduce things to a mechanistic explanation and you can start seeing when you add more and more agents together and we start getting these forces being exerted or these impacts being exerted, some canceling out, some amplifying themselves, you can see that as you scale up with this structural mechanistic interpretation, you start getting dominant structures starting to dominate the overall performance. And that's why if I look at a one minute chart and I compare it to a monthly chart, when I look at a monthly chart, I see these major trends in the monthly chart. Now this is what most participants in that much larger scale have been trading towards. They're being focused on these long term trends, these the major causative drivers of trends, which we interpret as things such as macro and fundamental reasons for why these major price moves occur. But when you look at it from the fractal structure of these complex adaptive systems, you start seeing, ah, this outcome is actually because of this positive feedback reinforcement that occurs and as you scale up, the positive feedback starts dominating the mean reversion and noise. And this also explains why trend followers do far better in the medium to long term space than in the short term space because in the short term space, the fractal structure is dominated by mean reversion. But when you start scaling up, you start getting access to these major branches in the fractal system, which is really dominating returns. This is where the positive feedback concentrates and dominates in this auto correlation or serial correlation mistakes. So look fascinating stuff. You've written a quite a detailed blog post on atstradingsolutions.com, the fractal feedback asymmetry of markets. I recommend that anybody who is interested have a look at that article. And I think the summary is basically that the negative feedback is fragile, but the positive feedback system is exponential wealth creation. And I think that as trend followers, we certainly want to be in the latter, ah, compounding an exponential wealth creation mode rather than the fragile mean reversion mode that some traders are still stuck in. Let's move on to the next topic, expectancy versus path dependence. What really matters? Now, Rich, do you want to compare expectancy against path dependence? Yes, I do, Adam. And this is another post, and this is a lead on from the fractal nature of markets. And it's a function of dealing with complex adaptive systems. And what we find is that when we start looking through this different lens, fractals, complex adaptive systems, we see that the typical statistics that we've used to assess performance measures are invalid. a good example as we all know. sharp world thinking. That type of thinking is inappropriate for a fractal system or a complex adaptive system. We see that Gaussian assumptions of how frequently and omelies should occur are much less frequent. One in every 13,000 years of a five sigma event, but we've seen about 30 or 40 of them in the last 100 years greater than five sigma events. So we know markets are doing something that's non-Gorsian. We know that they have these fat-tailed properties. And this changes the entire ballgame in looking in terms of statistics versus what I call engineering. I would say that trend followers are more concerned with an engineering solution than a statistical solution. Because when you look at the statistical properties of markets, there are so many assumptions built into those statistics that often they're incorrect. And a good example is the expectancy equation. So our listeners would be aware of the expectancy equation. And most traders are taught, provided they have positive expectancy, that's all they need to do to have a winning system. Now that is fine in a particular class of systems that we're not dealing with in the financial markets. And those systems are called Ergodic systems. However, the reality is these financial markets are non-Eurgodic systems. So think of it this way. In an Ergodic system, the average across possibilities is the same as the average over time. So if we flip a fair coin forever and your long run results will converge to 50/50, the path you take to get there does not matter. And this is how the expectancy equation works. But in a non-Eurgodic system, the path changes everything. When you compound returns like we do in the markets, the sequence of wins and losses can permanently alter your trajectory. You can get wiped out before the average ever shows up. So wealth path paths in markets are non-Eurgodic and that's why expectancy on paper can look positive, yet in reality, compounding and absorbing barriers drag results towards zero unless your process is engineered to survive. So we know markets are non-Gorussian. They have fat-tailed properties, which means the traditional statistical tools we are taught to to use often leave us wanting. So take the expectancy equation which we will put up on the screen. Traders are told it's a holy grail. If you have positive expectancy, you have an edge that leads to wealth. At least that is the promise of the equation. But in a world of fractals and complex adaptive systems, where we have a finite lower bound and an unbounded upside, expectancy has serious shortcomings. So look closely at that expectancy formula and notice what is missing. There is no reference in that formula to the sequence of returns. It assumes the world is organic, that the average works out over infinite time, but no absorbing limits exist in it. And we'll talk about what those absorbing limits are. But in real markets which are non-Eurgodic, the path matters and the sequence matters. And it's not just expectancy, the same issue undermines a classic risk of ruin equation which I'll also put up on the screen. That is often touted as a universal truth. So both ignore the path dependency of compounding wealth. So I'm going to run a simple experiment and show you the outcomes. And this experiment is going to start with $100, where you flip a coin. Heads, you win 60%. Your capital grows therefore from $100 to $160. Tails, you lose 50%. Your capital falls to $50. Now on paper, expectancy looks great, a positive 5% edge. But now if I run many wealth paths over time, and I'll do this by hitting F9 and we'll see it updating over the course of time, each time I hit F9, I'm running a new sequence of events. But you'll notice