Count how much each win brings in
70 wins out of 100 trades means a 70% win rate. To find the result in money, you also need the size of the wins and losses. In this fictional sample, a winning trade makes an average of 60 USD before costs, while a losing trade costs an average of 120 USD. Across all 100 trades, the total entry and exit costs are 8 USD per trade.
The 70 wins bring in 4 200 USD, while the 30 losses subtract 3 600 USD. That leaves 600 USD before costs. After subtracting 800 USD in costs across all trades, the net result turns negative: −200 USD. These amounts describe only the sample given in this learning exercise.
| Item | Calculation | Result |
|---|---|---|
| Gross profit from wins | 70 × 60 | +4 200 USD |
| Total gross losses | 30 × 120 | −3 600 USD |
| Costs for all trades | 100 × 8 | −800 USD |
| Net result | 4 200 − 3 600 − 800 | −200 USD |
The average result is minus two dollars per trade
Dividing −200 USD by 100 trades gives −2 USD per trade. That is the average net result for this sample. You can write the same calculation using the proportions of wins and losses: 0.70 × 60 − 0.30 × 120 − 8 = −2.
If a model of future results assumed exactly the same outcome distribution and costs, its calculated expectancy would also be −2 USD per trade. But the observed sample does not, on its own, prove that those assumptions will hold. A sample average and a forecast of future results are different claims.
Where a positive result begins in this example
With a fixed gross win of 60 USD, a gross loss of 120 USD, and costs of 8 USD, the mathematical break-even point is approximately a 71.11% win rate. It comes from the equation p × 60 − (1 − p) × 120 − 8 = 0. Here, p represents the probability of a win under the model's specific assumptions.
A complete sample of 100 trades cannot contain 71.11 wins. 71 wins and 29 losses would produce −20 USD, while 72 wins and 28 losses would produce +160 USD. This is not a target for every strategy: changing the average profit, loss, or costs also changes the break-even point.
Count costs exactly once
In this example, the 8 USD covers the whole trade from entry to exit. We deduct it from both winning and losing trades. If your data already gives a net result, deducting the same costs again would be a mistake.
The same applies to the spread between the bid and ask prices. If its effect is already reflected in the actual execution prices and the result calculated from them, a separate deduction would distort the calculation. First, establish what each column in your report means. Our fixed cost of 8 USD is a simplifying assumption.
The average does not show the order of trades
You could arrange the same sample so that all 30 losses come first. The average loss after costs would be 128 USD, so that run would produce a decline of 3 840 USD. Reordering the trades would leave the total result of −200 USD unchanged, but the account would take a different path to get there.
That is why a performance report needs the number of trades, average amounts, cost accounting, and the sequence of losses alongside the win rate. These details let you test specific assumptions. A single percentage without them leaves too many questions unanswered.
Also state how you chose the period under review. If you omit some trades because you dislike their results, the calculated average no longer describes the full sample you originally selected. It should be possible to repeat the calculation from the trade list and arrive at the same total.