Spain
Mexico
Brazil
Poland
Taiwan
Philippines
South Korea
The essentials
- A 60% win rate can produce a negative average result.
- Use realized average wins and losses, with costs counted once.
- Positive historical expectancy does not establish a durable future edge.
Count money as well as wins
A win rate answers how often a defined trade was profitable. It does not say how much each winner earned or each loser lost. Six small gains can be outweighed by four large losses, so an attractive percentage can coexist with a shrinking account.
For a simplified two-outcome model, net expectancy per trade is p × W − (1 − p) × L − C. Here p is win probability, W is average gross profit, L is the positive magnitude of average gross loss, and C is average round-trip cost. Historical inputs produce an estimate.
Work through a 60% example
Assume, hypothetically, that 60% of trades gain $40 before costs, 40% lose $80, and every completed trade costs $2. The calculation is 0.60 × $40 − 0.40 × $80 − $2 = −$10 per trade. Winning more often has not offset the larger average loss.
A constructed batch of 100 trades with exactly 60 winners and 40 losers earns $2,400, loses $3,200 and pays $200, leaving −$1,000. The table splits those contributions. It illustrates the arithmetic; a future batch will not necessarily contain the same proportions or payoff sizes.
| Component | Per-trade contribution | 100-trade total |
|---|---|---|
| Gross wins | 0.60 × USD 40 = +USD 24 | +USD 2,400 |
| Gross losses | 0.40 × USD 80 = −USD 32 | −USD 3,200 |
| Round-trip costs | −USD 2 | −USD 200 |
| Net result | −USD 10 | −USD 1,000 |
Compare the payoff, not just the percentage
Now use a second hypothetical set: 40% winners averaging $100, 60% losers averaging $40, and $2 average cost. Its expectancy is $40 − $24 − $2 = +$14. A lower win rate can therefore accompany a better average outcome when the payoff distribution differs.
In this two-outcome model the break-even win probability is (L + C) ÷ (W + L). It is 68.33% in the first example and 30% in the second. These thresholds assume the stated averages hold. Setting a distant profit target does not make that target an achieved average win.
Reconcile the calculation to actual records
Use execution records to rebuild each completed position: entry and exit quantities, actual fill prices, commissions, and any financing charges. Decide how partial exits and additions form one trade before calculating the win rate. Otherwise splitting one profitable position into many exits can inflate the count of wins.
Keep cost treatment consistent. Realized profit based on actual fills already reflects the prices paid, including execution effects. Do not subtract an estimated spread or slippage again from that same profit. If results are already net of commissions, do not deduct those commissions a second time.
Ask how stable the sample is
Record the sample dates, trade count, position sizes, largest loss and average holding time alongside expectancy. A handful of unusually large winners can dominate a positive mean. Inspect results with and without those outliers as a sensitivity check, while retaining every trade in the official record.
A backtest selected after trying many rule combinations can look stronger than an untouched sample. Compare the same rules on separate periods and distinguish simulated executions from live fills. Positive measured expectancy describes that dataset; uncertainty, changing conditions and unobserved losses prevent it from guaranteeing an actual future edge.
Official sources
An explanation of financial mechanics based on official sources. Hypothetical calculations are not actual trading results or forecasts.
