Post-trade review (the execution-quality loop)

Definition

Post-trade review is the closed-loop audit of completed trades against their plans: was the thesis right, was the process followed, and what did execution actually cost versus paper (ms-implementation-shortfall’s framing brought to account scale). It is distinct from disc-journaling-review — journaling captures the record and the psychology continuously; post-trade review is the periodic ANALYTICAL pass that turns the record into process changes. The institutional version is TCA plus strategy attribution; the KB carries the account-scale protocol, which needs only the journal, fill data, and honesty.

How it works / structure

• The three audits per trade: THESIS — right, wrong, or unresolved, judged against the stated falsification marker (not the P&L — disc-process-vs-outcome’s quadrant: good-process losses and bad-process wins are the informative cells); PROCESS — checklist compliance, size rule, exit execution vs plan (disc-pre-trade-checklist override log feeds here); EXECUTION — fill vs decision price per Perold’s decomposition (delay, spread/impact, opportunity cost of unfilled intentions — the account-scale shortfall ledger).
• The aggregation pass (where the value is): single trades are noise — the review’s unit of insight is the COHORT: by setup type (disc-playbook-setups — which playbook entries actually carry the P&L), by axis compliance (style-trading-types-taxonomy drift costs), by exit type (stopped vs target vs discretionary — discretionary exits’ cost is a standard finding), by hour/day, by override status; sample-size discipline applies (dozens per cell before conclusions — quant-backtest-hygiene inference standards).
• The change protocol: findings become ONE process change at a time, dated and tested forward — batch changes make attribution impossible (the same single-variable discipline the KB requires of backtests); the review cadence is calendar-fixed (weekly light, monthly full — practitioner convention, labeled).
• The bias audit (engine-relevant): the review is where Pillar 13 becomes measurable — disposition effect (winners’ hold time vs losers’ — bias-disposition-effect’s per-account test), revenge-trade clustering after losses, size creep after win streaks (bias-overconfidence); each has a journal-computable statistic.

When it applies

Every account, every cadence — the review is the only mechanism by which experience becomes improvement rather than repetition (the documented deliberate- practice distinction); mandatory after drawdowns (disc-drawdown-protocol triggers a full review) and after strategy changes (forward-test verdicts come from here).

Risk profile & failure modes

• P&L-only review (the dominant failure): grading outcomes instead of process re-trains the trader on noise — the quadrant discipline exists because variance swamps skill at single-trade scale.
• Selective review: skipping the review after bad weeks — exactly the weeks with the information; calendar-fixed cadence exists for this (bias-loss-aversion avoidance behavior).
• Metric gaming: once a review metric drives decisions, behavior optimizes the metric (fewer logged overrides via not logging) — the override log’s integrity is the review’s foundation, and auditing it is part of the review.
• Change churn: acting on every cohort wobble — sample-size impatience converts the review into a random-process generator; one tested change at a time.

Evidence & limits

Perold (1988) anchors the execution decomposition; the bias statistics are the cited entries’ peer-reviewed effects (Odean’s disposition evidence et al) made per-account; cadence and protocol specifics are practitioner convention, labeled. The review’s efficacy per account is itself testable — the before/after process-change records are the evidence.

Falsifiable-thesis examples

Illustrations only, not signals:

• “Discretionary exits cost this account money vs plan exits over 100+ trades (exit-attribution check)” — falsified by the paired exit-cohort comparison.
• “Average holding time of losers exceeds winners in this account (disposition-effect audit)” — falsified by the journal’s duration statistics.

Cross-references

• The record it consumes: disc-journaling-review, disc-pre-trade-checklist
• The judgment frame: disc-process-vs-outcome; the execution ledger: ms-implementation-shortfall
• The bias instruments: bias-disposition-effect, bias-overconfidence, bias-loss-aversion
• The inference standards: quant-backtest-hygiene