Profit tells you what happened to the account during the week. It does not tell you whether the decisions were good. A profitable week may have been built on excessive risk, luck, and rule-breaking. A losing week may contain disciplined trades with positive expectancy that happened to fall into an unfavorable run of outcomes.
The central question in a weekly trading review is therefore not “How much did I make?” but “Given the information available at the time, did I make repeatable decisions that respected my risk rules?” P&L stays in the review, but it becomes one layer rather than the final verdict.
A useful review separates three things: the quality of the strategy process, the quality of execution, and the market outcome. This distinction helps a trader avoid rewarding a lucky mistake, avoid punishing a correctly executed loss, and choose one concrete improvement for the following week.
Why profit can mislead
Trading takes place under uncertainty. At the moment of a decision, no one knows which possible scenario will occur, even when a strategy has an edge over a long series. One trade, or even one week, is too small a sample from which to judge the quality of a system reliably.
This is where outcome bias appears: the tendency to judge an earlier decision by how it turned out rather than by the information and process available when it was made. In a study by Jonathan Baron and John Hershey, participants judged decision-makers more favorably when a decision under uncertainty produced a good result, even though the information available before the decision was unchanged. In trading, the effect is especially dangerous because profit delivers immediate emotional reinforcement.
A simple way to expose this bias is to place every trade into one of four categories.
| Decision quality | Outcome | What it means |
| Good decision | Profit | The process was sound, and the market rewarded it |
| Good decision | Loss | A normal cost of the strategy; review it, but do not automatically “fix” it |
| Bad decision | Profit | The most dangerous category because luck can reinforce poor behavior |
| Bad decision | Loss | A visible process error with an immediate financial cost |
A “bad win” is often the most valuable material in the review. Imagine entering without a valid signal, increasing the position, and making money because the market reverses sharply. Judged only by P&L, the trade looks excellent. Judged by process, it is a warning: repeat the behavior often enough and a future loss may exceed the planned risk.
The weekly review begins before the first trade
After the event, it is easy to build a convincing story. The later price move is already visible, an important level looks obvious, and the entry can be explained more intelligently than it was actually made. A sound review therefore depends on records created at the time of the decision.
Before entering, or immediately afterward, record:
the market context and strategy name;
the specific signal and the conditions confirming it;
the intended entry price or zone, initial stop, and invalidation level;
planned risk in money, percentage terms, or R;
the position-management and exit scenarios;
a screenshot that does not reveal the later price path;
a short note about any state that may affect execution, such as haste, fatigue, or an urge to recover a previous loss.
R is the unit of initially planned risk. If the maximum planned loss was €100, then –1R is a €100 loss and +2R is a €200 profit before costs. R makes trades with different position sizes comparable, but only if the original risk is not rewritten after the result is known.
CME Group’s guidance on keeping a trade log recommends recording entry and exit points, targets, time, market levels, and indicators. Its post-trade guidance also treats P&L as secondary to understanding why and how the performance occurred. That sequence protects a review from convenient edits to memory.
Three layers that should not be mixed
1. Strategy process
This layer asks whether the trade belonged to the written plan at all. Was the setup valid? Was the market regime appropriate? Was there an event during which the rules prohibit trading? Were the thesis and invalidation point clear before risk was taken?
Process quality cannot be judged by a candle that appeared later. The question is whether the decision had sufficient, predefined justification at the time.
2. Execution
A valid setup does not guarantee a well-executed trade. This layer covers the actual entry, position size, stop placement, partial exits, management, and final exit. It captures late entries, chasing price, moving a stop farther away, or increasing size outside the plan.
If the strategy signal was valid but execution created an additional loss, changing the setup rules would solve the wrong problem. The execution habit needs attention.
3. Outcome
Only after process and execution have been assessed should the result be revealed: gross and net P&L, R result, commissions, slippage, and maximum adverse excursion if those data are collected. Outcomes are necessary for testing whether an edge persists over a longer period. They are not a reliable score for the quality of one decision.
The three layers make diagnosis more precise. A valid setup executed correctly for –1R is a strategy loss. A valid setup with a planned –1R that became –2.3R because the stop was widened is both a strategy loss and an execution error. An unexpected +1.5R gain after an entry without a signal is a positive outcome produced by a poor process.
A practical 45-minute weekly review
The review needs enough depth to produce a useful conclusion and enough restraint to be completed every week. The sequence below can be adjusted to the trader’s frequency.
Hide P&L and reconstruct the original context
Open the pre-trade screenshot and the original notes, but temporarily hide the monetary and R result. Check what was knowable at entry. If the platform cannot hide results easily, read the notes before opening the trade summary.
Classify every trade
Assign one of the four combinations: good or bad decision, profit or loss. “Good” does not mean perfect. It means that the trade met the predefined minimum standard and kept risk under control.
If the rules are too vague to support a classification, that is itself a finding. The next task is not to guess the grade but to make the trading-plan criteria more precise.
Score the components of the decision
A simple scale from 0 to 2 prevents the review from becoming an unstructured essay: 0 means the rule was broken, 1 means partially met or unclear, and 2 means fully met.
| Criterion | What to check |
| Context | Did the market regime and timing suit the strategy? |
| Setup | Were all mandatory signal conditions present? |
| Risk | Were size, stop, and total exposure within the plan? |
| Entry | Was the order executed in the intended zone without chasing? |
| Management | Was the position managed according to a predefined scenario? |
| Exit | Did the exit follow the plan or a documented rule? |
The total score is not a universal truth. It is a consistent instrument for comparing your own decisions. If a “good trade” means something different every week, the metric loses its value.
