Win rate is a vanity number, expectancy after costs is the real one
Posted: Sat Oct 31, 2026 8:32 am
Two strategies. A wins 70 percent of the time with 3 pip winners and 8 pip losers. B wins 40 percent with 9 pip winners and 4 pip losers. Most forum comparisons would pick A on win rate alone.
Gross expectancy for A: 0.7 x 3 minus 0.3 x 8 = 2.1 minus 2.4 = minus 0.3 pip per trade. It is already a loser. Add 1 pip of cost and it is minus 1.3.
Gross expectancy for B: 0.4 x 9 minus 0.6 x 4 = 3.6 minus 2.4 = plus 1.2 pips. After the same 1 pip cost it is plus 0.2 pip.
So the low win rate system is profitable and the high one is not. Worse, B has only 0.2 pip of margin, so a small rise in costs erases it. That is the thing to watch: cost as a share of gross expectancy. When the cost is 80 percent of the gross edge, any change in spread, slippage or commission decides whether the system survives.
Multiply the net figure by the number of trades per day to see what the strategy is really paying for the effort. 0.2 pip over 30 trades is 6 pips a day, before variance.
What is your net expectancy per trade, and what share of the gross edge goes to costs?
Gross expectancy for A: 0.7 x 3 minus 0.3 x 8 = 2.1 minus 2.4 = minus 0.3 pip per trade. It is already a loser. Add 1 pip of cost and it is minus 1.3.
Gross expectancy for B: 0.4 x 9 minus 0.6 x 4 = 3.6 minus 2.4 = plus 1.2 pips. After the same 1 pip cost it is plus 0.2 pip.
So the low win rate system is profitable and the high one is not. Worse, B has only 0.2 pip of margin, so a small rise in costs erases it. That is the thing to watch: cost as a share of gross expectancy. When the cost is 80 percent of the gross edge, any change in spread, slippage or commission decides whether the system survives.
Multiply the net figure by the number of trades per day to see what the strategy is really paying for the effort. 0.2 pip over 30 trades is 6 pips a day, before variance.
What is your net expectancy per trade, and what share of the gross edge goes to costs?