The Programme, From the Inside
Introducing the System Anatomy Series: seven essays on the structural logic of how a systematic trend following programme is actually built and run, written for the practitioner who wants to understand the mechanics, not just the philosophy.
The Foundations Series answered one question: why is the Outlier Hunting programme built this way?
It started from first principles. Fat-tailed markets. Complex adaptive systems. The geometry of loss and recovery. The impossibility of prediction. The structural logic of why process beats judgment, why diversification is opportunity amplification rather than simply risk reduction, and why the Cut Back Rule is an engineering requirement rather than a precaution.
Ten foundations. One coherent argument.
This series answers the question that naturally follows.
How does the thing actually work?
Not in theory. From the inside.
The System Anatomy Series is seven essays examining the structural logic of the programme, component by component.
I am not going to give you the specific parameters. There will be no entry thresholds, proprietary lookback periods or ATR multipliers to copy into a spreadsheet and call a trading system.
That is not the point.
What matters is understanding why each component exists. What problem is it solving? Why is it built the way it is? What does it depend upon elsewhere in the programme? And what starts to break when it is removed or implemented incorrectly?
This is the anatomy of the machine.
Not its settings.
What the Foundations Series Opened
The Foundations Series ended with an important claim.
Those ten foundations were never ten separate ideas. They were one argument expressed in ten different ways.
Remove one and the others become less coherent.
Position sizing without a complete system is just a formula looking for a framework. Diversification without appropriate position sizing can simply give you broad exposure to the same underlying vulnerability. An exit mechanism designed to capture outliers is of little use if the trader cannot survive the losses required to reach them.
The pieces depend upon one another.
That was true at the philosophical level.
It becomes much more obvious when we open the machine.
Each of the seven components in this series exists because it solves a particular problem created by trying to capture fat-tail events inside a noisy, non-stationary and adaptive market environment.
Entry logic deals with the problem of acting when certainty is unavailable.
Exit logic deals with a different uncertainty. Once we are in, we cannot know how far the opportunity will travel.
Position sizing determines how much exposure we take while allowing the programme to survive the many unsuccessful trades that are the price of finding the few extraordinary ones. It also determines which capital we allow the programme to compound from, and how exposure responds when the realised path deteriorates.
Portfolio construction determines how widely we search when we cannot know where the next outlier will appear.
The ensemble reduces our dependence on any single structural expression of trend. No single way of seeing the market is asked to carry the programme.
Rollover mechanics goes beneath all of them. Futures price histories are constructed from contracts that expire. If the measuring instrument is inconsistent, everything built on top of it inherits the error.
And finally, execution and process deal with the point where the architecture meets the world. A signal is not a trade. A model is not a position. The integrity of the programme ultimately depends on whether what was designed and tested is what actually gets executed.
None of these mechanisms lives alone.
Position sizing depends on a measure of market movement, which in turn depends upon the integrity of the price history used to calculate it. Portfolio construction determines where we search; the ensemble gives us different ways of encountering what happens there. Entry and exit logic depend on the same constructed price series that rollover mechanics is responsible for preserving. And all of it ultimately depends upon execution faithfully translating the architecture into actual positions.
Pull one piece out and the load does not disappear.
It moves somewhere else.
The intelligence is not sitting inside any one component.
The intelligence is in the architecture.
“The Foundations built the argument. The System Anatomy is the argument, expressed in metal.”
Why Mechanics Matter
There is a trap in systematic trading education.
We can become very good at explaining the philosophy while remaining surprisingly poor at implementing it.
A trader can understand perfectly well why discretion is dangerous, then tighten a stop because this particular retracement feels different.
They can understand the importance of diversification, then build a fifty-market portfolio that turns out to contain the same underlying exposure wearing fifty different hats.
They can backtest using one method of constructing continuous futures contracts and trade live using another, then wonder why the behaviour of the live programme begins to drift away from the historical model.
The philosophy can be right.
The mechanics can still be wrong.
And markets do not pay us for having the right philosophy.
They expose the implementation.
That gap between understanding something and actually building it correctly is where a great deal of systematic trading fails.
The System Anatomy Series is about that gap.
