Out of session defaults to Normal regardless of what the volatility reading says.

The asymmetry it encodes

Defaulting to static costs a capped outcome. Defaulting to trend costs an uncapped one.

The direction of the default is chosen by comparing the two ways of being wrong. Handling a genuine continuation as a static trade produces a completed trade at a fixed target with a tail left on the table — a real cost, bounded, and paid on a winning trade. Handling a thin-liquidity range as a continuation produces an unprotected position past the static protection point, a trail engaged against movement with no flow behind it, and an exit determined by whichever direction the shallow book happens to drift. The first error is capped by construction. The second is not, and that comparison decides which way the switch falls.

Why no exception clause

An exception for exceptional conditions is an exception that fires exactly when judgement is worst.

The obvious refinement is to allow trend handling out of session when the reading is extreme enough or the structure clean enough. That refinement would fire in precisely the situations it should not: a dramatic overnight move with an extreme print is the most persuasive possible case, it arrives at the hour when the operator is least likely to be reasoning carefully, and it is also the case most likely to be thin-book movement rather than participation. A rule whose exception triggers on the strength of the appearance is not a rule with an exception. It is a rule that is suspended whenever it matters.

What is actually given up

Real trades are refused by this, and the refusal is the price rather than a bug.

It would be dishonest to claim the default costs nothing. Genuine continuation does occur outside major sessions — regional catalysts, scheduled events in other time zones, moves that begin overnight and carry into the next session. Trades in those conditions get static handling, reach a fixed target, and leave the rest. Over a year that is a measurable amount of forgone outcome, and an operator who tracks it will be able to name specific trades that would have paid more. The defence is not that those trades do not exist. It is that identifying them in advance, at the moment, is exactly what the operator cannot reliably do.

• The forgone tail is real and can be counted after the fact.
• Counting it after the fact is not the same as recognising it in advance.
• A rule justified by hindsight-identifiable exceptions has no stable form.

The route that does exist

If overnight continuation is a persistent edge in your book, it is a scenario, not an override.

There is a legitimate path and it does not run through the live moment. An operator who believes a specific class of out-of-session conditions genuinely supports continuation can define it as a scenario contract with stated expected behaviour, run it against evidence, compare it to the standard handling over a persistence window, and promote it deliberately with a rollback trigger. That process can change the doctrine. What it cannot do is change it tonight, on this trade, because the reading looks compelling — which is the entire difference between revising a rule and abandoning one.

Why fail-safe direction beats accuracy

The default is chosen for the cost of its errors, not for how often it is right.

This is the design principle underneath, and it is worth separating from the intuition that a rule should be as accurate as possible. Accuracy is the right objective when the two errors cost the same. When they do not, the correct default is the one whose errors are survivable, even at the cost of being wrong more often. The switch's out-of-session rule is almost certainly wrong on some nights. It is wrong in the direction that produces a completed trade with a smaller outcome, which is a category of mistake an operator can make indefinitely without it mattering much.

The key idea

Decide which way a rule should break before you are standing in the situation where it breaks.

Every rule fails somewhere, and the only question that can be answered calmly is which direction it should fail in. Answered in advance, it becomes a default with no exception clause and no negotiation at the moment. Left unanswered, it becomes a judgement made at whatever hour the situation arrives, under whatever conditions made the situation compelling. The out-of-session default is a small rule, and it is a complete worked example of the larger doctrine: settle the direction while nobody has money on it.