“Time and again, our work is judged by instruments built for a different game and against objectives we never chose. This is a long look at what actually separates the classic trend follower from the rest of the investment world, and why naming that difference is a matter of precision rather than pride.”
There is a sentence I keep returning to in conversations with the people I trade alongside. We are not them, and they are not us. It began as a half-joke, the sort of thing you say after the tenth time someone has measured your life’s work with a ruler designed for something else. The longer I sit with it, the less it sounds like a joke and the more it sounds like the most accurate thing I can say about what we do.
Here is the experience that produces it. A classic trend follower puts up a record, and the conventional world reaches for its standard instruments to assess it. The Sharpe ratio. Correlation to equities. Realised volatility. The information ratio. Drawdown smoothness. The numbers come back looking modest, or lumpy, or hard to place, and the verdict follows quickly. Decent diversifier, nothing special, probably better in a small allocation. The reflex at that point is either to assume we have done something wrong, or to defend our figures on the same terms and argue that the Sharpe is actually fine if you squint. Both responses concede the game. Both accept that the instrument was the right one and only the reading is in dispute.
It took me a long time to see that the instrument is the dispute. The readings are not wrong. They are answers to a question we are not asking, about a quantity that does not carry our edge. When you measure a positively skewed, convex, survival-driven process with tools built for symmetric, smooth, mean-reverting returns, you do not get a flattering picture or an unflattering one. You get a category error wearing the costume of a measurement.
The argument is almost always about the instrument
Once you start looking for it, the pattern is everywhere. Nearly every long-running argument about trend following turns out, on inspection, to be an argument about the measuring instrument rather than the strategy itself. Change the instrument and the disagreement quietly dissolves, with both sides turning out to have been right about different things. So before the examples, it is worth stating the thread that runs through all of them.
The conventional world and the Outlier Hunter are not measuring the same thing in two different ways. We are measuring in different spaces. They work in price space, in the world of daily co-movements and second moments and bell curves, where risk is wobble and the average is the thing that matters. We work in a different space, the space of realised profit and loss compounded along a single path through time, where the shape of the tail is the thing that matters and the average is a fiction nobody actually lives. Almost everything that follows is a consequence of that one separation.
They measure the price. We live on the equity curve. Those are not the same object, and the gap between them is where our entire edge hides.
This is not a slogan. It is a testable claim, and we have tested it. What follows are four places where the conventional instrument and the lived reality come apart, and where the appearance of a weakness turns out to be an artefact of measuring in the wrong space.
One. Correlation, and the redundancy that is not there
Three respected research groups, two analyses from Graham Capital and one from Quantica, have argued in recent years that once a trend portfolio holds fifty or sixty markets, adding more does little, because the new markets are correlated with the ones already there. On their own terms they are correct. Add a market that moves with the others and you barely move a price-based, volatility-managed information ratio. The redundancy they describe is real, in price space.
But a trend follower does not hold prices. We hold positions, and we earn equity curves, and the correlation that bears on our diversification is the one between those profit-and-loss streams, not between the underlying prices. When you measure it properly, every pair of markets is less correlated on its equity curves than on its prices, and the gap is widest exactly where the price correlation is highest. Two markets that look like near-twins on a chart produce noticeably different streams of profit, because what a trend system captures depends on when it enters, when it exits, and how long it holds, and those rarely line up even when the prices do.
The proof of the pudding is in the tails.
Across sixty-eight markets and forty years, in the ten worst months the equity market has suffered since the mid-1980s, a broad trend book was positive in every single one. That is the behaviour the redundancy story cannot explain, and it falls straight out of the thing the conventional instrument treats as a flaw, namely holding many markets that can move together. When they break together, a book with no fixed view is short whatever is falling. Measured in price space, breadth looks like duplication. Measured in the space we actually occupy, it is the engine.
Two. Sharpe, the average, and the single life you actually live
The Sharpe ratio, and the whole family of mean-variance measures around it, rest on a quiet assumption that almost nobody examines. They assume that the average outcome across many parallel versions of you is a good guide to the outcome you personally will experience over time. In a world of symmetric, well-behaved returns that assumption is harmless. In our world it is false, and the technical name for why it is false is non-ergodicity.