Each episode takes one component of the programme, opens it up and asks what it is actually doing. The objective is not to hand you a parameter. It is to make the structural logic clear enough that the implementation begins to follow naturally from the problem being solved.
That distinction matters.
Parameters age.
Structural problems remain.
The Shape of the Series
There are seven episodes.
The sequence broadly follows the life of a position through the programme: finding it, holding it, sizing it, placing it within a portfolio, combining different ways of responding to the same uncertain market, preserving the integrity of the instrument being measured, and finally translating the architecture faithfully into live execution.
Episode 1: Entry Logic
Every trade begins with a signal.
Unfortunately, the signal does not arrive carrying a little sign saying this is the big one.
Entry is threshold detection inside a noisy system. We act without certainty because certainty is unavailable.
That is not a defect in trend following.
It is the price of admission.
We will examine why simple entry rules tend to survive better than elaborate ones, not because they are necessarily superior in any particular regime, but because they present fewer surfaces of failure across many different regimes.
The signal does not know whether it is right.
That is not its function.
Episode 2: Exit Logic
If entry gets us into the game, exit logic determines the shape of the game we are playing.
This is where the asymmetry we seek is allowed to emerge.
An initial stop defines the loss boundary we are prepared to accept under normal execution. A trailing mechanism does something psychologically much harder: it gives an expanding winner room to remain alive.
The first part is easy to understand.
The second is where many traders discover that understanding an asymmetric payoff and actually living through one are very different things.
We do not know how far a trend will travel.
So we do not require the exit to know either.
Episode 3: Position Sizing
Different markets move differently.
ATR normalisation allows us to scale positions so that markets with very different recent movement can be placed on a more comparable footing when determining exposure.
But there is an important boundary here.
Normalising volatility is not the same thing as measuring risk.
ATR tells us something about the recent scale of movement. It does not tell us what the market is capable of doing next. It cannot see a discontinuity waiting behind tomorrow’s open, nor can it tell us that a period of apparent calm is about to end violently.
That distinction matters.
The sizing formula has a job. We should not pretend it does jobs it was never designed to perform.
And there is another question hiding inside the calculation.
What equity are we sizing from?
Open profits are not the same thing as realised capital.
If unrealised gains are allowed to inflate the sizing base, successful open trends can quietly increase exposure elsewhere in the portfolio before those gains have been banked.
So position sizing becomes part of something larger.
It determines how the programme compounds, how realised growth becomes future capacity, and how exposure responds when the realised path deteriorates.
The formula does not find the optimal position.
It finds the survivable one.
Episode 4: Portfolio Construction
If we do not know where the next great trend will appear, we need to be there before it arrives.
That is why diversification, for the Outlier Hunter, is fundamentally about opportunity amplification.
We will examine different dimensions of diversification and how meaningful structural variety is translated into a broad search across markets.
But breadth does not mean pretending markets are independent.
They aren’t.
Markets exist inside a connected global system, and those connections themselves change as conditions change.
There is also an important distinction that is frequently missed: price correlation and the return paths produced by trading rules are not the same thing.
Markets that appear similar at the price level need not produce identical strategy-return paths once a trading rule interacts with them.
And markets that appear different can reveal common dependencies when conditions change.
The portfolio does not prevent the storm.
It determines whether the structure survives it.
Episode 5: The Ensemble Approach
One system is one particular way of interacting with an unknowable market.
An ensemble gives us several.
The objective is not to collect signals for the sake of having more signals.
It is to diversify structural sensitivity.
Different rules can respond differently to the same evolving market path.
One may act while another waits.
One may remain while another leaves.
Neither has to be wrong.
That disagreement can be the reason the ensemble is useful.
We are not looking for a committee of systems that votes on what the market will do next. We are allowing different mechanisms to respond independently to the same uncertain path.
This also changes the way we think about robustness.
A rule that looks wonderful in the environment where it was created tells us less than we might imagine.
The harder question is whether the logic travels.
Does it remain recognisably useful when it leaves home?
No single instrument sees the whole sky.
That is why we build the array.
Episode 6: Rollover Mechanics
This is the episode about the measuring instrument.
Which is precisely why it matters.
