Conservative variants are permitted by default. Aggressive ones have to be earned.

The stated position

Aggressive variants must be harder to qualify than conservative ones, and the matrix is built to enforce it.

The asymmetry is written into every layer of the selection machinery rather than left as an attitude. The score bands allocate more range below the baseline than above it. The gate permission table removes aggressive postures progressively as drawdown deepens, and removes them entirely well before trading itself stops. The veto list is populated almost exclusively with conditions that block aggression rather than conditions that block caution. The disagreement rule downgrades. Each of these could have been built symmetrically and none of them was, because the two directions of error are not equally recoverable.

Why the errors are not equivalent

Under-monetisation costs opportunity. Over-extension costs capital state, which costs future opportunity.

Taking a conservative posture during a clean trend surrenders part of the right tail. The cost is real and it compounds, but it arrives as expectancy foregone from a stable equity base, and the next cycle begins with full deployment authority intact. Taking an aggressive posture into a deteriorating structure produces losses that are wider than the branch's normal profile, and those losses move drawdown, and drawdown moves the gate, and the gate constrains the tier ceiling for every subsequent cycle until it is recovered. One error costs a slice of one trade's outcome. The other costs a slice of every trade for as long as the recovery takes. That difference, not risk aversion, is why the burdens of proof differ.

The over-conservatism failure is still a failure

The asymmetry is not permission to default to caution forever.

It would be a misreading to conclude that defensive postures are always safe choices. Overusing early-banking management in genuinely clean trend regimes under-monetises the right tail systematically, and on the trend branches the right tail is the entire economic case for their existence. A Trend No-Partial trade managed with staged early monetisation is not a cautious version of TNP; it is a Normal trade wearing a TNP label, contributing variance to the branch statistics without the compensating tail. The matrix is designed to prevent unearned aggression, not to prevent aggression, and a book that has not used an aggressive posture in six months is describing either an unusual market or an operator ignoring half the instrument.

• Chronic conservatism shows up as a compressed right tail in the far-rung hit rates.
• It is invisible in expectancy for a long time, because the losses it prevents are also real.
• The variant usage audit exists to make the drift visible before it becomes the habit.

Why the same variant is not equally aggressive on every branch

A posture inherits the variance profile of the branch it is applied to.

The five variant names are constant and their consequences are not. Delayed protection on Normal stretches a branch that was designed to be static and closes at a fixed target — the posture is out of character with the branch, and the risk is that it quietly converts a Normal trade into an unmanaged trend position. The same delay on Trend No-Partial is an intensification of what the branch already does, which is why it is described as the highest-variance combination available anywhere in the system. Judging a variant request therefore requires naming the branch first, and a scorecard total assembled without that context is answering an incomplete question.

How the burden is discharged

Evidence, not conviction — and the evidence has to come from lenses that could have said no.

The practical test for whether an aggressive posture has been earned is whether the supporting evidence came from instruments capable of refusing. Expectancy readings, gate state, structural diagnostics, volatility regime and recent execution quality are all measured before the trade exists and none of them is influenced by how the setup looks. A conviction about the chart is not evidence in this sense, because it was formed at the same moment as the desire to act on it. The whole apparatus of the seven lenses is a device for ensuring that the answer to 'may I be aggressive here' comes from somewhere other than the part of the operator that wants to be.

The key idea

Symmetric permission for asymmetric consequences is how a good system develops a bad month.

If both directions were equally easy to reach, the postures would be selected by whichever felt right, and what feels right correlates strongly with recent outcomes — aggressive after wins, defensive after losses, which is precisely backwards relative to what the evidence usually supports. Making one direction harder to reach than the other does not remove that pull. It just means the pull has to survive contact with six instruments before it becomes a position size.