Separate strategy losses from the cost of mistakes
A strategy loss is a valid trade that ended negatively within the planned risk. The cost of a mistake comes from a documented deviation: excessive size, delayed exit, an unauthorized re-entry, a trade outside the plan, or another rule violation.
Calculate the cost in R only when the planned alternative had been defined before the trade. Otherwise, the number creates false precision. With hindsight, it is always possible to invent the perfect entry and exit. Mark such cases as qualitative process errors without fabricating a counterfactual amount.
Summarize a small set of useful metrics
A weekly dashboard does not need dozens of indicators. Keep the measures that can change behavior:
rule-adherence rate — the share of trades completed without a material deviation;
valid-setup rate — the share of entries that genuinely met the strategy criteria;
average planned risk and its dispersion in R or as a percentage of the account;
number of execution errors and their documented cost in R;
number of unplanned trades and valid signals that were skipped;
net result in R and money;
expectancy in R by setup over a longer rolling sample.
Expectancy can be expressed as the average R result per trade. A five-day result, however, should not be declared the new truth about a strategy. Weekly data are a signal for investigation; changing the strategy requires a sufficiently large and comparable sample, together with market-regime context.
Risk must be defined before the trade. If that part of the process still needs structure, see the related article on combining risk per trade with a daily loss limit.
Find one recurring cause
At the end of the week, look for a repeated mechanism rather than simply the worst trade. Perhaps two late entries followed a missed first signal; three stops were widened in the afternoon; or unplanned trades appeared immediately after the first loss of the day.
That connection points to a cause that can be influenced. “Be more disciplined” specifies nothing. “After missing a signal, do not chase; permit another entry only after a new complete setup” is testable.
Choose one experiment for next week
Do not try to repair five problems in one week. Choose one behavior, its trigger, and a measure. Examples include completing three mandatory fields before every entry, taking a 20-minute break after two losses, trading only A-grade setups, or using a prepared order template.
At the next review, check whether the experiment was followed and whether it reduced the targeted error. Profit may not improve immediately. The purpose of the process is to build controlled, repeatable behavior.
Example: a profitable week with a weak process
Consider a hypothetical five-trade week ending at +2.2R. The first impression is positive. One trade, however, generated +4R after an entry without the complete signal and with 1.5 times the allowed risk. The other four trades lost –1.8R in total. Two were valid strategy losses; the other two included a widened stop and a re-entry without a new setup.
If the review stops at +2.2R, the trader may gain confidence in the very behavior that increases the risk of failure. A process review reaches a different conclusion:
the financial result was positive;
decision quality was poor in three of the five trades;
the main risk was not setup quality but deviation from entry and risk rules;
the next experiment is to block the order unless every mandatory setup criterion has been checked.
The opposite example is a –1.5R week in which every trade followed the plan, sizing was consistent, and no limit was breached. It is not an enjoyable week, but it may be a good one in terms of decision quality. The next step is to check whether the losing run is within the strategy’s historical dispersion and whether the market regime has changed, rather than rewriting the rules impulsively.
A weekly template that can actually be completed
One page is enough for the final review:
Weekly result: net P&L and R, number of trades, maximum daily loss.
Process quality: rule-adherence rate, average decision score, number of bad wins and good losses.
Execution: main deviations, slippage, sizing consistency, and the documented cost of errors.
Strategy data: results by setup and market regime, compared with a longer rolling sample.
Recurring pattern: one observation supported by specific trades.
Next week’s experiment: one action, one trigger, and one measure.
If the review regularly takes several hours, it probably contains too much free-form text and too few predefined fields. If it takes five minutes and ends with “trade better,” the structure is too weak. A good format makes weeks comparable without discarding trade context.
What to evaluate over several weeks
One week is useful for correcting behavior, but strategy conclusions require a longer period. Every four to eight weeks, review:
whether rule adherence is improving;
which setups generate positive or negative average R;
whether errors cluster at a particular time of day, in a market regime, or after a specific event;
whether actual drawdowns and losing streaks remain within the tested range;
whether commissions and slippage materially reduce the theoretical edge;
whether position size is affecting decision quality.
P&L becomes more informative here because a larger sample is better suited to evaluating strategy economics. Even then, outcomes must be read together with risk and execution. Two strategies with identical profit are not equivalent if one has a deeper drawdown, unstable sizing, and frequent rule violations.
Common weekly-review mistakes
The first mistake is to begin with the equity curve and then invent a story that fits it. That strengthens outcome bias. The second is to rewrite a strategy after a few losses, confusing normal dispersion with the disappearance of an edge. The third is to build a scoring system so complicated that it cannot be used consistently.
Self-punishment is not analysis either. If the review turns into an emotional verdict — “I am a bad trader” — it does not identify a different action. Evaluate observable behavior: the entry was outside the zone, risk exceeded the limit, the stop was changed without a rule, or the journal was not completed.
Finally, do not remove P&L from the review. The goal is not to pretend that money does not matter. The goal is to put it in the right place: after process and execution have been assessed, but before making a long-term conclusion about the strategy.
Conclusion: evaluate what can be repeated first
A trader cannot control the next market move, but can control setup selection, position size, risk limits, order execution, and the response to a loss. Those are the elements a weekly review should measure first.
Profit matters, but on its own it is a noisy teacher. A better review recognizes a good loss, a dangerous win, and a specific error that can be reduced next week. When every review produces one measurable experiment, the journal stops being a trade archive and becomes a decision-quality system.
This material is educational and does not constitute individualized investment advice. Trading, especially with leverage, can produce losses beyond those initially expected; understand the instrument’s terms and assess your own risk tolerance before trading.
