A trading journal becomes useful when its data supports a precise decision: keep a method, correct an execution, or test a new rule. The goal is not to accumulate screenshots, but to build a comparable history and then establish a repeatable analysis routine.
In brief
- 1R represents the initial risk of a position: this unit lets you express each result relative to the risk planned before entry, according to the Van Tharp Institute’s method.
- A loss can exceed −1R: initial risk is a calculation reference, not a guarantee of a maximum loss, as the Van Tharp Institute points out.
- A 95% confidence interval helps quantify uncertainty around a proportion, according to the methods presented by NIST; applying it to the win rate requires, among other things, checking the statistical assumptions.
What should you record in a trading journal?
A usable journal should bring together the initial plan, actual executions, and the information needed to classify each position.
CME Group recommends recording the reasons for entry, targets, entry and exit points, and the timing. The institution also recommends keeping daily conclusions to help identify recurring mistakes.
To put this approach into practice, use a structured form rather than free-form notes that cannot be filtered:
- Identification: instrument, contract type, account, currency, date, and time zone.
- Plan: setup name—the configuration you are looking for—entry condition, initial stop, target, and planned monetary risk.
- Execution: size, actual execution prices, partial exits, timing, and stop adjustments.
- Result: net profit or loss, costs, duration, and result expressed in R.
- Context: session, trend or range-bound market, and proximity to an economic release.
- Observation: whether the rule was followed or broken, emotional state before entry, and chart screenshot.
Enter the plan before placing the order, the executions after the position is closed, and the comments at the end of the session. Keep labels simple: the same setup should not change names depending on its result.
Also adopt a written convention for multiple entries and partial exits. To analyze an overall decision, group the executions belonging to the same plan while retaining their details to monitor costs.
How should you use R multiples?
The R multiple measures a position’s result relative to its initial risk, not relative to total account capital.
The formula is: result in R = net monetary result ÷ initial monetary risk. The Van Tharp Institute explains this normalization and defines expectancy in R as the average of the multiples obtained.
In your journal, specify whether initial risk includes an estimate of costs. Then keep that convention unchanged: changing the denominator after a stop adjustment would distort the comparison.
Also distinguish the reward-to-risk ratio considered before entry from the multiple actually achieved after the exit. The first describes the plan; the second measures its result.
Normalizing in R makes it easier to compare trades of different sizes, but it does not replace tracking results in euros. Keep both views: one to examine the method, the other to monitor the impact on the account.
Which statistics should be analyzed together?
The win rate should be considered alongside average profit, average loss, and expectancy, because the frequency of winning trades alone does not describe the overall result.
The win rate is the proportion of closed positions with a positive net result. State how breakeven positions are counted and keep that convention consistent. The general definition of a proportion—successes divided by trials—is recalled by NIST.
Expectancy, or observed expectancy per trade, is calculated as follows: winning proportion × average win − losing proportion × average loss, with the latter expressed as an absolute value. CME Group presents this relationship to show why the size of results matters as much as their frequency.
In your dashboard, also display the sample size, cumulative net result, and median result. Examine exceptional gains separately: an average driven by an atypical position deserves specific analysis.
How should you track drawdown correctly?
Drawdown measures the decline from a previous high in the account curve, not simply the sum of losing trades.
The CFA Institute defines maximum drawdown as the largest decline observed from a high to a subsequent low during the period under review. Expressed as a positive amount, it corresponds to the difference between the high and the low divided by the high.
For your records, separate the balance curve, based on closed trades, from the total account value curve, which includes open positions. Specify the observation frequency and neutralize deposits or withdrawals so they are not confused with performance.
Measurement frequency matters: widely spaced observations can conceal intermediate declines, as Goldberg and Mahmoud explain in their research on drawdown risk.
In a prop firm, track the relevant program constraints separately. Check the contractual documents for the calculation basis, whether unrealized results are included, and the reset time; do not use your historical statistic alone as a compliance check.
When do statistics become interpretable?
Statistics become more informative when there are enough observations, when those observations are comparable, and when uncertainty is measured.
NIST highlights the limitations of common approximations when the sample or the number of failures is small. An observed rate remains an estimate, not a definitively established property.
For your trades, do not automatically apply a binomial model: positions taken during the same move may be dependent, and market conditions can change. An interval calculated without checking these assumptions may create a misleading impression of precision.
Start by comparing results by setup, then by session or context. Avoid adding filters until you find an attractive combination. If a segment contains few positions, classify your conclusion as a hypothesis to test rather than a validated rule.
How should you organize a weekly review?
A weekly review should produce a written diagnosis and a verifiable change, not simply a reading of the net result.
Here is a practical process:
- Verify the data: reconcile the journal with the execution history, and check for duplicates, fees, and incomplete positions.
- Review the big picture: sample size, net result, expectancy in R, average gains and losses, and drawdown.
- Compare categories: setups, sessions, and contexts, always displaying their sample sizes.
- Inspect anomalies: unusual loss, early exit, modified risk, or incorrectly recorded cost.
- Write the conclusion: what is documented, what remains uncertain, and what will be tested.
Keep successive versions of your plan. Without a history of changes, you will no longer know which rules applied to the positions under review.
In Edgyx, the automatic journal and the AI coach Ora can support this review; you remain responsible for verifying the data and formulating the rules.
The useful output of a review is not an impression: it is an observable rule and a reassessment criterion.
How can you turn data into rules?
An actionable rule connects a documented observation to a condition, an action, and a control method.
Use this template: “When this condition appears, I follow this procedure; I reassess it based on these observations.”
For example, if your notes show impulsive entries after a loss, test a requirement to rewrite the rationale before placing another order. If partial exits make your statistics inconsistent, first define how they should be grouped before changing the strategy.
Distinguish journal-keeping corrections from changes to the method. A data error can be corrected immediately; a trading filter deserves evaluation on new observations, separate from those that inspired the idea.
Change one element at a time whenever possible. Record the reason for the change, its implementation date, and its abandonment criterion. Also assess compliance: a rule that is never applied cannot be tested properly.
Key takeaways
A useful trading journal connects a reliable history to explicit decisions that can be reassessed.
- Record the plan before entry and the actual executions after the position is closed.
- Establish consistent conventions for costs, breakeven positions, and partial exits.
- Analyze win rate, average results, expectancy, and drawdown together.
- Display the sample size for each category and distinguish observations, hypotheses, and validation.
- End every review with an observable procedure, then test it on new data.
