Skip to short-horizon research
Short Term Trading StrategiesCondition / Request / Position

A short-term trading journal that explains the result

A detailed journal for plans, orders, fills, cancellations, costs and amendments, with worked R-multiple examples and a blank CSV.

Original editorial revision: 9 October 2026. Expanded edition: 14 October 2026.

Educational execution planning. Invented costs and clocks are assumptions, not forecasts, broker instructions or a measured strategy record. Check the exact venue, instrument and account rules.

A useful trading journal lets you reconstruct a session without relying on memory. It records what was planned, what was sent to the broker, what filled and what the completed position earned after costs. Notes about confidence can help explain behaviour, but the numbers need a consistent record before they can support any conclusion about a strategy.

Give the signal and the trade different identifiers

One signal may produce no trade, a partial fill or several orders. Give the original signal a stable ID and attach all related order and fill IDs to it. Preserve the first-version entry, stop, target, size rule and timing rule, then append corrections with their own times. Overwriting the first plan makes it impossible to distinguish a changed instruction from an execution difference.

Use one stated timezone, ideally storing an unambiguous timestamp alongside the displayed session time. Keep the market session date separate from the calendar date where sessions cross midnight. A clock labelled “10:45” without a timezone and date is not enough to reconcile a signal with broker execution.

Download the blank event journal (CSV). Use one row per plan, order, fill, cancellation or correction, linked by signal and position IDs. The file contains headers only, not a fictional trading record. Record charges once, preserve the source reference and identify corrected rows rather than overwriting them.

The minimum fields for a reproducible session record
RecordFields to preserveQuestion it answers
Original planSignal ID, first publication time, direction, levels, units and deadlineWhat was decided before the outcome?
OrderOrder ID, type, submission time, requested quantity and amendmentsWhat instruction actually reached the broker?
FillFill ID, timestamp, quantity, price and venue where suppliedWhat executed, including partial fills?
ExitExit fills, reason code and remaining position quantityWhy and when did the position close?
CostsCommission, fees, financing and any separately estimated slippageWhat separates gross P&L from net P&L?
Unfilled instructionCancellation or expiry time and reasonWhich signals produced no funded trade?

Keep three result columns

The planned outcome is the calculation from the original levels. Gross executed P&L comes from the actual fills. Net executed P&L subtracts recorded costs. Showing all three exposes whether a weak result came from the rule, the execution or the cost assumption. Avoid subtracting slippage twice: if P&L already uses actual fills, the difference from the planned price is already reflected in it.

Choose the original cash risk as the denominator for R. If a hypothetical trade originally risks 20, a net result of minus 23 is minus 1.15R. Keep that denominator fixed even if the stop was widened later. Recalculating the denominator after the loss can make a breach look smaller without changing what the account lost.

A hypothetical long trade, line by line

Suppose the written plan is long 10 units at 100, stop 98 and target 104. Planned cash risk is 10 × 2 = 20. The entry actually fills at 100.10 and the exit at 98.00. Gross executed P&L is 10 × (98.00 - 100.10) = minus 21. With 2 of recorded round-trip costs, net P&L is minus 23, or minus 1.15R against the original 20 budget.

That record shows a 1.00 entry-price difference across the position plus 2 of costs beyond the clean planned loss. It does not require a story about whether the trade felt unlucky. The entry price, quantity, exit price and costs explain the difference. A stop that fills below 98 would create a further execution loss; the SEC stop-order bulletin explains why the trigger price is not a guaranteed fill.

Partial exits must reconcile to one position

Using the same hypothetical entry of 100.10, suppose five units exit at 102 and five at 104. Total proceeds are 5 × 102 + 5 × 104 = 1,030 against an entry cost of 1,001. Gross P&L is 29; subtracting 2 of costs gives 27, or plus 1.35R. The weighted exit price is 103. Counting each exit as an independent winning trade would inflate the trade count for one original position.

If five units remain open, retain their marked value separately. Do not describe the whole position as completed while exposure remains. A cancelled order with no fill contributes no executed trade P&L, but it still belongs in the signal ledger. Declare how unfilled and cancelled instructions affect any signal-level statistics.

Compare the deadline with the actual fill time

The same-session exit clock produces a planning boundary. With a 09:45 entry and a 60-minute maximum hold, that boundary is 10:45 unless an earlier session cutoff binds. An exit fill at 10:46:30 is later than the boundary even when the trade makes money. Preserve the order submission time too: an order sent on time and filled late is a different event from a late decision.

The clock also compares optionally entered closing-order and fill times, including seconds. Its JSON export preserves those inputs and the timing difference, not a verified execution record. Match the entered times to the original broker IDs, date and timezone in the journal.

Tag timing exceptions with a specific cause such as a late instruction, rejected order, partial fill or market halt. Keep the original deadline visible. Replacing it with the actual exit time would remove the very information the journal is meant to explain.

Review the session without optimising from one anecdote

First reconcile quantities and cash with the broker record. Then total completed positions, remaining exposure, costs and net P&L. Retain sessions with no trades. A trade win rate and a profitable-session rate have different denominators, so label them separately. An unfinished position also changes what can be concluded from that session alone.

Across a declared continuous period, inspect average winning and losing R, total costs, rule breaches and the share of instructions that never filled. Look for repeated execution problems before adjusting model thresholds. A small group of disappointing trades may justify investigating a process, but does not establish which parameter change will improve future outcomes.

The FINRA day-trading risk disclosure identifies costs and execution difficulties as consequential on the fast clock. Its disclosure concerns securities day trading; the appropriate costs and venue mechanics for other instruments need their own records.

A complete journal preserves both the plan and the funded result. Use it with the risk-management playbook and strategy evidence check to evaluate a process over a continuous run.

Keep reading

Freeze cash inclusion beside original risk

The new lens includes the full incurred entry fee and both declared fragment fees at the snapshot, without prorating the entry fee into a claimed broker realised-P&L amount. Its closed-leg and total-at-mark R values share the original entered risk denominator. The existing journal's fixed-risk and split-exit lessons remain useful, but neither convention supplies actual orders or fees. Preserve the chosen cash inclusion and any amendment with real source records rather than borrowing favourable numbers from different examples to invent a completed trading result.