that every single wealth path decays to zero over time. Not one compounds to wealth. Why? Because there are two destructive forces that are always at play in path dependent systems. The first is this absorbing barrier. So if a sequence of losses drives your capital to zero, you are out of the game. No money left, no future compounding. That barrier is fatal. And the second is compounding drag. When returns are applied conditionally to a non-agotic path, losses bite harder than wins lift. From 100 for example up to 160 on a win, then down to $80 on a loss, keep flipping and the negative compounding pressure pushes your trajectory towards zero. And we can see that in these examples as I'm hitting F9. So despite this positive expectancy, the wealth path is compromised. Statistics tell you that you have an edge, but the lived wealth path actually says otherwise. This is why survival, not just statistics, is paramount. And the traders job is not to trust expectancy, but to engineer a process that can withstand a symmetry and sequence risk. That's why I say a trend follower is more of an engineer than a statistician. So we design systems with small bets, wide diversification and rules that cut losses and let winners run. And this process generates enough lifting power to dwarf the ever-present drag of non-ergodicity. So expectancy may look like the holy grail, but in practice it is mute on the one thing that really matters, the path. And in trading, the path is everything. So now let's flip the problem around. Instead of letting chance dictate our path, what happens when we engineer the outcome with a positively skewed process? So a classic trend follower does not need to win often. In fact, the typical profile is only about, say, 30% wins and 70% losses. On paper, that sounds terrible, but here's the key. The winners are on average three times larger than the losers. And this builds in the necessary asymmetry to change the wealth path. So how do we make that work in practice? So in this example I'm giving now, by cutting leverage right down, in other words, by placing very small bets that are as small as possible that takes a sting out of the inevitable string of losses. Small bets make losing survivable. Then when the big winner does show up, that positive skew takes over and the asymmetry flips some math in our favour. So now instead of compounding dragging us down towards ruin, compounding accelerates a wealth path upward. And this is the difference between statistics and engineering. The statistical mindset says, if expectancy is positive, you have an edge. But the engineered mindset says, structure the process so that the edge actually survives in the real world. So by deliberately embracing small bets and positive skew, with our massively diversified portfolios that hunt for outliers, we turn these wealth paths from fragile into resilient wealth paths, and we transform compounding from a drag into a tailwind. And that is what engineered trading looks like. You need extra asymmetry to lift yourself. And so when you look at the rules of a trend follower, what are they? One is small bet size. Why small bet size? Well, what you find in this equation of path dependence is that the higher volatility you have, the greater the impact on this degradation of returns through compounding drag. The relationship between small bet size and the volatility drag is one of the fundamental engineering reasons to maintain this path dependent return. Another thing, why do we maximise as wide as possible? We maximise as wide as possible to get these fractal branches that exist in this system. We don't know where these outliers exist, but if markets are fractal and they've got these major dominant branches driving compounding, our nets need to encompass the entire system to be able to drive up that positive asymmetry to exceed that expectancy problem with the compounding drag. All of our rules are based around engineering. It's not based on statistics, because statistics leads us astray. It's about how do we achieve a path of returns that survives long enough to compound. And as we'll get into more of these topics, you'll see that the principle of survival comes at the forefront of everything we do. It's not about statistics, it's not about maths, it's not about being a quant, it's about engineering a survival path. We do that with small bet sizes, we do that with stop losses, we do that as Jerry will get onto with a cutback rule. Well, all of these are engineering responses to ensure that we've got a path of survival, which maximises compound annual growth rate. The geometrical returns in a non-agotics system. Right. I think it's good for you to make sure you're always separating because until the very end I thought that my equity was eventually going to zero. But you sort of eventually said if you're writing these outliers and you're taking small losses and diversifying, and you're having to say symmetry, then this is how you overcome all these bad things. And I just didn't understand, I probably don't understand exactly what you mean by expectancy and how asymmetry overcomes expectancy. So because when we run our back test, we all have an expectation per trade. It is just a number. And if it's a high number, that's good. And so I want my expectation per trade to be high. But so that's the issue I had, which I was wondering like, what's wrong with that? And of course, if you're having issues with drawdowns or whatever, you could do the cutback rule or you can keep following the system and hunt these outliers. But I think we're all having a high expectation per trade is a good thing. And I didn't really understand why, which you meant by it. You obviously didn't mean that. There is no problem having a high expectancy. But the expectancy must exceed the negative asymmetry that the wealth path is curating. Yeah. And it's counterintuitive because 60% is higher than 50%. But you're saying that because of the compounding effect, the 60 is not enough relative to the 50. You have to have some ratio between the winning percentage, not the winning percentage, but the winning amount. And the losing amount. That can overcome the negative of the loss. And path dependence is the only way to overcome that. And what I mean by path dependence, you have got to manage that path to prevent the negative drag or that absorbing barrier I talked about, taking you out