How can an 80% win rate still lose money?

Win rate is intuitive, but it says little about profitability unless you also know the average size of a win and the average size of a loss.

You can win 80 times and still lose money if the 20 losses are large enough.

For example, if eight of ten trades make +0.2% and two lose -1.0%, the win rate is 80%, yet the simple pre-cost total is already negative. Conversely, a 45% win rate can still have positive expectancy if the average win is large enough relative to the average loss.

A single summary number is not enough

Averages alone may still hide risk. Losses may cluster in one market regime, a few outsized winners may support most of the total return, or losing streaks may be longer than the operator can tolerate.

We do not use win rate as the final score

Win rate is useful as one descriptive metric, but it should be read alongside average win, average loss, net expectancy after costs, maximum drawdown, and the full P&L distribution. A strategy can manufacture a pretty win rate by taking tiny profits quickly while delaying large losses.