Last Updated: 2026-10-26
A loss does not tell you whether a rule, an execution assumption or an accounting record failed. Use this checklist to reconstruct one trade before changing a strategy. The examples are synthetic. For order mechanics, use the beginner reference; for different trading rules, use the strategy comparison.
1. Preserve the decision that existed before the result
Record the rule version, timestamp and timezone, instrument, direction, observed signal, intended entry, invalidation condition, planned exits, quantity and maximum planned loss. Save the data available at that moment. A reason added after the loss is a new hypothesis, not evidence of the original decision.
Compare the plan with actual orders and fills. “The stop was at 95” does not mean an order was accepted, triggered or filled at 95. Record those events separately. A stop-market order prioritizes execution after its trigger but may fill worse; a stop-limit constrains price but may remain unfilled. Neither removes gap, outage or liquidity risk. Do not describe changing the order type as a universal fix.
2. Recalculate size with costs and a worse-fill case
In a deliberately simplified example, entry is $65,000, stop reference $63,700 and planned loss budget $200. Ignoring costs gives $200 / $1,300 = 0.153846 BTC, or $10,000 notional. That is a reference calculation, not a guaranteed loss cap or recommended budget.
At hypothetical 0.1% fees on entry and exit, loss per BTC at the reference exit is $1,300 + $65 + $63.70 = $1,428.70. The same budget would correspond to about 0.139987 BTC, before quantity precision and other expenses. If 0.139 BTC instead exits at $62,000, the loss is 0.139 × ($3,000 + $65 + $62) = $434.653. A worse fill can exceed the planned budget even after fee-aware sizing.
3. Use the correct denominator for trading costs
Twenty round trips with $1,000 filled on each side create $40,000 turnover. At a hypothetical 0.1% on every fill, fees are $40: still 0.1% of turnover. Against $2,000 account equity those fees are 2% of equity. Do not multiply by trade count again after summing all fills.
Use actual fill prices for cash P&L. If slippage is already reflected in those prices, do not subtract the same slippage again. A separate comparison with a reference price can explain execution quality without changing the cash result twice.
4. Keep incomplete records visibly incomplete
Scroll sideways to see every column. Keyboard: focus the table region and use the arrow keys.
| Field | Example or check |
|---|---|
| Identity | Venue, account, market, order ID and fill IDs |
| Execution | Requested quantity versus every partial fill; timestamps and fees |
| Inventory | Opening and closing units, including unclosed positions |
| Transfers | Link outgoing and incoming records without calling both trades |
| Cost status | Known, estimated or missing; state the valuation currency/time |
| Plan comparison | Followed, departed from, or impossible to determine |
A portfolio showing realized gains can still have losing open inventory. A transfer missing on one side can look like profit or a sale. Reconcile these before using the result for tax records. This checklist does not determine legal tax treatment.
5. Separate uncertainty from a reason to continue
For 55% wins averaging $10 and 45% losses averaging $15, expected gross outcome under those assumed rates is 0.55 × $10 − 0.45 × $15 = -$1.25 per trade. A high win rate alone is not an edge. An observed positive average likewise needs its sample dates, dispersion, dependence, costs and selection process; 100 trades is not an automatic certificate.
A review pause can be triggered by missing records, an order-state mismatch, an exceeded preset loss limit or a result that no longer matches the tested assumptions. Investigating does not require opening another trade. Retiring an unsubstantiated rule is a valid outcome. There is no evidence here that a losing strategy will recover if you persist.
Social posts, a wallet transfer or a token scanner flag do not establish someone's intent or future price direction. Record the observation and what it cannot show. Use a verified support route for account problems; never supply a recovery phrase in a trading journal or support message.
Finish with one concrete action: correct a record, revise a paper-test assumption, investigate an execution discrepancy, or stop using a rule. “Be more disciplined” is not a diagnosis. Keep the old result and the revised hypothesis separate so the next review can tell what actually changed.
Educational review process, not investment advice or a promise of profitability. Trading can lose capital. AITR may receive commissions from other commercial links on the site; this checklist does not endorse a venue.