Post-Trade Review: A Decision-First Process

A post-trade review is a structured reconstruction of a completed trade. It compares the plan and information available at each decision point with what the trader actually did, then records the financial result separately. The purpose is not to explain every win or loss. It is to preserve enough evidence to assess strategy application, risk, execution, and behavioral drift without allowing the known outcome to rewrite the decision.

That makes a post-trade review different from simply adding a note to a trade log. A log records what happened. A review asks which standard applied, whether the action matched it, what changed during the trade, and what—if anything—should change next. A review is only as reliable as the evidence it starts from; a session that ends without a defined session shutdown often leaves the trigger, exposure state, and any re-entry undocumented before memory fades.

What should a post-trade review include?

A useful review needs six parts:

The plan before entry: setup criteria, invalidation, intended risk, management rule, and relevant session boundaries.
The information available at the decision: observations the trader could actually see at the time, not facts learned later.
The action taken: entry, size, management changes, exit, and any re-entry.
Rule status: aligned, planned exception, deviation, or unclassified.
The result: P&L, costs, and other outcome measures kept distinct from process classification.
One next action: retain the process, clarify a definition, collect more evidence, or test one specific change.

These fields answer different questions. A valid setup can lose. An ineligible entry can win. A correctly sized trade can still be entered late. Combining everything into “good trade” or “bad trade” hides the part of the process that needs review.

Review layer	Question	Evidence
Strategy application	Did the opportunity meet the trader’s defined setup?	Eligibility criteria and contemporaneous observations
Risk	Did planned and actual exposure follow the risk process?	Planned risk, invalidation, size, and changes
Execution	Were entry, management, and exit actions carried out as intended?	Orders, timestamps, prices, and decision notes
Behavior	Did pressure change a rule or boundary?	Trigger, intended response, actual response, and sequence
Outcome	What happened financially?	Gross result, costs, net result, and comparable outcome unit

Costante does not determine whether a setup has an edge or whether a trade should have been placed. Those remain strategy and execution decisions for the trader. A behavioral review can show that the recorded process changed; it cannot validate the underlying method.

Post-trade review checklist

Use this checklist without letting the result become the verdict:

Did the setup meet the eligibility criteria recorded before entry?
Was planned risk recorded before the order decision?
Did actual size match planned size and the stated sizing method?
Did the entry occur only after the defined trigger?
Did management and exit actions follow the original rule?
Was every claimed exception defined before it was used?
Did a session cutoff, re-entry limit, after-loss rule, or other active boundary apply?
Does any decision lack enough contemporaneous evidence to classify?
Are strategy, risk, execution, and behavior classified separately?
Is P&L recorded as an outcome rather than used as the process classification?
What is the smallest next action justified by the evidence?

A “no” is not automatically a mistake: an exception may apply, and missing evidence may require an unclassified label. The checklist identifies deeper review; it does not validate the strategy.

Why review the decision before the P&L?

Once the outcome is known, it can influence how the earlier decision is judged. In Baron and Hershey’s foundational experiments, participants evaluated otherwise similar decisions more favorably when they were paired with favorable outcomes (Outcome bias in decision evaluation). Aiyer and colleagues conducted a preregistered replication and extensions of Baron and Hershey’s Experiment 1 with a substantially larger sample and replicated the outcome-bias effect (Outcomes Affect Evaluations of Decision Quality). A broader review by Roese and Vohs distinguishes several forms of hindsight bias, including distorted recollection of an earlier judgment and beliefs that an event was foreseeable or inevitable after it occurred (Hindsight bias). These studies concern general decision making, not traders or trading outcomes. They support a narrower practice here: preserve what was known and planned at the time, then compare the decision with that record before interpreting the result.

Use this order:

planned standard → information available → action taken → rule status → result

Do not start with the chart after the move and ask what the trader “should have known.” A later chart contains information that was unavailable at entry. Preserve screenshots or notes from the relevant decision points when they genuinely help, but do not treat a screenshot as a substitute for a written rule.

A six-step post-trade review process

1. Select the review unit

Decide whether the review concerns one trade, one position with several orders, or a sequence such as a stop-out followed by re-entry. The unit should match the question.

