Sizing, stops and invalidation

Every guide in this academy is worthless without this one. Edges are small and streaky; the only reason a small streaky edge compounds into anything is that no single expression of it can hurt you. That is a sizing statement, and sizing is arithmetic, not feel.

Work backwards from the stop

The order of operations is fixed: invalidation first, then risk, then size. Decide where the idea is wrong — beyond the wall, back inside value, through the flip. Decide what fraction of the account one wrong idea may cost; for most intraday traders something under one percent is the ceiling that survives a losing week. Size is whatever makes the distance to the stop equal that risk. It is an output, never an input.

Sizing first and hunting for a stop that “gives it room” inverts the logic — the position now dictates the analysis, and every added tick of room is added risk with no added edge.

Invalidation is a price, not a feeling

The level-based framework makes this concrete: a fade at the call wall is invalid on acceptance above it, not on discomfort. Writing the invalidation down before entry does two things — it makes the stop non-negotiable, and it makes post-trade review honest, because you can check whether the exit followed the plan or the pulse. The reluctance this discipline fights is the disposition effect, documented across tens of thousands of retail accounts by Odean (1998).

The compounding rules

• Risk a fixed fraction, not a fixed amount — size then shrinks in drawdowns automatically, which is precisely when your judgement is worst.
• One loss should be boring. If a single stop-out changes your mood, it was too big by definition.
• Correlated trades share a budget: three longs against the same put wall are one idea wearing three tickets.
• When the regime is unclear — mixed gamma, no confluence — the correct size is frequently zero, and zero is a position.