Welcome to the Machine: Part II

The 12 Temptations blog is an ongoing series examining how markets behave under stress. We deliberately avoid prediction and advocacy, focusing instead on structure, incentives, and behavioural dynamics.

When Optimisation Becomes Exploitation

The Way We Explain What We See

There are moments in markets that feel difficult to reconcile with the usual ways of understanding price movement. A level that has held for some time gives way quickly, stops are triggered in a cluster, and price reverses not long after. The sequence can appear precise, almost as though it followed a defined path rather than emerging from a series of independent decisions.

The instinctive response is to interpret this through intent. It feels reasonable to assume that someone observed those levels, formed a view, and acted on it. The language that follows reflects that interpretation. Stops are described as being hunted, liquidity as being targeted, and price moves as though they were anticipated in advance.

This is not simply a feature of financial markets. It reflects a broader tendency in how people interpret patterns. When outcomes appear coordinated, we look for a coordinating force. When behaviour feels directed, we assume direction. In many areas of life, this is a useful shortcut. It allows us to simplify complexity and respond quickly to what we believe we are seeing.

Markets, however, do not always lend themselves to that kind of interpretation. They are not single systems acting with a unified objective, but collections of participants, constraints, and processes interacting in real time. What appears deliberate at the surface level can, in some cases, emerge from the interaction between these elements rather than from a single source of intent.

This does not mean that deliberate behaviour does not exist. It clearly does. But it does suggest that not all structured outcomes require deliberate design. Some arise because of how behaviour tends to organise itself under similar conditions.

Structure and Response

As discussed in Part 1, many of the systems now active in markets are designed to optimise. Their role is not to interpret or explain market behaviour, but to operate within a defined set of objectives. These objectives are often practical rather than directional. They include reducing execution cost, improving fill probability, managing inventory, or responding to changes in liquidity.

In doing so, these systems respond to what is present in the market rather than what it might mean. They do not need to form a view on whether a level is significant in a broader sense. They only need to recognise that activity is likely to occur there.

Liquidity, in practice, tends to concentrate. It gathers around recent highs and lows, around levels that have been tested repeatedly, and around areas where positions are likely to be adjusted. Stop-loss orders, in particular, are rarely distributed evenly. They accumulate in places that feel logical to participants managing risk, often for similar reasons and using similar reference points.

This creates a form of structure that is not imposed from above but emerges from behaviour itself. When enough participants act in similar ways, the market begins to reflect those patterns. Certain areas become more active, not because they were designated as such, but because behaviour has clustered around them.

An optimising system does not need to identify these areas in a deliberate or strategic sense. It simply responds to the fact that they exist. Where activity is concentrated, execution tends to be more efficient. Where execution is more efficient, interaction becomes more likely. Over time, this reinforces the same areas of activity, not through intent, but through repeated response.

As more systems operate under similar constraints, the likelihood of similar responses increases. Without coordination, the outcome can still appear consistent. What might once have been occasional becomes more regular, and what was once interpreted as coincidence begins to feel structured.

This is where the experience of the market begins to change. The behaviour is not necessarily new, but its consistency can be.

When Outcomes Begin to Feel Extractive

From a human perspective, these patterns can feel uneven. Price appears to move toward areas where participation is most exposed, and then away again once that participation has been absorbed. When this happens repeatedly, it can create the impression that the market is acting against those participants in a deliberate way.

It is understandable that this interpretation emerges. The sequence of events aligns closely with how intentional behaviour would appear if it were taking place. There is a point of vulnerability, a move toward that point, and then a reversal once that vulnerability has been realised.

A different way to approach this is to consider that the market is responding to the behaviour within it. The outcome does not need to be designed in order to feel one-sided. When optimisation consistently interacts with predictable behaviour, the effect can resemble exploitation, even if no participant is explicitly attempting to produce that result.

This distinction matters, particularly in how participants respond. If the behaviour is interpreted as intentional, the natural reaction is to try to anticipate or counter the perceived intent. If it is understood as structural, the focus shifts toward understanding where behaviour is likely to concentrate and how systems are likely to respond to it.

This does not make the outcome any less real. For those on the wrong side of the move, the experience is the same. What changes is the framework used to interpret it.

Over time, this dynamic can influence how participants behave. If certain actions repeatedly lead to similar outcomes, those actions may be adjusted or avoided. In that sense, behaviour and structure begin to shape each other. Participants adapt to what they observe, and those adaptations become part of the environment that systems respond to.

Interaction Rather Than Attribution

This shift in perspective moves the focus away from attribution and toward interaction. Rather than asking who is responsible for a particular move, the question becomes how different elements within the market are interacting at that point in time.

Markets have always involved a degree of feedback between behaviour and outcome. What may be changing is the speed and consistency with which that feedback operates. Systems that are designed to optimise can identify and respond to patterns more quickly than human participants, and they can do so repeatedly without the same cognitive constraints.

This does not remove the role of human decision-making. Expectations, positioning, and reaction remain central to how markets function. What it does change is the environment in which those decisions take place. Behaviour that might previously have been absorbed or overlooked can become more visible, and therefore more consequential.

For individual participants, this introduces a slightly different challenge. It is not simply a matter of forming a view on direction or value, but of understanding how behaviour is likely to be expressed within the market. Actions that feel reasonable in isolation may not be unique. They may be shared by many others and therefore become part of a broader pattern.

Levels that appear logical may also be widely recognised. Responses that feel instinctive may also be anticipated. In an environment where systems are responding to patterns rather than narratives, this can affect how those patterns play out.

A Subtle Shift in Experience

None of this suggests that markets have become something entirely different. The same underlying elements remain in place. Participants act, positions are taken, and prices move in response to changing conditions.

What may be changing is the way in which those elements interact. As systems become more effective at identifying and responding to patterns in behaviour, those patterns themselves take on greater importance. The outcome is not necessarily a market that is more intentional, but one that is more responsive to what is already there.

The distinction is subtle, but it helps explain why certain experiences feel more frequent or more pronounced. Behaviour that might previously have passed without notice can become more visible when it is consistently engaged with. Over time, this can alter how participants perceive the market, even if the underlying mechanisms remain grounded in interaction rather than intent.

The challenge, then, is not that markets have developed a new form of intention.

It may be that they have become more effective in responding to the behaviour that is already present.

And in practice, that can feel much the same.