Review one trade to inspect setup, size, management, and exit.
Review one position when scaling or multiple legs are part of a single idea.
Review a sequence when the prior result may have changed the next decision.

If the question is revenge trading, isolating only the second entry removes the loss and re-entry condition that made the pattern reviewable. If the question is an order-entry error, expanding the unit to the entire day may add noise.

2. Restore the pre-trade plan

Copy the plan as it existed before entry. Do not improve its wording during reconstruction. Record:

setup and eligibility conditions;
intended entry or trigger;
planned invalidation;
planned risk and size method;
management and exit rules; and
any active session, re-entry, or after-loss boundary.

If no relevant rule existed, mark it undefined. An undefined standard is a useful finding. Inventing one after the outcome creates false evidence of either adherence or failure.

3. Build a decision timeline

List the moments when the trader chose, changed, confirmed, or declined an action. Keep observations separate from interpretations.

Time	Observable event	Applicable standard	Action
09:42	Entry condition recorded as present	Enter only after defined confirmation	Entry submitted
09:48	Price moved toward, but did not reach, the planned management level	Hold full size until invalidation	Partial exit submitted
09:55	Invalidation reached	Exit remaining position	Position closed

This hypothetical timeline shows a management change even if the trade later made money. It does not prove why the trader changed the plan. Record a stated reason if one was captured, but do not infer fear, greed, or another internal state from the order sequence alone.

4. Classify each applicable decision

Use neutral labels with explicit meanings:

Aligned: the recorded action met the applicable rule.
Planned exception: an exception defined before the decision applied.
Deviation: the action conflicted with the applicable rule.
Unclassified: the rule or evidence is too incomplete to decide.

An exception created during the review is not a planned exception. If the original rule was impractical, preserve the deviation and consider revising the rule prospectively. This keeps the historical record intact.

Applying one of these labels correctly is not the same as verifying the label wasn’t influenced by something that shouldn’t have mattered — the trade’s own P&L, the account’s current drawdown state, or the prior trade’s result. How outcome bias distorts a post-trade review walks through exactly how a profitable or losing result can flip the strategy, risk, execution, or behavior label on its own, and how to test for cognitive bias in a trading review covers the general check for that failure mode across any ineligible variable, separate from getting the classification categories themselves right.

5. Add the outcome as a separate layer

Record gross and net results, fees, slippage, and a normalized unit such as R only when its definition is stable — verify the underlying journal data before that step, since a fill error, a missing execution, or an unreconciled gross-vs-net gap will misstate the record no matter how carefully the review itself is run. Then ask two independent questions:

What does the result contribute to strategy or risk evaluation across comparable trades?
What does the process record show about adherence on this decision?

One aligned loss does not invalidate a strategy. One profitable deviation does not validate a new rule. Strategy evaluation needs an appropriate sample and method; the post-trade review supplies a cleaner record, not a conclusion from one outcome.

6. Choose the smallest justified next action

End the review with one of four dispositions:

Finding	Next action
Process was clear and aligned	Retain it and continue collecting comparable observations
Evidence was missing	Make one essential field easier to capture at the decision point
Rule was ambiguous	Rewrite it with a condition, response, and exception before the next use
One deviation appeared	Record it; avoid redesigning the process from one occurrence
The same sequence repeated	Test one intervention aimed at the specific decision point

“Be more disciplined” is not a testable action. “Show the active re-entry condition before the next post-loss entry” identifies a cue, a decision, and an observable response. If the review points to a skill that has never been demonstrated reliably — rather than a single deviation from an established one — structured trading practice builds and measures that skill in isolation before the next live review has to judge it again.

Post-trade review example

In this hypothetical intraday trade, the plan defined a pullback setup, confirmation trigger, 0.5% account risk, and no partial exit before the first management level. A new-entry cutoff applied after 11:00.

