How often should you run a trade review?

Two appointments with different purposes work best. Weekly you check completeness and rule compliance: whether every trade is recorded, whether each has a stop documented in advance, and which trades deviated from your rules. That slot is short and concerns process. Monthly you look at numbers, because shorter intervals mostly show noise and tempt you into premature changes. A daily review is unnecessary on most timeframes and tends to make you react to individual outcomes rather than to patterns. More important than the exact frequency is that both slots sit in your calendar rather than depending on whether time happens to be left over.

What do I do with trades that have missing fields?

Leave them empty and mark the trade as incomplete. The most common review mistake is filling missing entries in afterwards, because the sheet then looks complete. This is particularly damaging for the planned stop: a value recorded after the exit is not the decision you made but your memory of it, and memory adjusts to the outcome. That makes the single most important analysis impossible, namely separating a weak system from poor execution. A journal with honestly flagged gaps is far more useful than one where the gaps are invisible. For metrics, simply exclude incomplete trades and track their share as a number in its own right.

How many trades before analysis is worthwhile?

For the numbers, roughly thirty to fifty trades under identical rules. Below that randomness dominates and you are mostly interpreting noise. Consistency matters as much as count: if you changed your entry signal or position sizing partway through, the count effectively restarts. The weekly review is exempt from this, because it analyses no metrics but checks completeness and rule compliance. That check works from the very first trade and is the more valuable of the two early on. Recording cleanly and flagging deviations during your first weeks leaves you with a data set worth something after two months.

Why examine winning trades critically?

Because a gain says nothing about whether the decision was good. A trade that broke your rules and still closed in profit is the most dangerous case in the whole journal: it rewards behaviour that does not hold up across repetitions and sends no bill to correct you. This is exactly why you score every trade with two questions instead of one, namely rule compliant yes or no and outcome positive or negative. What matters is comparing the two compliant boxes against the two deviating ones. If results drop sharply on the deviations, your problem is execution. If they do not, your rules themselves deserve a closer look.

Trade Review: A Weekly and Monthly Routine

Q: How many changes should follow a review?

One per period, and it belongs in writing. The reason is purely about measurement: change your entry signal, position sizing and timeframe together after a weak month and you cannot attribute next month's result to any of them. You end up collecting the opening phase of several systems rather than the result of one. Note three things for every change: what you are changing, from which date, and how you will tell after four weeks whether it helped. That third entry is almost always missing and is the decisive one, because without it any later assessment comes down to a feeling again.

A trade review is a fixed appointment with a fixed sequence, not an occasional scroll through recent trades. Weekly you check completeness and rule compliance; monthly you look at the numbers. That separation is the entire trick: combine them and you always end up at the outcome question and never answer the more important one, namely whether you followed your own rules.

Why do reviews fail for most traders?

Three reasons, and none of them is laziness.

First, there is no appointment. A review that happens when there is time never happens, because in trading something more urgent is always running.

Second, there is no sequence. Without fixed questions the review becomes a look at the equity curve, which answers nothing. It only produces a feeling.

Third, process and outcome are not separated. A winning trade feels right even when it broke every rule you set. As long as you only look at outcomes, that is exactly the behaviour you train.

What belongs in the weekly review?

Four questions, no more. The slot should be short enough that you keep it even in a busy week.

Are all of this week's trades recorded in full? Including the uncomfortable ones. Missing trades do not distribute randomly, they cluster around poor results.
Are there positions with no documented stop? If so, the stop was not fixed before entry, and that trade does not count towards your risk analysis.
How large was the gap between intended and actual entry? A number, not a feeling.
Which trades did not fit your rules, regardless of outcome?

The fourth question is the most important and the most frequently skipped, because rule breaking trades that worked out do not feel like a problem. Flag them consistently anyway.

What belongs in the monthly review?

This one is about numbers, and about few of them. Look at the distribution rather than the total.

Three analyses are enough to start: results expressed in multiples of risk rather than in currency, profit factor across the period, and how many trades account for most of the result. If two trades out of forty carry everything, that is not a flaw, but you should know it, because the next forty without such an outlier will probably look different. How to calculate profit factor cleanly is covered in calculating profit factor, and the remaining metrics in the overview of the metrics that matter most.

