A Trading Journal for Decisions, Not Stories

A trading journal should preserve what was known and planned at the moment of decision. Without that record, hindsight can make a losing trade look obviously foolish and a winning trade look inevitable. A journal is not a diary of market stories or proof that a strategy works. It is a structured dataset for comparing rules with actions, expected risk with actual execution, and hypotheses with later evidence. Its value depends more on consistent fields and honest timestamps than on elaborate commentary.

Behavioral notes can reveal contexts associated with decisions, but they do not diagnose a mental-health condition. Use neutral descriptions of actions rather than clinical labels.

Capture the decision before the result#

Record the pre-trade fields before submitting an order:

• date, time, instrument, direction, and strategy tag;
• market context and the evidence making the setup eligible;
• exact trigger and order type;
• entry assumption, invalidation, and planned exit;
• expected risk per unit and total cash risk;
• current portfolio and correlated exposure;
• scheduled event risk and expected costs; and
• one fact that would argue against the trade.

Add a timestamped chart if useful. Do not edit the original thesis after the trade. Append updates separately so the record shows how information evolved. A blank required field should block the order under the journal process, not be reconstructed at the end of the day.

Record execution and management separately#

After each fill, capture requested and executed price, size, spread, fees, slippage, and any partial fill. During management, log only decisions: stop movement, scaling, cancellation, manual exit, or response to new evidence. Include the rule authorizing each action.

Use neutral context labels such as “entered 45 seconds after a prior stop-out” or “size exceeded calculated amount by 20%.” “Angry” or “overconfident” may feel explanatory but cannot be independently reviewed. A trader can optionally note self-reported intensity on a simple scale, provided it is treated as personal context rather than a diagnosis.

At close, record realized outcome in both currency and R-multiples, where one R is the initial planned cash risk. If planned risk was $80 and the net loss was $60, the result is -0.75R. Keep planned R separate from actual maximum exposure if rules changed.

Turn entries into a small review#

Suppose 30 journaled trades have 18 losses averaging -0.80R, 10 gains averaging +1.20R, and two flat results. The sample average is:

[(18 × -0.80) + (10 × 1.20) + (2 × 0)] / 30 = -0.08R

This does not prove the strategy has a negative long-run expectation; 30 observations may be too few and may come from one market regime. Next, separate the trades by rule adherence. Assume six violations averaged -0.70R while 24 compliant trades averaged +0.075R. That suggests violations deserve immediate control work, but the compliant result is still too small and uncertain to claim an enduring advantage.

Also compare planned and realized loss. If average losing execution is -0.80R because exits were efficient, that differs from a value caused by repeatedly moving stops. Numbers require the underlying decision records.

Run a fixed review cadence#

Use three layers:

1. Daily reconciliation: verify fills, fees, position state, and missing fields.
2. Weekly process review: score rule adherence, repeated execution errors, and exposure-limit use.
3. Periodic strategy review: examine setup performance, regimes, costs, and whether enough observations exist to consider a change.

Choose metrics before viewing results. Useful measures include compliance rate, average slippage, maximum planned exposure, holding-time distribution, and results by one or two predeclared setup tags. Change one major rule at a time and version it. Mixing results from different rule sets creates misleading averages.

Prevent journal failure modes#

Selective journaling—recording clean trades but omitting impulsive ones—makes the dataset reassuring and useless. Too much free text can also hide missing risk data. At the other extreme, dozens of fields may make completion impractical.

Other failures include optimizing rules to a tiny sample, creating new tags after seeing outcomes, and assuming correlation is causation. “Trades after lunch lost” may reflect a particular instrument, event schedule, or chance. Screenshots can be cherry-picked, and simulated fills can differ from executable prices.

Protect sensitive financial data, credentials, and account identifiers. A journal should never contain passwords, authentication codes, or unnecessary personal information. Back up records securely and control who can view them.

Key takeaways#

• Preserve the pre-trade thesis before hindsight can alter it.
• Separate planning, execution, management, and outcome fields.
• R-multiples aid comparison but do not replace cash exposure and cost data.
• Review adherence and strategy evidence on different schedules.
• Small, selectively tagged samples cannot establish a reliable edge.

This guide provides behavioral education, not a psychological diagnosis, therapy, or personal investment advice. Journaling cannot make a strategy profitable or prevent loss. If records show repeated loss of control or trading causes persistent distress, stop risking capital and seek qualified support.