You do not get to live the average. You live one path, in sequence, and you compound along it. On a single compounding path the order of returns matters, a large loss can remove you from the game before any future gain can arrive, and the geometric growth rate, not the arithmetic average, is what determines where you end up. A strategy can have a forgettable Sharpe and a superb terminal-wealth profile, because the Sharpe is averaging away the very asymmetry that does the compounding. Many small, bounded losses and a few enormous gains is a shape that a variance-based measure actively penalises, scoring the large gains as just more volatility to be docked for.
Expectancy and the Sharpe ratio describe a casino full of parallel gamblers. We are one person walking through time, and the only question that matters is whether we are still walking when the outlier arrives.
So when someone reports that trend following carries a modest Sharpe, I do not argue the number. The number is roughly right and entirely beside the point. We are not optimising the ensemble average of a symmetric distribution. We are protecting and exploiting a single, skewed, compounding path. Judged as what it is, a low Sharpe on a convex, positively skewed strategy is not a weakness to be apologised for. It is what a healthy outlier-hunting equity curve looks like from inside the wrong instrument.
Three. Risk is not the wobble
Ask the conventional world to define risk and you will be handed volatility. Standard deviation of returns. The width of the wobble. It is a definition of enormous convenience, because it is easy to compute and it makes the mathematics tractable, and it is wrong for us in a specific and important way.
For an Outlier Hunter, risk is not how much the equity curve shakes. Risk is ruin, the permanent loss of the ability to compound. It is the sequence of events that takes you below a level from which you cannot recover, or that strips your capital so far down that the next great trend, when it comes, lands on too small a base to matter. Volatility and that kind of risk are not the same animal, and at times they point in opposite directions. Upside volatility, the violent favourable move, is not a danger to be suppressed. It is the entire prize. A process designed to minimise the wobble will, as a matter of mechanism, trim the upside surprises along with the downside ones, and in doing so it cuts away the fat right tail we exist to capture.
This is why the oldest rule in the book, cut losses short, is not a platitude. It is the actual risk control, aimed at the actual risk. We bound the downside per trade so that no single position can compound into ruin, and then we let the winners run without a volatility cap strangling them. We treat the deepest hole, not the average shake, as the thing to fear, and we treat the rare violent gain as the thing to protect. A manager who measures risk as volatility and a manager who measures it as ruin will build two different machines, and only one of them is built to survive long enough to be paid by the tails.
Four. We do not forecast, and that is the design, not the gap
The conventional critique that stings the least, because it is true, is that we cannot tell you in advance which market will produce the next great trend. We cannot. Nobody can. Our own reading of markets as complex adaptive systems, where most large moves are driven from inside the system by the behaviour of its participants rather than by the arrival of outside news, leads us to expect exactly that unpredictability. The location and timing of the next outlier is not a solvable problem.
The conventional world treats this as a deficiency to be fixed with better prediction. We treat it as a fact to be designed around. If you cannot know where the outlier will appear, you do not place a clever bet on where it will appear. You make sure you are present everywhere it might, you keep the cost of being in the wrong places small, and you let the distribution itself select the winners by allowing your rules to stay in whatever begins to run. Breadth is not an admission that we cannot forecast. It is the structural answer to the fact that nobody can. The critique and the design are the same observation, read by two parties with opposite reflexes. They want to predict the path. We want to be standing on every path, lightly, when one of them ignites.
Even among trend followers, we are a distinct species
It would be too easy to draw the line between trend followers and everyone else. The more honest line runs through trend following itself. Among systematic trend followers there are at least four recognisable faces, and they are not variations on one another. They are answers to different questions.
There are the Replicators, who aim to deliver the broad trend-following return stream cheaply, liquidly and at scale, usually with tight volatility targeting and a focus on a smaller set of deep markets. There are the Core Diversifiers, who treat trend as a satellite holding inside a larger portfolio and tune it to play nicely with everything around it. There are the Crisis Risk Offsets, who shape the strategy primarily as portfolio insurance, sized to pay out when equities break. Each is a legitimate, well-built thing solving a real problem for a real client, and the people running them are serious.