of the game. Okay. How to trend followers intuitively know this and do this without knowing about non-agotic systems. Okay. So classic trend followers, classic trend followers do it naturally with their process. Now what I'm arguing is that other trend followers aren't doing it. Now what I'm saying is that classic trend followers, they ensure that their outliers are not clipped. If you start reducing the volatility of your upside, the beneficial upside, you are naturally going to change your wealth path to a non-agotic declining wealth path over the long term. If you have a smooth return stream that doesn't harness a power of outliers to lift yourself out of this negative compounded geometry, you're going to find that over the long term you are more exposed and more fragile. A classic trend follower, because when you start putting this growth path into a positively skewed system, you find that suddenly, you know we talk about the two-edge sort of compounding, you know, there can be beneficial compounding, there can be adverse compounding. We often talk about the two-edge sort of compounding. If you are losing money and compounding takes hold, you start exponentially losing money. If you are making money and compounding takes hold, you start exponentially making money. There is a sweet spot, which is not the expectancy equation. It's the asymmetry, the geometric path that needs to be exceeded to have this beneficial compounding. And I'm saying classic trend followers with their process naturally fall on this path that exceeds all of the negative outcomes, the absorbing barrier and the negative drag of compounding through their process, because we are adopting positive skew, we're maximizing outliers, we're giving every opportunity for the step-ups, the outliers, to massively overcompensate for these negative drag effects associated with it. Yes, with the typical classic trend follower system, have a higher average trade than the managed future security. Yes, yes I will. So we like our expectation for trade because it incorporates the outliers in the asymmetry and so if everyone treated like us, we wouldn't be complaining about the expectation for a trade because it would be much higher. Yes, Geron. So what we find is that because we've got a high average trade win against losses, that more than compensates for the effect of those losses, we also find that our small bet size, more than compensates for the asymmetry in the geometry of returns, our diversification, more than compensates because we are massively diversifying to achieve these significant compounded events or these massive outliers that drive our returns. Everything we do is actually an engineered outcome without maybe us realizing it statistically, it's to achieve maximum compound annual growth rate. And we might show and other people talk about how harmful drawdowns are to compounding, they're missing the fact that drawdowns from a classic trend point of view, they're not as harmful because they're overwhelmed by the uplifting. Exactly. And if we imagine a scenario, Jerry, where we had this change in the trend following model where we've started seeing smoother returns being generated, people are volatility targeting, people are dynamic position sizing, what that is doing is compromising the uplifting power which is necessary to overcome this asymmetry. And this is why my thoughts are, why do classic trend followers snap back very quickly from a drawdown? It's because of this beneficial asymmetry. Why do the others take longer to recover from this drawdown? It's because of this natural drag and the compounding equation. Well, I think that naturally leads us to the third topic that being why the cutback rule is essential for survival. And we know that Jerry has some strong views on this topic. Can you once again describe the cutback rule? What is its role in survival? It's laid us through it, Jerry. Okay. So I think the cutback rule is a very important rule. It's the ultimate rule. And that is when you're losing money, maybe define what that means to you. If you're down 10%, 20%, 30%, whatever you're, whatever you're, you know, however you want to set up your fund and have a certain risk profile, then you're going to reduce cutback rule. Okay. When I hit these drawdown levels, I'm going to reduce my current positions and my future trades faster than I'm losing money. So with the turtles, we traded very short term and we traded with very high leverage. So I do think that there's a different answer if you trade long term with low to moderate leverage. But for the example, though, the turtles would, if they were down 10%, just a 10% drawdown, they would cut back the current positions by 20%. So twice as fast as your loss is in the, and then the, the current, the future trades for the foreseeable future until you make this money back would be cut back at 20% as well. So if you're down 20%, you're going to make the 40% cutback. So it's a non-system trade. It's strictly there to help you stay in the game and allow you to continue following the system. I think this is one of the most amazing things about it and this philosophy is, what do I do? Well, I'm really destabilized. I'm losing money. This is not part of the back test. What do I do? Well, the most important thing for you to do is continue to take the same trades in the same way and exact same trades and not skip trades and use discretion because you know you're going to not come out of this drawdown if you're skipping trades and having to choose which trades to do. So the cutback rule says, nope, just reduce your, reduce your trading size at least or maybe even faster or greater than the loss and go right back to following these systems. So I think that is a really, but like I said, it is a non-system trade is undoubtedly going to make, ensure that you're going to make less money. You'd be much better off trading smaller and not ever having or rarely having to use this cutback rule. And I think now with my leverage and my, you know, the leverage that we use and the long-term nature of our systems were seldom. get into a situation where we're going to have to use the cutback rule. So, Jerry, if you rewind back to the '80s when you were trading with Richard Dennis and William Eckhart, how often were you invoking the cutback rule yourself or the other traders that you were trading with? Frequently. We were trading way too large and it was almost like this game. And an outlier trade, it would be much smaller than an outlier trade now. So, like an outlier trade for me might be 50 to 100 ATR. Back to the end, it was 20. That was the magic number. 