A useful journal is a decision record, not a scrapbook of winning charts. It preserves what was known before execution and compares the action with the plan, allowing a controlled loss to be distinguished from a preventable process failure.
Capture the pre-trade thesis
Record the market condition, evidence, activation, invalidation, intended horizon and why the instrument was selected. Screenshots should show the information available at the time rather than only the final outcome.
- Market regime
- Setup and confirmation
- Entry and invalidation
- Planned quantity and risk
Record execution separately
Note actual fill, spread, slippage, changes to quantity and whether the exit followed the plan. This separates research quality from order execution.
Score rule adherence
A binary or simple scale can track whether the decision followed defined rules. Profit should not convert a rule violation into a good process, and a normal stopped trade should not automatically be labelled a mistake.
Review patterns at fixed intervals
Look for repeated errors, market regimes, time-of-day effects and concentration only after enough observations exist. Change one rule at a time so the effect can be understood.
Turn review into one controlled experiment
At each review interval, identify the highest-cost repeatable error and choose one specific change. Define how long the revised rule will be observed before judging it, unless a safety limit requires an earlier stop.
Preserve the original data when changing a rule. Overwriting old classifications can make the new process appear better through hindsight rather than genuine improvement.
- Priority error
- Single rule change
- Observation period
- Success and stop criteria