Keeping a trade journal that's worth rereading
Recording what you did is nearly useless. Recording what you expected, before you knew, is the part that teaches you something.
Most journals are a list of trades with prices and outcomes. That information is already in your broker statement, and rereading it teaches nothing you did not already know.
What makes a journal worth keeping is recording your reasoning before the outcome exists. Once you know how a trade turned out, you cannot reconstruct what you actually thought beforehand — memory rewrites it toward whatever happened.
What to write down before
- The read. What you think is happening, in one sentence.
- What would make it wrong. A specific price or condition, stated as a number.
- Why now. What changed to make this the moment.
- A screenshot. The chart as it looked, not as it looks afterwards.
- Size, and how you got it. The arithmetic, not the feeling.
Two minutes. If the position cannot wait two minutes, that is itself information worth having a record of.
What to write down after
The outcome, and separately — this is the important part — whether you did what you said you would.
Score the decision and the outcome separately. A trade can be well executed and lose. A trade can be a rule violation and win. Recording only the outcome trains you toward the second, because it paid.
Four boxes: followed the plan and won, followed and lost, broke it and won, broke it and lost. The one that matters most is broke it and won, because it feels like success and is the most expensive habit you can build.
What the journal is actually for
Not motivation, and not record-keeping. It exists to answer questions you cannot answer from memory:
- Are the trades I skip better than the ones I take?
- Does my execution deteriorate at a particular time of day?
- Do I widen stops after a loss?
- Are my worst outcomes concentrated in one instrument, or one condition?
Each of those needs written evidence from before you knew the answer.
On sample size
Ten trades tell you almost nothing. The variation in short runs of anything is large enough to produce convincing streaks in both directions, and concluding a method works or fails from a handful is the same error as curve fitting at smaller scale.
Review on a schedule — monthly, or every fifty trades — rather than after any particular result. Reviewing after a bad day means reviewing while you are the least able to read the evidence fairly.