Return versus a benchmark — because if the index rose 25% and you made 12%, you made money and lost to the passive alternative. Maximum drawdown from the peak — the metric that best predicts whether you can keep following the method. Expectancy per unit of risk. And the share of trades executed to plan, which measures discipline rather than outcome.

The last one is the only metric fully under your control. A single trade's outcome is noisy and partly luck; execution quality is not. A trader who executes 90% of trades to plan will improve; one chasing a monthly return figure will not.

A month is noise. A quarter is a hint. A year starts to be information. Checking returns daily generates a hundred unnecessary decisions a year and reactions to movements that mean nothing. Set a fixed review date — monthly for a long-term portfolio, weekly for swing — and ignore it in between.

And above all: measure gross, costs and taxes separately, then net. A trader who looks only at the final line on the screen doesn't know how much of the return disappears into costs, and therefore can't improve it.