Futures expire. Positions have to move from one contract to another. Continuous price histories have to be constructed. Term structures change. Activity migrates between contracts.
The long historical price series sitting on a screen is not one instrument that traded continuously for forty years.
It has been assembled.
That makes its construction part of the architecture.
Small inconsistencies here can quietly contaminate everything sitting above them.
If the historical programme was tested on one representation of the market but the live programme is effectively trading another, the backtest no longer describes exactly what you are doing.
That is not a footnote.
It is an integrity problem.
The survey can look precise.
Whether it is accurate depends entirely on the calibration of the chain.
Episode 7: Execution and Process
Finally, the machine has to touch the world.
A signal is not a trade.
A theoretical position is not an actual position.
A backtest does not pay anyone.
At some point an instruction has to become an order, an order has to become a fill, and the live portfolio has to remain faithful to the architecture that produced it.
This sounds operational.
It is more important than that.
Because execution is also the point at which the human being returns to the machine.
We can build simple rules, control exposure, diversify broadly, combine different structural sensitivities and preserve the integrity of the data.
Then undo all of it with one sentence:
This time feels different.
The final episode deals with the discipline required to ensure that the programme we trade remains the programme we designed.
No selective signals.
No discretionary improvements made in the heat of the moment.
No quiet rewriting of the rules because the market has become uncomfortable.
The architecture only exists in the real world if we actually execute it.
A Note on Complexity Science
Throughout this series I will use the language of complexity science: complex adaptive systems, fat tails, power laws, non-stationarity, regime shifts and structural dependence.
These terms are not decoration.
They describe the environment in which the programme has to survive.
That matters because system design cannot be separated from the nature of the system being navigated.
If markets were stationary, normally distributed and reliably predictable, I would build a very different machine.
They are not.
So I don’t.
Readers who have worked through The Foundations Series will already know this territory. If you are arriving here for the first time, the Glossary will help, but I would strongly recommend beginning with the Foundations.
This series does not replace that argument.
It opens the bonnet.
Where This Series Sits
I think of the progression this way.
The Foundations Series is the on-ramp.
It establishes the intellectual framework and explains why the programme has the shape it does.
The System Anatomy Series is the engine room.
It takes the machine apart and examines what each component actually does.
A future course planned for the second half of 2026 becomes the workshop, where the emphasis shifts again from structural understanding to detailed implementation.
Why.
How.
Then build.
They overlap because they should.
Understanding becomes more useful as we move between levels. A mechanical rule means more when we understand the structural problem it solves. An implementation decision becomes easier when we understand why the mechanism exists in the first place.
But there is an important boundary.
This series is about principles, not prescriptions.
There are many legitimate ways to construct a systematic trend-following programme. Different rules. Different definitions. Different implementations.
I am not interested here in arguing that one particular configuration is the answer.
That would contradict much of what this series is about.
The mechanics can vary.
The principles travel.
And that brings us to where every trend begins.
Not with certainty.
Not with a forecast.
Not with somebody knowing what happens next.
With a threshold being crossed.
Episode 1 begins at that precise moment.
The signal fires.
The position opens.
And at the instant we enter, the trade that will eventually become the extraordinary outlier looks structurally indistinguishable from the trade that will stop out a few days later.
We do not know which one we have.
That is not a weakness in the system.
It is the reason the rest of the system has to exist.
READ DEEPER
→ The Paradox of Simplicity: Why the Best Trading Rules Are Counterintuitive
→ Selection, Not Skill: Why Simple Strategies Outlive Brilliant Ones
→ Built for Purpose: Why Every Trend Follower Has a Different Mission
→ Adaptation Replaces Prediction. Reaction Replaces Belief.
→ The Cut Back Rule: Engineering Survival in a Fractal World
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, Complex Adaptive Markets, Carved by Impossibility and The Aussie Turtles Trend Following Guide.
Want to explore why structure exists at all?
Carved by Impossibility: What Remains When Everything Else Is Eliminated
The book explores the architecture of constraint, emergence, and reality itself, and what it means for how we understand markets, life, and the universe.
Available now on Amazon in paperback, hardcover, and Kindle.
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.