Review field	Recorded evidence
Setup eligibility	All three predefined pullback conditions were recorded before entry
Planned risk and size	0.5% account risk; 100 shares calculated from the planned invalidation
Entry condition	Confirmation close above the reference level
Management rule	No partial exit before the first planned management level
Active boundary	New-entry cutoff at 11:00; entry occurred at 10:12
Actual sequence	Confirmation occurred; 100 shares entered; 50 shares exited early; remainder exited at the planned target
Financial result	Net profit after recorded costs

The review keeps each classification separate:

Layer	Classification	Reason
Strategy application	Aligned	Recorded setup conditions and entry trigger were present
Risk	Aligned	Planned and actual size matched
Execution	Deviated	Half the position exited before the management rule permitted it
Behavior	Unclassified	The record contains no applicable behavioral-rule conflict or evidence of motive
Outcome	Profitable	Net result was positive after costs

The profit does not retroactively convert the early-exit deviation into an aligned decision or establish why it occurred. The next action is to preserve the deviation and, on future management changes, capture the applicable rule, any predefined exception, and the stated reason for the change at the decision point.

A compact post-trade review template

Use the smallest template that preserves the evidence needed for the review question:

Review unit:
Review question:

PRE-TRADE STANDARD
Setup and eligibility:
Planned entry or trigger:
Invalidation:
Planned risk and size method:
Management and exit rule:
Active session or behavioral boundary:

DECISION TIMELINE
Time / observable event / applicable rule / action:

CLASSIFICATION
Strategy application: aligned / planned exception / deviation / unclassified
Risk: aligned / planned exception / deviation / unclassified
Execution: aligned / planned exception / deviation / unclassified
Behavioral rule: aligned / planned exception / deviation / unclassified

OUTCOME
Gross result:
Costs:
Net result:

NEXT ACTION
Retain / clarify / capture / test:
Evidence needed at the next review:

Not every field must be completed for every trade. If the review question is size drift, the plan, invalidation, intended size, actual size, trigger, and sequence matter more than a long narrative. The trade journal examples compare several record formats and show when a smaller log may be enough.

Review one trade, then look for a pattern

A single review can identify an obvious operational error or rule deviation. It usually cannot establish a recurring behavioral pattern or a strategy conclusion. For pattern review, group decisions using stable definitions and count both occurrences and opportunities.

For example:

Post-loss re-entry deviation rate
= deviations after the defined loss trigger
  / recorded occasions when that trigger occurred

Keep the numerator and denominator visible. Three deviations across four applicable occasions means something different from three across forty. Also retain unclassified occasions rather than silently treating them as aligned.

Compare like with like: the same rule version, review unit, and classification standard. If a rule changes, date the new version and begin a new comparison period. The broader trading performance framework explains how results, risk, and execution can remain connected without being collapsed into one score; process-versus-outcome feedback keeps the review from treating a favorable result as proof that a deviation was sound.

What metrics can a post-trade review track?

Decision-level metrics should retain their rule and denominator. Useful measures include setup eligibility, risk adherence, execution adherence, re-entry deviation, early-exit deviation, and unclassified-decision rates. Also count profitable deviations and aligned losses to keep outcome and process separate.

Execution adherence rate
= aligned applicable execution decisions
  / total classifiable applicable execution decisions

“Applicable” excludes decisions where the rule did not apply. “Classifiable” excludes insufficient records, which must remain visible through an unclassified-decision rate rather than being counted as aligned. Show counts with percentages and keep the review window and rule version consistent.

Do not collapse unrelated measures into a universal discipline score unless its purpose, weighting, and denominator are explicit. Broader analysis belongs in the linked trading-performance framework; these metrics ask how decisions compared with rules.

Post-trade review vs daily and weekly review

These are analytical units, not mandatory schedules. Match the level to the question.

Review level	Best-suited question	Evidence emphasis
Immediate capture	What fragile fact may be lost later?	Active rule, observed trigger, action, stated reason
Individual post-trade review	Did this trade follow the applicable plan?	Setup, risk, execution, behavior, and outcome
Session or daily trade review	Did decisions interact across this session?	Sequence, re-entry, cumulative risk state, cutoff
Weekly pattern review	Did the same defined deviation recur?	Counts, opportunities, rule versions, unclassified records
Longer-period performance review	How did results, risk, and execution behave across a relevant sample?	Comparable outcomes and process classifications; performance diagnosis when the limiting layer is unclear

Immediate capture does not require interpretation. Preserve fragile facts near the decision when practical; classify later when review will not compete with execution. No universal interval fits every workflow. For how to assign a distinct question to each of these horizons instead of repeating the same review at every interval, see trading review cadence.

