Efficiency ahead of profit and profit ahead of efficiency are different illnesses.

The five states

Two agreements, two divergences, and one honest refusal to decide.

The engine classifies each composite each week into one of a small set of named states. Alignment means the two metrics agree, and the agreement can be positive — structure and monetisation both healthy — or negative, in which case the agreement is bad news stated clearly, which is still preferable to ambiguity. Under-monetisation means efficiency is the stronger of the two. Profit without efficiency means the reverse. Dual weakness means both are impaired, which is the structural danger condition. Transition means the signals are mixed or changing and no conclusion should yet be drawn. Alongside these sit the leading states, which describe one metric improving ahead of the other and are the constructive cousins of the two divergences. The set is small on purpose: a divergence taxonomy with twenty entries is a taxonomy nobody applies consistently.

Under-monetisation

The trading is working and the exits are leaving the result behind.

Efficiency stronger than profit quality means risk is being deployed well — the positions taken are converting into returns at a good rate relative to what was risked — while the profit-side reading fails to confirm it. The trades are being found and entered competently and the result is not arriving in the form the profit measure recognises. In practice this points at the monetisation side of the exit architecture: targets set too near, partials taken too early, trends released before they pay, or a branch mix weighted toward structures that cap the upside. It is the more benign of the two divergences precisely because the hard part is already working. Finding edge is the difficult problem; converting it is a tractable one, and the repair lives in exit rules and branch weighting rather than in the strategy itself.

Profit without efficiency

The account is up and the risk consumed to do it does not justify the result.

The reverse case is the dangerous one, and it is dangerous for the same reason profit with pain is dangerous in the capital layer: the operator is being paid while the process deteriorates. Profit quality reading high while efficiency fails to confirm indicates gains arriving through over-aggression or variance rather than through efficient deployment — larger risk producing proportionally smaller returns, with the shortfall masked by the fact that the absolute numbers are positive. The named diagnosis is that the profit appears high but risk efficiency does not confirm it, and the phrase does the work: confirmation is what is missing, not profit. This state resolves in one of two ways over subsequent weeks. Either efficiency recovers and the divergence was a lag, or the variance turns and the enlarged risk is still deployed when it does.

Why the direction changes the repair

The two states share a symptom and share almost nothing else.

Both divergences present identically at the level the operator first encounters them: two quality metrics that have stopped confirming each other, flagged with a severity. If the direction is not read, the natural response to either is to look at the strategy, and that response is wrong in both cases and wrong in different ways. Under-monetisation treated as a strategy problem leads to changing entries that were working. Profit without efficiency treated as a monetisation problem leads to holding for larger targets with risk that was already too large — which is the single worst available response, because it compounds the exact behaviour producing the divergence. This is why the engine names the state rather than reporting a magnitude. A gap of a given size tells the operator nothing actionable. The direction of the gap is the entire instruction.

Reading the leading states

Improvement also produces divergence, and it should not be treated as illness.

The leading states exist because a system that is getting better does not improve on both axes simultaneously. Efficiency improving ahead of monetisation is the expected signature of a repair to risk deployment beginning to take effect before the profit side registers it. Profit efficiency improving ahead of broader capital efficiency is the corresponding case on the other side. Both look like divergences and neither is a failure. Distinguishing them from the two problem states relies on trajectory rather than on the gap alone — which metric is moving, and in which direction, rather than merely which is currently higher. This is also why a single week's reading is never the unit of judgement here: a snapshot cannot separate a system pulling apart from one converging unevenly, and only the accumulated pattern can.

• Efficiency ahead: the edge is real, the conversion is not — repair exits and weighting.
• Profit ahead: the gains are not being earned — reduce aggression before anything else.
• Both metrics rising unevenly is improvement, not divergence. Read trajectory, not the gap.

The key idea

A named state carries an instruction; a measured gap does not.

The engine could report the numerical separation between the two readings and leave interpretation to the operator, and it would be simpler to build and considerably less useful. A number invites the question of how large is too large, which has no stable answer and which the operator will resolve differently depending on how the month has gone. A named state — with a defined meaning, a direction, and a routed review — converts the same underlying measurement into something that can be acted on identically every time it appears. That consistency is the point. The value of the engine is not that it detects divergence, which an attentive operator might notice unaided, but that it responds to the same divergence the same way in a good month and a bad one.