Journaling & review

Definition

The trading journal records what was done AND WHY at the time of decision — thesis, setup, size, emotional state, expected outcome — creating the dataset that makes self-review possible. Its purpose is to defeat memory: the documented biases (bias-recency, hindsight revision, selective recall of wins) systematically rewrite un-recorded history, so a trader without a journal reviews a fiction. For this platform the journal is automatic — every agent decision is logged with its inputs — which shifts the discipline from record-keeping to actually running the review loop.

How it works / structure

• The entry schema (engine-native): per trade — the setup tag (disc-playbook-setups taxonomy), the thesis and its falsifier (the platform’s structural requirement), size and risk at entry, the plan’s relevant clauses, and at exit — outcome vs falsifier (thesis right/wrong vs trade won/lost, the disc-process-vs-outcome grid), execution quality (slippage vs intent), and rule adherence (in-plan or deviation).
• The review cadences: per-trade closeout (was the falsifier respected?), weekly pattern scan (setup performance, deviation frequency), and the periodic deep review — cohort statistics by setup, regime, and holding period, hunting the specific leaks (documented common findings: a profitable strategy inside a losing account, the losses living in unplanned trades and oversized exceptions).
• What the record enables (the payoff): expectancy by setup (which playbook entries earn their place), deviation cost measurement (what rule-breaking actually costs, in dollars — the number that changes behavior), and bias fingerprinting (holding losers longer than winners is measurable — bias-disposition-effect diagnosed from one’s own ledger).
• The labeled evidence: journaling’s support is practitioner canon (Steenbarger) plus the general documented value of decision records against hindsight bias; controlled trading-specific studies are thin (labeled) — the platform’s position is that the MEASUREMENT layer it enables is where the value is proven, account by account.

When it applies

Every live strategy (unreviewed strategies decay undetected — philosophy-adaptive-markets lifecycle monitoring requires exactly this data); deviation management (the deviation log is the discipline metric); strategy retirement decisions (the journal’s cohort statistics are the evidence base).

Risk profile & failure modes

• Outcome-only journaling (the common failure): recording P&L without thesis and process captures nothing reviewable — the WHY at decision time is the irreplaceable field.
• Review theater: journals written but never aggregated change nothing — the cadence and the cohort statistics are the active ingredients.
• Hindsight contamination: entries edited after outcomes destroy the dataset — append-only records (the platform’s logs are immutable by design).
• Self-flagellation drift: reviews that grade outcomes instead of process teach fear, not skill — disc-process-vs-outcome is the grading rubric.

Evidence & limits

Practitioner canon (labeled); the bias literature it counters is peer-reviewed (pillar 13); the platform’s automatic logging removes the compliance problem that undermines most human journaling. The review loop’s value is measurable per account — stated as an auditable claim, not an article of faith.

Falsifiable-thesis examples

Illustrations only, not signals:

• “This account’s out-of-plan trades cost more than 2% of annual return (deviation-cost measurement)” — falsified by the tagged-ledger arithmetic.
• “Setups in the bottom expectancy quartile for 2+ quarters, retired, improve account-level expectancy the following quarter (pruning-loop check)” — falsified by the before/after comparison.

Cross-references

• The constitution it audits: disc-trading-plan
• The grading rubric: disc-process-vs-outcome; the loss ritual: disc-drawdown-protocol
• The taxonomy: disc-playbook-setups
• The biases it defeats: bias-recency, bias-disposition-effect