20 ATRs, that's a really big outlier trade. So, it's a whole different game when you're using massive leverage, making 200% of a year, trading very short term. But if you remember this guy, Jez Liberty, he ran a website and he took some turtle systems and he figured that the turtles just traded a little bit smaller and tried to make a 100% or 150% rather than 200%. And the cutback rule would have been used a lot less and they actually would have made more money. I have no doubt that that's probably true. I have no doubt about it. Would it be correct to say that the trading has quite a quite slowed down in terms of you're using wider parameters? And there's less overall trading now in the program, your program than there was back in the 80s? Yes, there's much longer term and much lower leverage. So, much better diversification. And I really focus a lot on protecting the closed trade equity, the capital and the account, the P&O, the real-life profits and losses plus the investments. And I've, you know, it's seldom happens that the drawdown and the closed trade equity is not exceeded by the open trade equity. So, to some degree, CPAs never have drawdowns. If you look at it, your capital account, the original money that the client gave you and including the real-life trades, that never, the drawdown never is exceeded. The drawdown is never greater than the open trade equities. And you could, you know, for a startup period, you could liquidate the account on any day and recover your entire closed trade equity drawdown with the real-life. With, with the open trade profits. So, to some degree, I think that can change. Probably is going to change one of these days where we're going to see that happen. I don't think it really is fatal, but to some degree, I think if you use closed trade equity as your trade level and you sort of have as a mindset of protecting that and then allowing the yourself the freedom to not care so much about open trades. In the sense that you're going to override the system, you've developed these systems based upon the back test, what makes the most amount of money. And you want to carry those rules forward and carry them out. So, you can do that. And the best way to do that is to not put yourself in a situation where you're even thinking or looking at a drawdown of profitable trades. Rich, what are your thoughts, Rich? I view this as an engineering solution again to ensure that you survive. In other words, it's all about the path of returns. And what I'm saying is that when we do our back testing and our model development, we inevitably have our first line of defense, small bets, stop losses, trailing stops, massive diversification. That's what our models give us. But then we might enter an environment we have never experienced in the past, which can actually start compromising our core equity, what I call the closed realize balance. So in that instance, you need to intercede with the discretionary rule because this is a regime that has never occurred in your back test. We recognize that the future is unbounded. There are so many different possible paths. And what we've experienced in our back test to get us our front line of defense. That's been integrated into the model. But now we've got this environment where we're finding that our closed trade equity, the thing that we've got to protect it all costs to survive, to survive the absorbing barrier to a survive compounding drag. That's being compromised. So we need to intercede with the discretionary rule to massively cut back our exposure to at least survive until we are out of that regime. So it, for instance, I'd have a rule where a cutback rule be applied as soon as my closed balance deck, what he was being threatened. And it would only be turned back on or once I achieved a high watermark. So I'd massively reduce my position sizing faster than what the drawdowns were occurring to just defend myself and protect myself and wait until the regime has passed. And I'm actually back up to my high watermark where I am. And then I'd revert back to my original position sizing. So it's the last line of defense as far as I'm concerned. And it's a necessary discretionary rule to override when a system encounters a regime that's never been experienced in their back test. It's not necessarily discretionary in the sense that you have a rule and you're going to, you know, when you're going to use it and you're going to know when you're going to go back to trading the full speed. It's more like it's just, it's no sample size. And thankfully, like the sample size is zero in the back test. That's fantastic. And we basically are trying to do legitimate robust research come up with these systems that have a lot of us have a large sample size. And we have these great systems that that we're going to maximize a cager. And but they're going to let hunt these outliers and maximize the profit on these outliers. And we give that an eight plus with our systems. And we still say we got to have to cut back will in case we see a future that looks really bad. And so if that's genius, you're kidding. Why do you have to worry about this? You came up with these monster really most profitable systems. They're fantastic. And you're still going to have a cut back rule. Yeah, because we're humble and we're afraid of the future and we're risk managers. We have to stay in the game. And bad things can happen even though we're trading the best systems possible. And you're still going to not rely upon that back test to say, you know, we've never seen it. So we don't have to worry about it. Or if we do see it. Then we know the systems no longer work. That's another thing that people say that's kind of silly. You've never seen something in the back test. And if you see it, if it's bad enough, you can just shut the systems down. As someone told me that a famous trader did that. And a week after doing that. You know, months after shutting his systems down, he figured out that it was making new highs. So yeah, that's I think every single time I've used the cut back rule. It was at the bottom. You know, that's just going to happen. It was at the bottom of the drawdown. And I made