You closed out the month up four percent. You feel good. You should feel good about the dollars — but you should not feel good about your edge, because you have no idea where the dollars came from.
Your PnL is an outcome metric. It tells you what happened. It does not tell you why it happened, and it does not tell you whether the same process repeated a hundred times would put you up or down.
That is the gap a real journal closes.
Three things a PnL hides
A profitable month can come from any of these — and they imply very different things about whether you can keep doing it:
- You called the direction right. Your thesis was correct, the market moved your way, you sized into it, and you took profit when you said you would. This is the only one that compounds.
- You got lucky on timing. Your thesis was directionally fine, but you would have been stopped out twice if the same idea had landed on a different week. The dollars are real; the skill signal is noise.
- You got lucky on the wrong reason. The trade went your way for a catalyst you never considered. You credit yourself with insight you did not have, and you size the next one bigger.
A column of dollar amounts cannot distinguish these. Only a record of what you thought was going to happen, and why, before you put the trade on, can.
What a thesis-first journal captures
The trade journals retail traders default to are spreadsheets of fills — entry, exit, size, PnL, maybe a one-line "comment" tacked on after. Those are useful for tax season. They are nearly useless for self-knowledge, because the thesis — the actual bet you were making about the world — is missing or written after the fact.
A thesis-first journal flips the order. Before the trade goes on, you log:
- What you expect to happen. The catalyst, the timeframe, the level you think matters.
- Why you think so. The reasoning, in your own words. Not a chart annotation — a sentence.
- What would prove you wrong. The state of the world that would invalidate this thesis cleanly. Not a stop-loss; an idea-loss.
When the position closes — or the catalyst window passes — you grade it. Not against the dollars. Against the thesis. Did the thing you said would happen, happen?
Why this is the version that compounds
Once you have a hundred logged hunches, the question stops being "did I make money." It becomes:
- Are my macro calls better than my single-name picks?
- Am I better at one-week catalysts than three-month theses?
- When I sell volatility because I think a name is too quiet, am I right more than half the time?
Those answers are invisible in a PnL log, and they are the only ones that tell you where you actually have edge — and where you are funding the parts that work with the parts that do not.
You don't need a hundred logged hunches to start getting signal. You need ten honest ones.