“Trends are patterns. Outliers are structure.”
Introduction – Seeing Beyond the Surface
Markets are full of patterns. Prices rise, fall, and repeat. We draw lines, name formations, and build systems to capture what looks familiar. But what we are seeing are patterns, not structure. Patterns describe form, what appears. Structure defines cause, what drives those appearances. When we recognise that structure itself is fractal, repeating across scales, we realise that outliers are not accidents. They are the visible consequences of deep asymmetry, moments when feedback loops align, equilibrium breaks, and the market reveals its hidden architecture.
Patterns: Fleeting Order on the Surface
A pattern is an observable recurrence in data such as a trend, a cycle, a correlation, or a regime. Patterns are descriptive. They tell us how price has behaved, not why. They attract traders because they are visible, measurable, and actionable. Yet once recognised, they are quickly exploited and decay. The market adapts faster than any fixed pattern can endure.
Patterns are quickly exploited. Structure persists.
Patterns are the waves on the surface of the ocean, shaped by invisible currents beneath. They are what the market looks like, not what the market is.
Structure: The Architecture Beneath
Structure is the deep geometry that governs how markets behave. It is the configuration of relationships, incentives, feedbacks, and constraints that determines how energy flows through the system. Structure is causal, not descriptive. It persists through time, connecting events that appear unrelated on the surface. This is why serial correlation, the persistence of returns, is not a pattern but a signature of structure. It shows that today’s outcomes are not independent of yesterday’s and that memory is embedded in the system through feedback.
Patterns are what we observe. Structure is what connects those observations through time.
Structure defines the possibility space. When structure changes, patterns rearrange, but the underlying architecture endures.
Regimes and Cycles: Patterns of Structure in Motion
Regimes and cycles feel structural because they last longer, but they remain patterns, surface expressions of slow structural reorganisation. A low-volatility regime, a credit cycle, or a risk-on environment are all manifestations of deeper processes. When the structure beneath them shifts, those visible states dissolve. Structure does not oscillate, it reconfigures. Regimes and cycles are simply its temporary faces.
Trends and Directional Bias
Trends are among the most alluring of all patterns, yet not all trends are alike. Some are mere noise, some are recurring habits of visual behaviour, and a few are structural manifestations of geometry itself. They give the illusion of order, with price moving consistently in one direction. Yet trends can arise for many causal reasons such as random alignment, short-term mean reversion, or genuine structural imbalance. The difference lies beneath the surface. Only when feedback loops reinforce in one direction does structure create directional bias. Directional bias is the signature of structure, the persistent asymmetry that drives non-linear outcomes. When that bias strengthens, feedback compounds, moving the system far from equilibrium. That is when the market begins to self-organise and when outliers emerge. It is important to distinguish between directional bias that emerges within mean-reverting behaviour and directional bias that defines structural change. Mean reversion can produce short segments of apparent trend, yet these movements are constrained by negative feedback that restores balance. The bias exists, but it oscillates around equilibrium. Structural directional bias, by contrast, arises when positive feedbacks align and reinforce one another, driving the system away from equilibrium. The result is compounding displacement rather than restoration. In fractal terms, one is bounded motion within structure, the other is the transformation of structure itself.
Compression and Expansion in Fractal Structure
Fractal structure does not move in straight lines. It breathes. It compresses and expands, storing and releasing energy through the interplay of opposing feedbacks.
- Compression occurs when positive and negative feedbacks offset one another. Volatility contracts, liquidity thickens, and prices appear stable. Beneath the surface, energy accumulates. The system is wound tight.
- Expansion occurs when those feedbacks align. The stored energy releases, volatility surges, and directional bias becomes visible. Structure unfolds through price.
Compression and expansion are two phases of the same process, the market’s adaptation to imbalance. Compression hides structure. Expansion reveals it. This rhythm occurs at every timescale, from intraday fluctuations to multi-year moves.
Compression hides structure. Expansion reveals it. Outliers are large-scale structural expansions.
Outliers as Phase Transitions
In complex adaptive systems, a phase transition marks the moment when a system reorganises. Small changes accumulate until feedback, tension, or correlation push it past a critical threshold and a new state emerges. That is exactly what an outlier represents in markets. It is a phase change in structure. For long periods, feedbacks balance and volatility compresses. Energy is stored invisibly within the system’s microstructure. As tension accumulates, the market approaches a critical point. Then feedbacks synchronise, correlations spike, and the entire system shifts abruptly into a new configuration. This transition is expressed through price, the sharp, asymmetric, and seemingly explosive event we call an outlier.
- Compression Phase – Opposing feedbacks cancel, volatility falls, and potential energy builds.
- Critical Threshold – Feedbacks align, local perturbations trigger global change.
- Expansion Phase (Outlier) – Energy releases, volatility spikes, and the system reorganises into a new state.
Outliers are therefore not random anomalies. They are structural reconfigurations, the punctuation marks in the market’s ongoing evolution. Because the same process repeats across scales, fractal systems experience phase transitions both small and large. A micro burst on a one-minute chart and a decade-long bull market differ only by magnitude, not by mechanism.
Outliers are structural phase transitions, the visible moments when the market reorganises itself.