the cut back. And but you only have to be wrong one time. So you need to keep preserving that cap. I'm doing whatever you need to do to do that. And suffer worse performance and use a rule that doesn't have sample size in order to stay in the game. Yeah, that's that's what it's all about. That risk of ruin and minimizing the chance that you can't keep trading tomorrow is super, super important. So we're looking at path dependence. We've worked out that we need to take small bets on individual trades. And I guess by extension, we would also say that if we're taking lots of trades, the aggregate of all those trades should not be a large amount such that it threatens the viability of the account. So if we are taking these really, really small bets and lots of them loose pants, I think is the next topic. How do we use loose pants to effectively leverage a system that's only taking really, really small bets? Jerry, could you please explain loose pants to us? Well, I wanted to make sure I knew what loose pants meant. So I asked Jackie PT and it told me and I know no surprise. Why it's why it's stop losses. I don't like the word stop losses. Wide trailing stops, you know, and not being too eager to get out of the trade and optimal stop losses. That's why I look at it. I have an optimal perfect place to get out of the trade not too fast, not too slow. You know, you're just doing less frequent trading and you're willing to tolerate drawdowns. And we do all of these things because the computer told us to like, I would not do it. I only do things that the computer says is something I should do and it's the best thing I can possibly do. So I think sometimes words fail us. I like loose pants, I like classic trend, like trend following, like all these words we use, but it's really very deep in the Python somewhere. And that's all I care about and that's all I do. And anything short of the actual code. load and the rules is inferior. But I do think loose pants gives a great philosophy of, build a look at the chart with your human eyes and say, it looks like the trend is ended or it looks like the trends should continue. These are just all subjective things. Just pay attention to the numbers. And the numbers in the back test says, you can't be as long term as you should be. You're a human. You try to have as loose pants as possible. But it's really difficult to do psychologically. I think people look at the numbers and look at what the computer tells them. And they're like, no, I'm not going to choose that. Or I'm going to throw in sharp. Yeah, that's the real bullshit idea there too, is that I'm going to throw in sharp, you know? Because that'll help me take profits on the way up. And that'll help me smooth things out. And everybody looks at sharp. And so I think that's how people convince themselves that don't maximize these outlier trends, don't maximize K-GRA, don't have loose pants because they want to live in the sharp role. The sharp role helps them sleep at night. Sleeping at night is overrated, making money is a lot more important. Yeah, I think people would probably be surprised at how small the position sizes are relative to the size of our respective funds and then how wide the stops are. Right? So I think when people think about trading, they think about buying 10, 20, 50, 100 lots of some contract and then eking out a few points here or there. And I think that's called scalping, something that we would obviously never advocate as classic trend followers. But we're almost the opposite. We're looking to trade very, very few contracts, which represent a very, very small proportion of the fund. But have these optimal stops as you describe, Jerry, chosen by the backtest, not necessarily the chart. And we love to look at these charts every month because we're living and breeding that PNL. And it's exciting. It's exciting to watch these markets move. But it's not necessarily that helpful when it comes to classical trend following, right? Yeah, I've seen her traders recently even say that they have altered their systems because they're trend followers. And their system was keeping them in a trade when it looks like, but with the human eye, that the trend has sort of ended. And I think that's just not set. It doesn't matter what your eye says. It doesn't matter if it's OK if you don't call what you do trend following. I think our highest loyalty should be to systematic trading. And that's what we should be faithful to is following the rules, following the systems. And for us, it is going to be trend following. And it is a pretty decent way of describing it. But sometimes I've even had people sit down with me, clients, and show me a chart and say, you didn't get out of this trend here on the chart. And everyone knows that that's when the trend ended. And so you have to listen to a lot of nonsense and people who don't know too much about trading have to listen to clients like that. But we are what we are. We do what we do strictly. If you-- like me, if you follow what the computer says, and I was talking to a famous trader one time, and I asked him about his stop loss. And I was like, don't you think the stops that you use are just too tight? And they just create a lot of whoopsaws. And he goes, yeah, that's just my personality. I've never said that in my entire life. Yeah, because you're relying on a backtest, which are facts and figures and statistics. And as you say, optimal, as opposed to a subjective view of a chart, depending on how you feel, or you slept the night before, or whether you jet lagged, or whatever, I think relying on the backtest. Yeah, but I think my point is that it's a tad different from that. And that is, I'm choosing whatever the computer says. The computer says it, then I'm going to do it. And I think that's rare. People say, no, no, no, go back to the drawing board. Don't choose that, because that's too difficult. Now, leading us to the last topic here, topic five, high-voltrend following, the most valuable alternative investment. And there is a seeking alpha article, Jerry, that you have found and shared with us. I've had a quick read. Could you please take us through the key takeaways? Well, I think the key takeaway here is that it really is helpful for portfolios to include trend following, diversified CTA trend following, and don't shy away from CTAs who have used higher leverage because your investment can be