How to Keep a Trading Journal and Review Decisions
A trading journal is a structured record of planned and completed trading decisions, market conditions, execution and later review. Its purpose is to improve the quality of evidence about a process rather than turn every losing trade into a mistake. Recording assumptions before outcomes are known is especially important for avoiding hindsight bias.
What to record before a trade
A useful journal begins with instrument, market, date, thesis, relevant data, alternative explanations and exact entry or exit conditions. Record the intended risk, available liquidity and what would invalidate the idea. A trade entered without a documented decision rule cannot later be evaluated fairly against a rule invented after seeing the outcome. Education or simulated trading can use the same discipline without placing any real orders.
Separate decisions from outcomes
A process can be reasonable yet lose money because markets are uncertain. Conversely, a poorly justified trade may make money by chance. Journal reviews should therefore score whether a decision followed the predefined method separately from the realised financial result. This avoids encouraging the trader to chase lucky outcomes and reject every losing setup without considering sample size.
Worked example: gross versus net return
Suppose a fictional position makes a gross gain of £120 but pays £8 commission and £22 in spread or slippage costs. Net gain is £90, before financing or taxes. If the original planned risk was £60, gross reward-to-risk is 2.0 and net realised reward-to-planned-risk is 1.5. This does not mean the strategy’s expectancy is positive: one trade supplies almost no statistical evidence.
Record market conditions objectively
Useful context includes pre-announcement consensus, realised volatility, bid–ask spread, session, trend regime and relevant scheduled catalysts. Use timestamps and link to market data snapshots rather than relying on memory. Avoid collecting excessive metrics that are never used in review. A journal should facilitate comparisons between like-for-like situations, not accumulate decorative statistics.
Performance statistics that matter
Across a suitable sample, examine average net win, average net loss, win rate, maximum adverse movement, drawdown, trade frequency and total trading costs. A common expectancy measure is win probability × average win minus loss probability × average loss, using mutually consistent net amounts. This does not establish future profitability: statistical uncertainty, selection bias and changing conditions can be substantial.
Worked expectancy calculation
Imagine 100 hypothetical trades: 45 winners averaging £160 net and 55 losers averaging £100 net. Net expectancy per trade is 0.45×£160 − 0.55×£100 = £17. The total expected sum based on these sample means is £1,700 for 100 trades. However, actual results can deviate materially, and the selected history may overstate edge if parameters were optimised on the same trades.
Tag errors without post-hoc invention
Distinguish rule violations, execution problems, data errors and planned decisions that simply failed. If a stop slips due to a gap, record the actual fill and compare with the realistic risk model rather than calling the whole loss a psychology failure. Track changes to strategy rules with effective dates so a later review can separate versions.
Protect records and personal information
A journal can contain account details, trading history and financial information. Store it securely and avoid putting passwords, API keys or unnecessary personal identifiers into unprotected documents. When sharing examples for education, remove sensitive data and clearly label hypothetical transactions. Review summaries periodically rather than reacting emotionally to individual outcomes.
For supporting explanations and common pitfalls, read trading plans and structured reviews.
Trading journal: a reproducible review template
This educational comparison uses hypothetical figures and is not a prediction or investment recommendation.
| Recorded field | Example entry | Reason to record |
|---|---|---|
| Instrument and session | EUR/USD, London session | Market and liquidity context |
| Idea and evidence | Hypothetical inflation surprise; pre-defined technical level | Original thesis before the outcome |
| Risk parameters | Fixed monetary budget, entry, invalidation, stop type | Planned exposure and constraints |
| Actual execution | Fill time, spread, slippage, commission | Reconcile costs with the plan |
| Outcome and review | Actual P/L, rule adherence, what changed | Separate process quality from a lucky result |
| Future adjustment | Change one testable process step | Prevent vague retrospective stories |
A useful journal distinguishes information available before the decision from explanations invented after the outcome. Store screenshots or chart references only where they support a dated record, and avoid placing personal account credentials in notes. Review a meaningful set of comparable trades to identify repeated cost, execution and discipline problems. A profitable individual trade can violate good process, while a planned loss can still be a correctly managed outcome. That distinction makes a journal more useful than a list of winning and losing tickets.
Frequently asked questions
Does every winning trade mean a good decision?
No. Outcomes and decision quality should be evaluated separately.
What is trading expectancy?
An estimate of average outcome per trade derived from win and loss frequencies and amounts.
Should commissions be included?
Yes. Execution costs can materially change a strategy’s net results.