Frequency matters. Monthly is enough, because shorter intervals mostly show noise and prompt premature changes.

Which questions do you ask per trade?

Three blocks work well for individual trades. Answer them briefly, in keywords rather than essays.

Setup. Which signal was present, and was it complete? Was the instrument on your list beforehand? Did the market regime suit this setup?

Execution. Did the entry land where it was planned? Was the stop fixed before the order? Did position size follow the stop distance? Was the exit decided by rule or by feel?

State. How were you at the moment of entry, in one word? Was there outside pressure, time constraints, a preceding losing streak?

The third block looks soft and often explains a surprising share of the deviations. Keep it deliberately short, or it turns into a diary and you stop filling it in after two weeks.

How do you judge a trade independently of its outcome?

With two questions instead of one. Every trade then falls into one of four boxes:

Rule compliant and won. The normal case you want to repeat.
Rule compliant and lost. Also fine. Losses within plan are part of the system and no reason to change anything.
Rule breaking and lost. Unpleasant but instructive, because the bill arrives immediately.
Rule breaking and won. The most dangerous box, because it rewards the wrong behaviour and sends no bill.

What matters is not the distribution itself but comparing the first two boxes against the last two. If your rule compliant trades collectively produce a clearly different result from the deviations, your problem is execution rather than the system. Only when the compliant trades fail to hold up across enough cases does the strategy become the candidate for change.

What about a quarterly review?

Alongside weekly and monthly, a third and rarer slot pays off. It answers questions that a single month cannot, simply because the data is not there yet.

Three analyses belong in it. First, comparison by setup: which entry type carries your result, and which consumes it? Very often one setup is financing another's losses without that showing up in the headline number.

Second, comparison by market regime. Almost every strategy has a phase where it is structurally weak. Knowing that phase is more valuable than trying to optimise it away, because you can then cut size during it instead of changing the rules.

Third, development over time. Track your metrics on a rolling basis across the last fifty trades rather than only since inception. A falling rolling value under unchanged rules is one of the earliest warning signs, and the lifetime number hides it behind older strong phases.

This slot may run longer but should still end with exactly one derived change.

What do you do with the findings?

Exactly one change per period, and it gets written down.

The reason is measurement. Change your entry signal, your position sizing and your timeframe all at once after a weak month, and next month you will not know which of them mattered. You end up collecting the opening phase of several systems rather than the result of one.

What works is a short note with three entries: what you are changing, from which date, and how you will tell in four weeks whether it helped. That third entry is almost always missing and is the decisive one.

Which mistakes happen in the review itself?

The most common is reviewing by feel rather than by list. It feels productive and produces no decision.

The second is analysing samples that are too small. Ten trades show randomness. Meaning starts at roughly thirty to fifty trades under the same rules.

The third is focusing on the losing trades. They attract attention, but the more expensive insights often sit in the winning trades that broke your rules.

The fourth is filling in missing fields during the review. A stop recorded after the fact is not your decision but your memory of it. If the field is empty, leave it empty and mark the trade as incomplete.

How long should it take?

The weekly review should be done in fifteen to thirty minutes once capture is running. If it takes considerably longer, the cause is almost always that data has to be assembled during the slot itself. That is a capture problem rather than a review problem, as covered in trading journal spreadsheet.

The monthly review may take longer but also needs a ceiling, otherwise it turns into a rumination session. A structured trading journal supplies the fields for both slots.

Conclusion

A working review needs three things: a fixed appointment, a fixed list of questions, and separation between process and outcome. Weekly you check completeness, documented stops, execution quality and rule compliance. Monthly you look at a few numbers and derive exactly one change, with a criterion fixed in advance for recognising whether it worked.

Disclaimer: this article is for informational purposes only and does not constitute investment advice. Trading securities and crypto assets carries the risk of loss, up to and including total loss of capital.

Trade Review: How to Properly Analyze Past Trades