less in dollar terms and get more bang for your buck, essentially. And of course, that's true. I think that's a difficult thing to pull off. If you're at one point, he says a 15% allocation to a high-volts strategy targeting 40% to 80% returns. I mean, who are these people who are targeting 40% to 80% because the drawdowns would be pretty hellacious. And 80%, you'd have to figure out a way to use the cutback rule frequently if you're shooting for those kind of returns. But I do think that institutions over the years, they say, have increased encourage CTAs to trade smaller and to use non-classic methods to smooth out the returns. And the article is saying, yeah, but you need CTAs. You need them in style and big doses. And it's more efficient to allocate to the CTAs who are not trying to make 5% or 10% a year, but maybe a lot more than that. And he does sort of take a shot at our favorite CTA groups that reduce the positions when volatility heats up. And I talked to the author today and I said, oh, yeah, remember that time where the CTAs were short to stocks in 2008. And they're making so much money and the ball was picking up. But the exact time their clients needed that short stock exposure, they were buying the shorts back to smooth out the returns. So I think you don't have to look for a far to be very thankful that you don't do things like that. And we're all sticking to the classic way. Yes. I know that some of these funds actually do offer multiple leverage options. So you might be able to go with a low volatility option or a high volatility option. And so sometimes that allows investors to make an allocation based on their preference. But yes, I think the principle that choosing a fund with a high vol allows you to allocate a smaller amount, probably makes sense. My only concern is that I talked about the two-edged sort of compounding. Now high vol and positive skew is a beautiful marriage in heaven. That actually amplifies returns. But no skew and high vol absolutely sacrifices returns. So only in those instances where we have positive skew can we apply this high vol principle? So I'd be saying-- Right. Classic trend is a perfect accompaniment to really move the needle as an allocation into a traditional portfolio. Because it naturally encompasses what I call a beneficial volatility. So the article that I read, which Dury referred me to, obviously criticizes sharp and these traditional metrics. It talks about how they prefer to use sortino. I'd say, well, actually sortino is not a path dependent metric itself. So I actually don't even like sortino. I much prefer the compounded geometry, geometrical paths as a way to describe the best outcome. Because you need to look at-- is the volatility beneficial or is it adverse? And under high volatility, one is crippling and one is magnifies your compounding. So in our case of classic trend following, we're fortunately in the positive skew side with high vol that just amplifies our trends further. So yeah, that's generally my thoughts. But I just in general, I think looking at volatility is mostly going to be incorrect. There's a couple of bullet pointean article one is, all volatility isn't bad volatility. I don't think any volatility is necessarily bad. It's a winning trade. Look, if you're going to get out with a small loss, I mean, I don't care how volatile it is, that never was a problem. You have a small bet. You took your small loss with your small bet. I don't think there is any sort of volatility that's a problem with a winning trade. I mean, look, it's not so much volatility that we hate. We had a big profit. Now it's a small profit. If it happens over a month or two, or it happens over a day or two, I'm really not happy in either scenario. So I don't really see any other bullet point. is volatility isn't risk when it's working for you. I don't think it's ever risk. I think you're looking at downside volatility and upside volatility. You're getting more information. I think I wouldn't get rid of upside volatility as a risk measurement because it's probably going to be a good indication of what the possible downside volatility could be. Your sample size will be increased, but I just don't think you should even care about it too much. Protect that capital, protect that close trade equity, and let those profits run and follow your system. That's really what you need to do. If you want those wonderful returns that you saw in the back test, the computer is asking you one thing. Do you want these? Yes I do. They are awesome. Yes they are. All you have to do is do the same, follow the same rules that I follow historically, follow them in the future. But what about? Shut up about what about? You know, profits can turn into losses. It's going to happen. It's going to have draw downs, primarily with your open trade. So I think CTA is waiting into this and trying to find a narrow path to like volatility sometimes and dislike it. I think just forget that. It's not there. The trend following has some magic to it and you're destroying the magic when you destroy the outliers and don't follow your rules. Well I think that's a good way to close this segment of turtle tidbits. Now over to Rich for Shellmail. Alright, thanks Adam. So we'll open the Shellmail inbox and we've got some excellent questions this month. So firstly a question from Dan who are from X. So Dan says, what do you think about sizing things larger that are better diversifiers or where we have fewer of them like commodities versus smaller sizing for something we have 30 of like indices. So that's his question. So over to you Jerry first then Adam. This is a good question. I think, well I think that the first step for every trader, the first step for every trader, the first step for me anyways is to put together this portfolio and I'm going to create my portfolio. It's going to be a fixed portfolio. I might pick some things out for liquidity reasons in the future and add things in the future because as more markets come on and I can further reduce my bet size. But overall I'm going to put together this portfolio and I trade single stocks and so I could have a portfolio that dominant, you know, had thousands or hundreds of single stocks and would overwhelm all of the other diversifier markets. So I think you need to put a break on, you know, and put your portfolio together with that in mind that you do want some