Fractality and the Power of Scale
In fractal systems, small and large events share the same underlying rules, but their impact scales non-linearly. Small fluctuations are abundant but cancel out. Large outliers are rare but dominate outcomes. The true power of markets lies in these medium- to long-term structural expansions, where feedback persists long enough to reshape the system. These are the moves that define portfolios, careers, and histories. To focus on them, we must truncate the distribution. We ignore the short-term and the dense bulk of small, cancelling events and concentrate on the tails where structure drives irreversible change.
The greatest power in the fractal landscape exists in the medium to long term, where structure compounds into outliers.
This is not a statistical preference. It is a philosophical one. We choose to engage with the regions of the distribution where reality expresses its architecture most clearly.
Clarifying Scale and Magnitude
Some might ask how a system can be described as scale invariant while its effects appear to compound toward the longer timeframes. The answer lies in understanding that scale invariance refers to form, not force. The geometry of feedback remains consistent across scales, but the magnitude of consequence grows nonlinearly with duration and persistence.
A tree provides the perfect analogy. Its branches, twigs, and trunk share the same fractal geometry, yet the trunk carries far greater magnitude. The same pattern repeats, but energy and material density accumulate as we move toward the larger scales.
Markets behave in exactly the same way. The feedback geometry that drives a short-term price burst is identical in form to the forces that drive a multi-year outlier. Structure repeats; energy compounds. In fractal systems, the rules of behaviour remain unchanged, yet their expression amplifies as feedback endures. The result is that form is self-similar, but outcome is non-proportional.
Fractality governs the form. Nonlinearity governs the consequence.
Trend Followers and Outlier Hunters
The trend follower and the Outlier Hunter both operate within the same universe of price, but they are not the same species. A trend follower trades form, reacting to the appearance of motion. A breakout, a moving average cross, or a new high in price are all examples of visual confirmation. It works while the pattern holds, but fails when form changes. An Outlier Hunter trades structure, responding to the conditions that make motion possible. We look for energy compression, volatility clustering, and feedback alignment across scales. Trend followers that trade patterns tend to be prescriptive about what constitutes a trend. Their models are fitted to specific visual forms of movement. An Outlier Hunter, however, is far looser in their interpretation of trend because it can take many shapes. They adopt loose-pants models and ensembles of systems designed to engage with structure rather than pattern. Trend followers trade the signal of motion. Outlier Hunters trade the geometry that creates motion. Trend followers expect trends to look similar. Outlier Hunters expect trends to take many forms such as slow grinds, fast shocks, long drifts, or parabolic expansions.
To align with fractal structure, your process must have loose pants.
It also requires small bets and open-ended profit targets to stay aligned with structure that is undefined. Markets express directional bias in many shapes, and you need room to move with all of them. Tight models fit patterns. Loose models adapt to structure.
Tight pants belong to pattern chasers. Loose pants belong to Outlier Hunters.
Structure bends, stretches, and expands, and your process must move with it. Both are reactive, but one reacts to form, the other to cause. The Outlier Hunter’s process is grounded in universality, the pursuit of asymmetry that can scale, repeat, and reshape the system.
The trend follower trades the appearance of motion. The Outlier Hunter seeks the geometry that makes motion possible.
Reactive Discipline and the Ensemble
Because structure is fractal, its expressions are infinitely varied. No single system can anticipate them all. That is why Outlier Hunters use ensembles of models, a family of systems each attuned to a different way structure might reveal itself. One listens for volatility expansions, another for smooth persistence, another for sudden breakouts. Together, they form a coherent, reactive framework that is not predictive, but receptive. We act when material trends appear, not because we know which will endure, but because structure eventually will. This is not pattern chasing, it is structural listening. Our rules create disciplined reactivity, a readiness to capture the few trends that reveal genuine structural transformation.
We do not forecast structure. We align with it.
Why I Call It Outlier Hunting
That is why I refer to my process as Outlier Hunting, not trend following. I am not chasing trends as visual patterns, because they can be superficial, as transient as the regimes and cycles they represent. What I am hunting for is structure, the architecture found in fractal systems, where deep asymmetries occasionally emerge as feedback loops align, equilibrium breaks, and price reveals its underlying geometry. These moments are not random. They are the causal drivers of structural change, the points at which the market reorganises itself and long-term returns are forged. To be an Outlier Hunter is to operate at this structural level. It is to respond to the unfolding order beneath apparent randomness, not to forecast it. It is to build systems that are reactive by design, yet tuned to recognise when the superficial gives way to the fundamental, when the pattern becomes structure.
Trends are the surface. Structure is the cause. Outliers are where the two converge.
The Shape of Understanding
Patterns are the transient forms of market behaviour, seductive, measurable, but fleeting. Structure is the enduring geometry of cause and effect, invisible, adaptive, and fractal. Regimes, cycles, and trends are patterns. Outliers are structure revealed. Fractality ensures that this interplay of compression, expansion, and transition repeats endlessly across all scales. The small cancels. The large defines. That is why Outlier Hunters truncate the distribution, to focus on the regions of greatest consequence where feedback compounds and structure becomes visible. We are reactive, not predictive, because we cannot know which patterns will reveal structure. But through disciplined process and ensemble design, we remain aligned with the unfolding geometry of the market itself.
Patterns are quickly exploited. Structure persists. Fractality ensures that power concentrates in the tails, and the disciplined process keeps us there when it arrives.