Common post-trade review mistakes

Writing a story instead of reconstructing evidence

A fluent explanation can sound convincing while hiding missing facts. Start with the plan, timeline, actions, and rule status. Add interpretation only after the evidence is visible. That risk does not disappear when the fluent explanation is AI-generated rather than self-written, and the same evidence-before-interpretation discipline applies before the trade when the input is AI-generated market or strategy analysis instead of a completed trade’s own record.

Reviewing only large losses

Loss-only review makes outcome the sampling rule. Include profitable deviations and rule-aligned losses so process classification is not reserved for painful trades.

Changing several rules at once

Multiple simultaneous changes make it difficult to tell which process was actually used. Unless an immediate risk issue requires action, choose the narrowest change supported by the record and evaluate it prospectively. When a review sample surfaces more than one candidate gap, rank them by recurrence, materiality, and evidence sufficiency before picking which one to address first, rather than defaulting to whichever one is most memorable.

Treating a reason as proof of a cause

“I entered because of FOMO” may be a useful contemporaneous note, but it does not prove a causal mechanism. Pair self-report with observable events and actions, and describe the finding as an association in the recorded sequence. The same caution applies to missed entries: “I hesitated” is a label, not a diagnosis, and a five-question sequence can separate a hesitation-caused miss from a rule-based pass, a deliberate decline, or a setup that never fully qualified.

Turning review into self-judgment

Labels such as careless, greedy, or undisciplined do not locate the failed decision. A neutral trading-mistake classification makes it easier to match the intervention to a strategy, risk, execution, or behavioral problem.

Where Costante fits

A post-trade review becomes less reliable when the trader must reconstruct the original rule, context, and decision entirely from memory after the outcome is known. Costante is designed to preserve more of the behavioral evidence before the review begins.

Stage	Costante’s role in the evidence trail
Before the trade	Session planning, self-defined guardrails, and pre-trade checks preserve the intended process and relevant boundaries
During the session	Active re-entry, after-loss, and session boundaries can remain visible; low-friction trade and behavioral logging retains decision and session context
After the trade	Structured behavioral review compares intended rules with recorded actions, while behavioral cost attribution, discipline trends, and repeated-drift detection make recurring deviations inspectable

This workflow does not decide whether the trade was “good.” Costante reduces how much behavioral evidence must be reconstructed after the result is known; the trader interprets the record and decides what should change.

Costante does not generate or validate strategies, connect to a broker or exchange, execute or route orders, block trades, enforce account rules, determine appropriate risk, or decide whether a trade should be placed. Strategy quality, risk, execution, and every trading decision remain the trader’s responsibility.

Frequently asked questions

How soon should you review a trade?

Capture fragile facts—such as the active rule, decision reason, and sequence—as close to the decision as practical. Conduct deeper interpretation outside the live decision when attention is no longer needed for execution. The appropriate timing depends on the trader’s workflow; accuracy matters more than an arbitrary universal interval.

Should you review every trade?

Keep a minimum dependable record for every trade you intend to analyze. A deeper review can be triggered by a planned sample, a specific rule, an unclassified decision, or a defined deviation rather than by P&L alone. The sampling rule should be written in advance so profitable and losing trades are treated consistently.

Is a screenshot enough for a post-trade review?

No. A screenshot can preserve market context, but it usually does not show the applicable rule, intended risk, decision sequence, or whether an exception was defined in advance. Pair visual evidence with a compact written record.

What is the difference between a trade journal and a post-trade review?

A trade journal is the record system. A post-trade review is the analysis performed with that record. The journal may contain transaction facts, plans, screenshots, and notes; the review compares the applicable standard with the recorded decision and selects a justified next action.

Sources

Costante provides educational workflow tools, not financial advice. Trading involves risk.