diversification, maybe 25% each sector or something like that. But I would be hesitant to use a different bet size or unit size for commodities for instance because they have more diversification and because we're like I said earlier, we like diversification, we have more of it than most everybody else in the hedge fund world. But we're really trying to hunt outliers and as we said earlier, if things are correlated, it doesn't mean one of the two highly correlated markets might have an outlier that the other ones don't have. So I think following on from what Jerry said, yes, like so we break our contracts, we allocate them to four asset classes being equity indices, bonds or interest rates and commodities and foreign exchange. So I think it would be a reasonable approach to limits each asset class to 25% of the aggregate risk if you wanted to. But I think Jerry, I think you're correct. Well you're definitely correct. The commodities do have unique properties and those commodities, you know, in our opinion, have a lower internal correlation than let's say related indices or interest rates or effects when you're trading a lot of dollar pairs. So an example, a euro to the US dollar or the Swiss franc to the US dollar, that is just going to have a higher correlation than let's say gold and orange juice. So in our opinion, a larger allocation to commodities doesn't make sense. But I could also see the argument, you know, if you're trading two units of energy that are very, very similar, but not identical. So Brent Crude and WTI Crude, you could make an argument to split that risk in half. We don't actually do that. So we do allocate a unit of risk each to each of those contracts. But I think that's just something to work within the back test, put it in the computer and see where it leads. Move on to the next question. This is from Matthew Young from Houston, Texas. And Matthew says, "I'm worried about another lost decade like 2008 to 2018 for trend following. Can you comment on what contributed to that and whether the pitch looks different now?" So that's a great question from Matthew and look, I'll throw it first to you Adam and then to Jerry. So Adam, your thoughts? So the lost decade in trend following, yes, there was a period of relative or underperformance for trend following systems and our analysis leads that to be due to zero interest rate policies, quantitative easing and the suppression of volatility. Now, is that a bad thing? We like to think of markets as accumulating and dispersing energy. So if you suppress volatility for a decade, well guess what, the next decade is probably going to be absolutely fantastic. And it has been. Right? So there is no way to get rid of this risk. You can contain it, you can control it, you can try to suppress it, but eventually the markets will need to find an equilibrium. So I would say that that lost decade is an anomaly. I think if you do look at 50 or 60 year track records and back tests, you will note that period is not necessarily reflective of the greater outlook for trend. And I would say that we've definitely moved into a different type of regime. I think there's a number of fiscal and monetary factors, inflation and other geopolitical sort of events and scenarios that are very constructive for trend following for the future. I mean, was it a lost decade? And it was an underperforming decade. The last time I heard the term lost decade was in the middle of 2008 when stocks had a negative return for 10 years. But I think you know that people have done studies on trend following underperformance. AQR was one and there was some other ones. And it's basically a conclusion is usually the same. And that's just a lack of trends for whatever reason. I don't think it's helpful to focus too much on fundamentals. And I don't think it's helpful to make it to say that's an excuse because maybe we shouldn't be trend following. You have my money. You told me you're going to make money with it. I'm not letting you off the hook just because you didn't have big trends. You chose a system that relies upon big trends. And so they're infrequent. And now they've become really infrequent. So we have to just suck it up and take our medicine. But this is a strategy that is totally dependent upon the huge trends. And when we get them, we feel really good. And we say, you know, that's just the greatest thing ever. No system on the planet can make up for missing cocoa. If you miss cocoa, you don't have a system that can make up for that dumb move of kicking it out of your portfolio. So keep it in your portfolio. Don't kick things out of your portfolio. It doesn't matter how they performed historically. You have it in there because it's diversifying. But we don't want it to be too diversifying. I don't want to pay attention to correlation and too much diversification or split things up 50% split the risk in two because we don't care about smoothness. We don't care about the diversification enough. We care about it a little or a lot. But what we really, really care about is if you're trading a lot of different markets in hitting these outliers. And once again, we will trade two markets that are 99% correlated almost all the time, just in the hopes that one of them might have an outlier of the other one doesn't. So I don't want to cut them into. Brit could have a huge move and WTI may not have a move. And now I've cut Brit into when the whole purpose of existence was the outlier and not having a smooth portfolio. I mean, with 400 markets or 100 markets, commodities, currencies, stocks and bonds, you can't help but have some smoothness in some diversification. Even if you're not trying, and even if you're trying. primary goal is to find out layers even in unlikely places. All right, well, the final question is from Wayne Jarvis and Wayne asks, "I hear often that fat-tailed events happen more frequently than models assume. But how do you quantify that? I've heard you mention the positive outliers occur 5 to 10% of the time, but 5% of what and over what period?" So look, I do have something to say here. So I might start first and then I'll pass it on to Jerry and Adam. What I'm referring to here, I've talked about it earlier in this episode, is the scale invariant nature of fractal systems. So what you find is that if I took a small sample over a fractal system, I would find that in that sample there'd be 5 to 10% of that