What Survives Is What Could Not Be Eliminated
Persistence is the visible residue of variation, constraint and history
Previous: Episode 2, The Future Is Not Hidden, It Is Unwritten
Next: Episode 4, The Same Pattern at Every Scale
The river at the bottom of a canyon looks small from the rim. The walls look permanent. But water cut through the landscape over a long history, carrying sediment and exploiting differences in rock, while the terrain directed the flow. As the channel deepened, it altered the route later water would take. The river cut the canyon, and the canyon is a record of everything the rock would not allow.
From above, the route can look chosen. We see the channel that exists and none of the passages that closed, silted up or failed to cut through the rock. A successful trend strategy creates a similar visual trap when its surviving winners appear alone on a presentation slide. Their apparent inevitability is partly the disappearance of the paths that did not last.
The channel and the routes that vanished
A canyon is no passive mould. Rock erodes differently according to its composition and fractures. Water arrives with changing force and sediment load. Uplift, slope, tributaries and weather also matter. The present channel is the result of their long interaction. [1] One flood can rearrange a reach that seemed fixed; a resistant layer may preserve a bend while a softer one gives way. The shape we see is a historical result, not a plan the landscape kept hidden for the river to discover.
Think of the routes the water cannot continue to take. One side of a channel resists erosion and sends more flow through another. A bend grows; a passage silts up; a flood opens an alternative. What we call the river’s course is the path that continued to be available as both water and land changed. The terrain constrains water, but water also changes the terrain. That reciprocal shaping is why the residue tells a story.
The explicate canyon is visible. We can map its walls and measure the water. In the language introduced in Episode 1, the implicate is the constraint structure under which these arrangements came to be and continue to change. The photograph cannot show that whole structure. It shows a result through which we can investigate it. The canyon preserves some traces and erases others.
There is a practical warning in this view. When a structure has lasted, we are tempted to tell the story backward and make its survival inevitable. We see this channel and imagine the water was always bound to use it. But inherited conditions and contingency matter. There may have been other workable routes, and a different sequence of floods could have produced a different canyon. Persistence records what occurred under real constraints, not every path that was possible.
Evolution keeps its history
Richard Lenski’s long-term E. coli experiment offers a clearer case of adaptive filtration. Twelve populations descended from a common ancestor were propagated under controlled conditions. After tens of thousands of generations, one lineage gained an ability to use citrate aerobically in that setting. Researchers could revive preserved samples from different stages and examine the role of earlier changes that made the later innovation more accessible. [2]
The story has variation, historical contingency and selection. Novel changes were generated; earlier changes altered the possibilities available to later generations; the environment filtered which variants could increase. Eleven other populations followed different histories. They were not simply “eliminated” because they did not acquire this particular trait. Their continued existence is part of the result we must explain.
That is why residue is a stronger word than “winner”. The residue is the visible arrangement remaining after filtration, the continuing loss of configurations that cannot persist under current constraints, acting on variation and inherited history. Filtration is no conscious judge. It does not select for my favourite outcome. It may leave multiple workable arrangements and remove one that would have flourished in a different environment.
The organism is not the river, and neither is a market. But both examples help us see an operation we often overlook. Conditions constrain what continues. What continues inherits a history. Its presence changes the starting conditions for the next round. We are residue all the way down in that sense: made of arrangements that have persisted through earlier constraints, without supposing that every persistent physical thing has adapted or that survival alone confers superiority.
The preserved samples make the argument unusually tangible. Researchers could restart a lineage from an earlier generation and ask whether the later innovation remained as accessible. They were investigating the effect of a population’s own history on what could emerge next. That is a stronger account than announcing that the final population was “fitter” in the abstract. The ability concerned a particular resource under specified conditions. It arose through changes that had accumulated in one line of descent. The outcome makes sense once that chain is examined, even though nobody could have named its exact endpoint at the beginning of the experiment.
A trend that keeps going
A market move is continuously met by other decisions. Some participants take profits. Others bet on reversal, hedge an exposure, reduce risk or position for what they believe is exhaustion. The resulting orders meet participants willing to trade the other way. Many apparent trends stall or reverse. Some continue far enough for a trailing exit to retain an unusually large gain.