The Outlier Hunter is the fourth face, and it is built around a different objective from the other three. For us, maximum breadth is not a preference, it is the oxygen the process breathes. We are not trying to deliver the smoothest possible version of the trend stream, nor to fit neatly beside a sixty-forty portfolio, nor primarily to insure someone else’s equities. We are trying to be holding the rare, structural moves when they arrive, in whatever obscure market they choose, and to have survived the wait. The same words, trend following, cover all four of us, and underneath the word we are doing genuinely different jobs.
What we actually are
It is easy to define ourselves by what we decline to be. The harder and more important thing is to say plainly what we are, because the affirmative version is the one worth standing behind.
We are Outlier Hunters. We hold that markets are complex adaptive systems, not efficient machines, and that their largest moves are generated from within by the collective behaviour of participants. We hold that returns are not normally distributed, that a small number of rare, high-magnitude events dominate long-run results, and that the central task is to be positioned for those events rather than to predict them. We cast the widest net we can, across many markets and many systems, because we cannot know in advance where the next outlier will form. We cut losses short, so that being wrong is cheap and survivable. We let profits run, so that being right is allowed to become enormous. We measure risk as ruin and the loss of compounding, not as volatility. We harness convexity and positive volatility rather than suppressing them. We protect the path of our equity above all, because the path is the only thing we actually live. And we accept many small, dull losses as the honest price of being present when the rare and decisive thing happens.
None of this is new in its parts. Cut losses, let profits run, diversify widely, respect the tails, these run back through the Turtles and the whole classic tradition. What is ours is the insistence on taking them all the way to their logical end, refusing to dilute the design to flatter a Sharpe ratio or to smooth a correlation, and the framing of the whole posture around the hunt for outliers rather than the following of trends. The trend is only the visible signature of an outlier forming.
Why the distinction is not vanity
If this were only about identity, it would not be worth a long essay. It matters because being measured by the wrong instrument has real and expensive consequences, and not only for our feelings.
An allocator who judges us purely on Sharpe and correlation will systematically under-allocate to the strategy that would have helped them most in the moments that actually threatened their portfolio, because the instrument is blind to skew and to crisis behaviour. A manager who internalises the conventional yardstick will be tempted, year after year, to volatility-target a little harder, to trim the lumpy winners, to prune the obscure markets that drag on the optics, and in doing so will slowly file the convexity off the strategy until it becomes a smoother, safer, lower thing. That is the quiet tragedy in our corner of the industry. The pressure to look respectable on instruments built for someone else gradually turns Outlier Hunters into Replicators, and the edge dies not in a crash but in a thousand small concessions to the wrong measure.
So naming the difference is a way of protecting it. When I say we are not them, I am not claiming we are better at their game. We would lose at their game, and they would lose at ours, and both of those facts are exactly as they should be. I am saying that we are playing a different game, with a different objective, that demands a different instrument to be seen at all. The respect I have for the Replicators and the Crisis Risk Offsets and the genuinely skilled people running passive and predictive strategies is real. It is precisely because I respect what they are built to do that I refuse to be judged as a worse version of it.
We hold the wide net. We bound the downside in the rules. We accept the long, quiet stretches as the price of admission. And when the outlier finally runs, we are already there. That is the whole job. The outlier is the quarry.
Richard Brennan writes on systematic trading, complex adaptive markets, and the philosophical foundations of trend following at atstradingsolutions.com. His books include The Fractals of Finance and Complex Adaptive Markets. The forthcoming Carved by Impossibility completes the trilogy.
Want the theoretical foundation for why trend following works?
The Fractals of Finance: Determinism, Adaptation and the Geometry of Markets
The book explores the full architecture of feedback, fat tails, and fractal structure in financial markets, and what it means for how we trade, invest, and understand risk.
Available now on Amazon in paperback, hardcover, and Kindle.
Want a practical field manual for trading trends and capturing outliers?
The Aussie Turtles Trend Following Guide: A Field Manual for Hunting Outliers adapts the timeless principles of the original Turtle traders into a systematic, rules-based approach for modern markets. Co-authored with Adam Havryliv.
Available now on Amazon in paperback, hardcover, and Kindle.