What to review first after a loss: separate trading results from the execution process

Author: Thomas | Trading Knowledge | October 7, 2026

A losing trade does not necessarily mean all the rules are wrong; a winning trade does not necessarily mean the order-placing process is worth repeating. Looking only at the final profit or loss makes it easy to mistake a random outcome for judgment ability. After a loss, the review should first reconstruct what was known and what was done at the time, and then discuss the next step.

Put results and process in two columns

The results column records actual fills, fees, realized profit and loss on closed positions, and the risk still open. The process column records the entry rationale, the originally defined invalidation condition, planned size, actual operations, and whether any deviation occurred. The two columns can explain each other, but a good result cannot erase a process problem.

For example, entering according to pre-set rules and exiting according to defined conditions while still taking a loss is a different record from placing an order on the spot without a triggered condition and then profiting by chance. This example only illustrates review classification; it is not the experience of any real account, nor does it prove that the former rule is effective over the long term.

First write down the information at the time of order placement; do not use later price action to add reasons

Keep the trading instrument, time and time zone, observation timeframe, price information visible at the time, and the basis for the judgment. If news or an economic calendar was used, record the time the information actually appeared; if chart conditions were used, save the screenshot or corresponding record from that time.

Seeing later highs and lows and then treating them as information already clear at the time of order placement will distort the review. Ex post observations can be written separately in the record, but they must be kept separate from the original basis. Mark missing information as unknown; do not fabricate the judgment at the time just to make the story complete.

After a loss, check whether the original rules were changed

The review can check in sequence: whether the originally defined trigger condition was met; whether the actual size matched the plan; whether the exit or invalidation condition was temporarily loosened; whether new orders were added because of eagerness to recover the loss. These questions check behavior; they do not directly prescribe a trade size for anyone's next trade.

If a deviation occurred, write down the specific operation and reason, rather than summarizing it only as a “bad mindset.” For example, the original plan was to wait for confirmation, but an order was placed early out of fear of missing out; or after an anomaly appeared, the defined process was not followed for checking. Describable behavior is easier to verify later than a vague evaluation.

Do not set immediate recovery as the task for the next trade

The amount of a previous loss cannot by itself make new trading conditions more reliable. Temporarily increasing size, loosening restrictions, or frequently changing methods in order to make up for one loss may add new exposure and at the same time undermine the comparability of records before and after.

You can write in advance which anomalies require pausing new trades and checking, for example unclear order status, actual loss exceeding the planned boundary, or an obvious impulse to violate rules. Once triggered, follow the predetermined process instead of constantly rewriting the rules when results are unfavorable. Pausing new operations also does not mean existing positions have been handled; they need to be checked separately.

The log does not need to be long, but it must be reviewable

Each record should be able to answer at least four questions: why participate; what information was available at the time; what was actually done; which parts differed from the plan. Then attach fill time, instrument, size, fees, and related orders, so the text can correspond to account records.

For accounts using MT4 or EA, program behavior and manual operations should be recorded separately. Modifying parameters, restarting the program, manually closing orders, or canceling pending orders may all affect later results. Do not attribute all differences to the strategy, and do not replace complete records with just one profitable screenshot.

Observe repeated deviations; do not change the system based on a single result

A single trade provides one case, which is not enough to prove that a certain behavior will necessarily profit or lose. After accumulating records under the same criteria, observe whether deviations occur repeatedly, in which periods they cluster, and whether they are related to fill conditions or specific rules.

If you decide to modify the rules, keep the modification date, basis, and planned checking method. Subsequent records should use a clearly defined new version; do not mix old and new methods when calculating performance, and do not repeatedly adjust the statistical interval to keep only the segments that look favorable. The recording method itself cannot solve all risks, but it can make unknown matters and changes clearer.

Daily review: leave one verifiable conclusion

Write down the one thing that most needs continued checking that day, along with the corresponding evidence. For example, “a certain operation's size was inconsistent with the plan; parameters and orders need to be checked” is more specific than “I will definitely do better tomorrow.” If information is still insufficient, record what materials are needed next, and do not rush to draw a conclusion about the entire strategy.

Leveraged trading may cause significant losses. This article is for knowledge sharing on trading records and behavioral review, and does not constitute advice to buy or sell, position sizing, or returns. Following processes and keeping logs also cannot guarantee future profits; historical results do not represent future performance.