← Guides

Keeping a trading journal that actually helps

Most trading journals die after three weeks. Not for lack of willpower: because too much goes in, too late, and nobody ever reads them again.

Separate what is calculated from what is judged

A journal holds two very different things. The facts: symbol, direction, prices, size, result, time. And the judgement: why you entered, why you exited, whether you followed your plan.

The facts should never be copied by hand. It is slow, error-prone, and it is what makes people give up on a journal. They can be pulled automatically from your platform. Your time should go to the judgement, which nobody can record for you.

What is really worth recording

The planned risk, before entering. Without it, there is no result in R.

The reason for entry, in one sentence. If you cannot write it, it may have been an urge rather than a plan.

The reason for exit: target, stop, manual exit, and why.

The execution quality: well executed, fair or poorly executed, regardless of the result. A losing trade can be perfectly executed, a winning trade a mistake that happened to work.

What is useless

Screenshots of every trade that nobody reviews. Long emotional paragraphs written in the heat of the moment. Columns filled out of habit and never used. A good journal is short: what will not be reread does not need to be written.

Reviewing it

A journal is only useful when reviewed. Once a week, look at your trades from three angles: by execution quality, to see what straying from your plan costs you; by pair, to spot the ones where you keep losing; by day of the week and by hour, to see when you trade badly.

A simple rule: every review should produce a decision, however small. Drop a pair, reduce size on Fridays, stop entering before a release, which the economic calendar lets you spot in advance.

In My Trading Book

MT5 trades come in on their own, with their prices, commissions and swaps. All that is left to enter is the judgement: risk, setup, reasons for entry and exit, execution quality, notes. Statistics can then be sliced by account, pair, quality, day and month, so a review takes ten minutes rather than an evening.