How to actually review a trading journal

Almost everyone who keeps a trading journal keeps it the same way: diligently for six weeks, then sporadically, then not at all. The usual diagnosis is discipline. It is almost never discipline. It is that the journal was set up as a place to write things down, and never set up as a place to read things back.

A record you never re-read is a diary. What makes it a journal is the loop: you write, you read, something changes, you write again. Below is the smallest version of that loop that still works, and it fits in about twenty minutes a week.

Why the review usually fails

Three failure modes account for most abandoned journals, and they compound.

1. The review has no fixed time. "I will look at it this weekend" loses to everything else on the weekend. A review that is not on the calendar does not happen.
2. The review has no fixed question. Opening a journal and scrolling is not reviewing. Without a question you are looking for something interesting, and interesting is not the same as important.
3. The review produces a list. Twelve observations feel productive and change nothing, because you cannot hold twelve changes in your head during a live session. One change is a change. Twelve is a wish.

The twenty-minute loop

Use the same slot each week: the end of the last session, or the morning before the first. Consistency matters more than which slot you choose.

1. Read the week without the P&L (5 minutes)

Hide the numbers if your tool lets you. Read only what you wrote at the time: the plan, the reason for entry, what you said while the trade was open. You are checking one thing: whether the decision was reasonable given what you knew at that moment. A well-reasoned loser is a different object from a lucky winner, and the P&L column makes it almost impossible to tell them apart.

2. Sort by one dimension, not all of them (5 minutes)

Pick a single cut and look only at that. By setup. By hour of day. By day of week. By whether it was your A-setup or something you talked yourself into. One dimension per week, rotating, beats every dimension at once. The latter produces a wall of statistics you skim and forget.

3. Write one sentence (5 minutes)

Write one sentence describing the single most expensive repeated behaviour of the week. Focus on the repeated behaviour, not the worst trade. "I moved my stop three times on Wednesday and Thursday" is a sentence you can act on. "Bad week" is not.

4. Set one rule, with a number in it (5 minutes)

Turn that sentence into a rule you can actually be caught breaking. Rules without numbers are intentions.

• Weak: "Stop revenge trading." There is no threshold and no way to check it.
• Strong: "After two consecutive losses, I am done for the session." It is binary, checkable, and obvious when broken.
• Weak: "Be more patient on entries."
• Strong: "No entry in the first fifteen minutes unless it is on the pre-market list."

What to check next week

The following week the loop gains one step at the front: did last week's rule hold? That single question is what makes the whole thing compound. It turns the journal from a record of what happened into a record of what you are changing, which is the only version that pays for the time it takes.

A journal you argue with is worth ten you agree with.

One more thing worth saying plainly: the point of this is not to find a hidden edge in your data. Most traders do not have a subtle statistical leak. They have an obvious behavioural one that they already half know about, and the review exists to make it impossible to keep not-quite-knowing.