sample, which dominate the structure of that sample. If I then get a larger sample, I'll still find that relatively speaking, 5 to 10% of those dominant structures. So it's this scale invariance that's important. And just how I measure it. So I don't use a Z score to measure my outliers in my distribution. When we find outliers occurring in our distribution, we find that the mean and the standard deviations are themselves influenced by outliers. And these extreme values, they distort the actual results, making Z scores sort of inappropriate, they're appropriate to use for Gaussian systems. But when you are looking at what fraction of the trades are outliers in your distribution, I'd be saying use a modified Z score and what you do there is you're using what we call the median absolute deviation to determine what is classified as an outlier. And so the way I measure it is anything that is greater than 3.5, which is the absolute value of the mad value, is what I call an outlier in my distribution. And when I look at that distribution number, I find that about 5 to 10% of that distribution can comprise these outliers. No matter what scale I use. So if I have a small portfolio, when I do that 5 to 10%, when I have a large portfolio, when I do that 5 to 10%, we find it's a scale invariant system. What nature of this system occurs in the financial markets. And that's why I'm saying that they're fractible systems. That's how I quantify the number of outliers in my distribution. I think Wayne has asked a good question here. And I think the word that he's missing is not 5 to 10% of the time. It's 5 to 10% of the trade. So I think that's what he's missing there. I've actually heard people say on podcast that 5 to 10 trades, which is totally ignoring the fact that some people trade 10 markets and some people trade hundreds of markets and four or 500 markets. I'm not the only person to trade 300 or 400 markets. I think foreign car trade is 400 maybe 500. And they don't trade half of my markets are stocks, which is kind of cheating. They don't trade individual stocks, I don't think. So it really is just a question of which your portfolio, you want to have this fixed portfolio. So keep it constant in doing the trades, doing all these trades in the same markets and suffering and not having an outlier show up like in cocoa than all of a sudden. Cocoa is still there, still one of your markets and you finally finally get that outlier. So it's think roughly 5 to 10% of the trades. And that could vary over time. It could be zero sometimes. That's another good reason of trading so many markets is the bad luck factor of this 5 to 10% will be smoothed out over time. If you only trade 20 to 30 markets or 50 or 60 markets, you could weigh more than 5 to 10 percent and 0 percent sometimes too. So that's something to try to be avoided. I think you've both covered that question pretty well. The only thing that I would add is that the way I do this is basically creating a histogram of the trades which has the frequency and the profit or loss. And you can just visually see the outliers on that histogram. And you will be able to basically differentiate between quite an quote average trades and then where there's a small loss or a small gain. And then these outlier trades which yes, they're 5 to 10% of those trades in aggregate. And you will see them on that histogram. The goal is to not filter yourself out of those trades. So the more markets, the better, the looser the pants. And you've got to be aware that there's clustering in those returns. So it may well be that a lot of those outliers come at once. And then you might have a period of time where there are no outliers. So small bet sizes and controlling aggregate risk through that small bet size, I think is key. That's it for the questions for the month. Now don't forget, listeners, to send them in by going to our contact page on the Aussie Turtles website. You can either record your question or write your question in the comments section on that field. No, it's time to wrap up this episode. So Jerry, are there any closing thoughts on this episode for you? Oh no, I'm not prepared for more closing thoughts. I gave it all I had during the regular time there. So it was good to be back with you guys and it's always fun. And I'm still looking for my missing wine. Well, yes, we enjoy the Adam and I did. We had a fun time right here. I think next year it's going to be recording from Vegas for something. What do you think? Oh, that's good. Yeah, I like that. Fantastic. All right, you're wrong. All right, Adam, any final thoughts for this episode? Oh, look, I would just say back test, back test, back test, it's fun looking at charts and watching the silver market close above $45. But I think the value is in the back tests and the stats and the trade levels that that delivers. Yeah, so back to you. That's great. Many thanks, guys. Well, that's it for episode seven of Turtle Talk. Thanks for joining us, everyone. We'll be back with more charts, conversation and trend talk in the next one. And before we go, I just want to mention that our first book, the Aussie Turtles Trend Following Guide, will be launched, I think next week on Amazon, a Kindle and Paper that back version to start. We'll get a hard copy a bit later happening, but we're looking forward to that being released and Jerry was kind enough to give us a great forward in that book. So many thanks, Jerry, for that. Without more ado, it's time to say goodbye and we'll see you next month. So until then, stay systematic, stay patient and may the trend be with you. Thanks for tuning into Turtle Talk. If you enjoyed the episode, don't forget to rate and review the podcast. It really helps us reach more listeners. Share it with your friends and invite them to join the conversation. Got a question for us? Send it to info@ AussieTurtles.com and we'll do our best to feature it on a future episode. Please note, Turtle Talk is for informational and educational purposes only. The views expressed by the hosts and guests are their own. And do not necessarily reflect the opinions of the podcast producers or affiliates. This podcast does not provide financial, investment or professional advice. Always consult with a qualified professional before making investment decisions. That's it for today folks. See you next time and as always, may the trend be with you. [BLANK_AUDIO]