Journaling and Metrics

The short answer

Most journals record entry, exit and profit. That is a bank statement with extra columns. It tells you what happened and nothing about why, which means it cannot tell you what to change.

A useful journal records the decision — what you saw, what you expected, what would prove you wrong — written before the outcome exists.

Fields that earn their place

Before entry (this is the important half):

• Setup name, from your plan
• Timeframe analysed
• Thesis, in one sentence
• Invalidation level, and the reason it is that level
• Planned size, and the risk in currency and percentage
• Target or exit logic
• Your state: rested, tired, up for the day, down for the day
• Time and session

After exit:

• Actual entry and exit
• Actual size
• Result in currency and in R (multiples of the risk taken)
• Adherence: did you follow each of the four core rules? Yes/no each
• Deviation type, if any: size, stop, setup, exit, unplanned entry
• One sentence: what would you repeat, and what would you not

The "before" block must be filled in before the trade is placed. This is the difference between a journal and a memoir. Reasoning written after the outcome is reasoning shaped by the outcome, every time, however honest you intend to be.

The metrics worth keeping

Four, and no more. Extra metrics mostly add noise and give you more ways to find a comforting number.

Adherence rate

Percentage of trades that followed every rule. This is the only metric with a usable weekly sample, because every trade contributes to it regardless of outcome.

Below about 90%, no other metric means anything, because the results were not produced by your method.

Expectancy

expectancy = (win rate × average win) − (loss rate × average loss)

Expressed in R, this is your average profit per trade in units of risk. An expectancy of +0.2R means each trade is worth, on average, one fifth of what you risk.

It needs a real sample. Over 20 trades it is noise. Over 100 it starts being informative. Compute it net of fees or it is measuring something you cannot collect.

Average win versus average loss

Watch this pair specifically for the disposition effect: if your average loser is bigger than your average winner while your win rate is around half, you are cutting winners and holding losers, whatever your plan says.

Maximum drawdown

The largest peak-to-trough decline. Track it because it is the number that determines whether you will still be following the method when it recovers. A method you abandon at the bottom has an effective expectancy of whatever you locked in.

R as the unit

Recording results in R rather than currency is a small change that removes a lot of distortion.

A $200 loss on a $10,000 account and a $400 loss on a $20,000 account are both −1R. Expressed in currency they look different and feel different; expressed in R they are identical, which is the truth. R also makes results comparable across instruments and across time as the account changes size.

Tagging mistakes

Free-text notes do not aggregate. A fixed set of tags does.

Something like: size-too-large, no-setup, moved-stop, early-exit, late-entry, revenge, fomo.

At the end of a quarter you can count them. "I over-sized eleven times, nine of which were within thirty minutes of a loss" is a finding you can act on with a specific constraint. "I need to be more disciplined" is not.

The uncomfortable part

A good journal will, with some regularity, tell you that your method does not work, or that you do not follow it, or both. That is what it is for.

The alternative is not avoiding the finding; it is finding out later, from a drawdown, with less information about why.