It would be false to claim that all traders are trying to kill a trend. They have different motives, and the same trade can serve different purposes on each side. The defensible point is stronger than a staged contest: a visible trend has persisted through the real orders and constraints it encountered. We observe the surviving path, not a list of all the moves that might have grown but did not. That is filtration without intent.
The exit in my trend system responds to this difference without needing to know why a move continues. When a breakout fails, the rule closes the position. When it extends, the rule leaves room. I cannot point to a particular current winner and declare it immune to reversal. I can construct a method in which a continuation, if it occurs, can matter to the account. The many abandoned beginnings and the few long paths belong to the same research record.
The continuing trend has met selling, hedging and changes in demand without ending yet. Its precise route was contingent, but its visible persistence means it remained possible under the pressures that actually arrived. The alternatives that ended early are absent from the chart of the winner. What survives is not what was cleverest. It is what could not be removed by those pressures.
The claim has a real limit. We see survivors more readily than failures, so survivorship bias can make a method or a manager appear stronger than the full field from which it emerged. Survival under earlier conditions is not a promise of survival under new ones. I must examine the missing cases, the rules and costs behind a long record, and the constraints that the record has yet to encounter.
The river did not have to trace this exact line. Yet a channel exists where water continued to pass through the ground it could erode, and its earlier passage changed the ground for later flow. The lineage in the laboratory likewise inherited earlier changes that made one new ability accessible. Persistence looks like design when the failed and foreclosed alternatives have vanished from sight.
Configure against elimination
I shift from designing around an imagined winner to keeping the method capable of another attempt. Two questions follow. Is there a credible advantage worth pursuing? Can I remain able to pursue it when its ordinary failures and extraordinary adverse events arrive? The first demands evidence across markets and time, with implementation costs included. The second concerns the size of positions, the capital and mandate available, and the ability to act after losses.
Neither question can replace the other. A beautifully funded strategy with no advantage can survive financially while steadily losing the reason to exist. A promising rule can be rendered useless by an exposure that forces me out before its rare outliers arrive. To configure against elimination is to make continued participation possible while remaining honest about whether the participation deserves capital.
Research on historical time-series momentum gives a reason to take trend following seriously across many markets and long periods. [3] But a broad family of historical results is a beginning for examining a particular rule, not an inheritance that any implementation receives automatically. I need to know the market universe, costs, revisions, capacity and which trades carried the record. I also need to look for the approach that ceased reporting, the manager who cut size, and the signal that was skipped after a drawdown. The surviving chart conceals these absences.
In my own work, the practical shift is to put endurance into design before I need it. A small initial commitment permits another attempt after a loss. A trailing exit leaves a live opportunity open without pretending I can calculate its final extent. A documented review can retire a rule whose evidence has genuinely weakened. The aim is not indefinite survival at any price. It is to remain present for an opportunity while preserving the ability to recognise when the original case no longer holds.
In trading, I want the capital and rules to keep me eligible for the next attempt without granting a failed method permanent support. Configure for continued participation, then test whether that participation is worthwhile. The next trend will not announce itself. My chance to hold it depends on having survived the attempts before it.
The river continues through the canyon it helped cut. Episode 4 asks how far we can carry this way of looking across different scales without pretending the mechanisms are the same.
Previous: Episode 2, The Future Is Not Hidden, It Is Unwritten
Next: Episode 4, The Same Pattern at Every Scale
Sources and notes
[1] U.S. Geological Survey, Geology of Grand Canyon National Park, on the interaction of water, rock and uplift.
[2] Zachary D. Blount, Christina Z. Borland and Richard E. Lenski, Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli, Proceedings of the National Academy of Sciences 105, no. 23 (2008): 7899–7906.
[3] Brian Hurst, Yao Hua Ooi and Lasse Heje Pedersen, A Century of Evidence on Trend-Following Investing, Journal of Portfolio Management 44, no. 1 (2017): 15–29.
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.