5 / 6 · 8 min
The journal: the only way to catch yourself red-handed
Memory rebuilds motives after the fact, and arguing with it is useless — you can only write the decision down before the outcome is known. The good news: your own habits show up on a sample twenty times smaller than testing an idea requires.
Ask someone why they entered a trade that turned a profit — they will name a reason. Ask about an identical one that lost — they will name a different reason, and both will sound convincing. This is not lying: memory genuinely rebuilds the motive to fit the known outcome. It happens without any act of will and cannot be felt from the inside.
What exactly gets rewritten
- THE MOTIVE. «I entered at the level» after a profit and «I rushed it» after a loss — about the very same trade.
- THE CONFIDENCE. After the outcome it feels as if you foresaw it, although at the moment of entry you were unsure.
- THE SELECTION. Rule violations are remembered selectively: the ones that ended in profit are recalled as «intuition», the ones that lost as «bad luck».
Hence the only requirement for a journal
Everything about the DECISION is written down BEFORE the outcome is known. Everything about the result comes after. A journal filled in during the evening from memory documents not your decisions but your reconstruction of them; such a journal always shows a sensible person who is occasionally unlucky.
What to record before entry and after exit
| Before entry (mandatory) | After exit |
|---|---|
| The instrument and why this one | Exit price and time |
| The entry condition — something already happened | Reason for exit: stop, target, by hand |
| Stop price and target price | Result in R |
| Size and the calculation it came from | What price did after the exit |
| Was this in the plan for the day | One sentence: did I follow the rule |
Why your habits show up fast and an edge does not
The «Testing an idea» course computes that distinguishing a 55% hit rate from 50% requires 1046 trades. A wrong conclusion is often drawn from this: since the sample is unattainable anyway, the journal is useless. But a journal is not looking for an edge — it is looking for YOUR systematic bias, and that is a different task altogether: the bias is large and constant, the edge is small and drowns in noise.
How many trades it takes to see it
| What we are looking for | Trades |
|---|---|
| Your exit bias of 1R per trade | 4 |
| Your exit bias of 0.5R | 16 |
| Your exit bias of 0.3R | 43 |
| Losers held twice as long as winners | 11 |
| An edge of 60% versus 50% wins | 258 |
| An edge of 55% versus 50% wins | 1046 |
A sixty-fold difference, and an explicable one. Your habit of cutting profit at 0.8% instead of 2% acts identically in every trade — that is not randomness but a shift. An edge, by contrast, has to be fished out of noise several times larger than the edge itself. So the first thirty journal entries will tell you far more about you than about your system — and that is the most valuable thing thirty trades can buy.
What a month of records shows
Twenty-six trades in a month. Compute four quantities. Average win +1.1R, average loss −1.4R — with a planned target and stop of 1R that is already a diagnosis: the exits are asymmetric, and expectancy has slipped 0.25R per trade without a single error in the entries. Average time in a winning trade 40 minutes, in a losing one 2 hours 10 minutes — the same thing seen from another side. Of the twenty-six trades, six were not in the day's plan; of those six, five lost. None of these numbers requires statistics: they are not about whether the system works but about whether you are executing it.
Keeping a journal of results instead of decisions
The commonest form of journal is a table of «entry, exit, profit». It is useless precisely because it contains only what the exchange's trade history already holds. The value of a journal is in the columns the exchange does not have: what you intended to do, how you justified it, whether this entry was in the plan. Without them the journal reports the result and stays silent about the cause — and the cause is what needs fixing.
Review once a week, not after every trade
Looking at the journal right after a trade is pointless: you are in exactly the state it is supposed to protect you from. The review is done on a calm day, across the whole week at once, and answers three questions: how many trades were off-plan, where the exit diverged from what was written in advance, and which rule was broken most often. The last one is the candidate for redesign — of the rule, not of your character.
If keeping a journal by hand does not stick, start with paper trading: there the time, price and size of every trade record themselves, and only the reason and the plan are left to add.
Open paper tradingThirty entries and four numbers
Set up a seven-column table on the model above and fill it in BEFORE entry. After thirty trades compute four quantities: average win, average loss, average time in a winning trade and in a losing one. Draw no conclusions about the system — there is nothing yet to draw them from. Draw a conclusion about yourself: which of the two pairs is asymmetric, and in which direction.
Why must the reason for an entry be recorded before the entry rather than in the evening?
Because once the outcome is known memory rebuilds the motive: a good reason attaches itself to the winning trade and a bad one to the loser, although the decision was identical. An evening entry documents a reconstruction, not a decision. The only defence is to write it down before the outcome exists.
Testing an idea needs hundreds of trades. What good is a thirty-entry journal then?
It is good because the journal is not hunting for an edge but for your systematic bias, which is far larger. An exit bias of 0.5R shows up in 16 trades, a two-fold difference in holding time in 11, whereas an edge of 55% against 50% needs 1046. Thirty entries will say nothing about the system and a great deal about how you execute it.