Skip to payoff research
Risk Reward Ratio CalculatorLevels / Quantities / Net payoff

Trading expectancy after costs

A worked guide to costs, weighted exits and completed outcomes.

Original edition: 9 October 2026. Expanded edition: 13 October 2026.

Price-level arithmetic and user-entered samples. A target ratio is neither a measured edge nor a guarantee of the stop fill.

Expected outcome, step by step

The basic expectancy formula is p × average win - (1 - p) × average loss, where p is the winning share and the loss is entered as a positive magnitude. Use gains and losses in the same units and on the same cost basis. A theoretical calculation from entered prices is useful for planning; a measured calculation needs the actual outcome ledger.

Start with a clearly defined sample

Choose a continuous period and a fixed strategy version. Include every completed position under the declared rules, and retain the statuses of unfilled, cancelled and still-open instructions. State whether a flat trade counts as a non-winner, has its own category, or is excluded under an existing rule. The formula must match that definition; a two-outcome model is not automatically appropriate for a ledger with several outcome classes.

If the cash risk differs between trades, a cash average and an R average answer different questions. Cash expectancy describes the average amount earned per trade at those historical sizes. R expectancy divides each result by its own original risk budget and describes the normalised distribution. Neither is a portfolio return without a sizing and capital calculation.

A hypothetical 40% strategy

Imagine 100 completed trades, all risking one unit before costs. Forty winners each make 2 units and 60 losers each lose 1. Gross gain is 80, gross loss is 60 and net gross P&L is 20. Expectancy is 0.40 × 2 - 0.60 × 1 = 0.20 per trade. The win rate is below half, yet the average outcome is positive because winners are larger.

Now add an assumed round-trip cost of 0.15 per trade. Net winners make 1.85 and net losers lose 1.15. After-cost expectancy is 0.40 × 1.85 - 0.60 × 1.15 = 0.05. Total net P&L across the hypothetical 100 trades is 5. A modest change in execution or costs could consume that remaining margin.

The example is synthetic arithmetic, not a performance record. The SEC fees bulletin explains why expenses reduce returns. A venue-specific cost estimate should come from the applicable fee schedule and execution record.

Derive the after-cost break-even rate

When there are two fixed outcomes, set expectancy to zero and solve for p: p = net loss / (net win + net loss). In the example, 1.15 / (1.85 + 1.15) = 38.33%. The clean 2:1 break-even rate is 33.33%, so costs move the required win rate upward. This calculation assumes the same outcome sizes on every trade.

Sensitivity example: 2-unit win, 1-unit loss
  1. No round-trip cost: net win 2.00, net loss 1.00, break-even 33.33%.
  2. Cost 0.15: net win 1.85, net loss 1.15, break-even 38.33%.
  3. Cost 0.30: net win 1.70, net loss 1.30, break-even 43.33%.

The trade's entered target and stop are unchanged; the cost assumption changes the required success rate.

Planned reward-to-risk and realised R differ

Suppose a long plan enters at 100, stops at 98 and targets 106. Its planned reward-to-risk is 3:1. With 10 units, the original cash risk is 20. If the entry fills at 100.10 and the exit at 105.50, gross P&L is 54. Subtract 2 of recorded costs and net P&L is 52, or 2.60R against the original 20 risk budget. The target ratio remains 3:1; this completed outcome is 2.60R.

A stop fill exactly at 98 gives a gross loss of 21 from that actual entry, before costs. A worse exit increases the loss further. The SEC stop-order bulletin explains why the stop trigger does not guarantee an execution price. Keep original planned risk visible when evaluating the result.

Partial exits need a weighted calculation

Consider a separate hypothetical position entered at 100 with a stop at 98. Half exits at 102 and half at 106. The weighted exit price is 104, so gross reward is 4 per original unit and the completed trade earns 2R before costs. Averaging the advertised 1R and 3R targets gives the same result here only because the quantities are equal. If 80% exits at 102 and 20% at 106, the weighted exit is 102.80 and gross outcome is 1.40R.

For a long position, calculate total exit proceeds minus total entry cost, then subtract actual charges. For a short, use the sale proceeds minus the cost of buying back, less charges. Aggregate the fills into the original position before calculating trade-level win rate. Counting both profit-taking fills as separate winners can change the denominator without adding another independent trade.

Variable outcomes: use the ledger mean

The realised-R analyser calculates the mean, outcome classes, R-based profit factor and closed-position cumulative-R decline from an entered series. Flat trades stay in its denominator. It does not verify the series or reconstruct funded account equity.

If gains, losses and costs vary, calculate each completed trade's net outcome first. The arithmetic mean of those outcomes is the sample expectancy. The win-rate formula gives the same mean when winning and non-winning groups are defined consistently. Do not replace actual realised winners with the original targets or actual losses with ideal stop distances.

Download the blank position-outcome ledger (CSV). Use one row per original position after reconciling its fills. Calculate net P&L as gross P&L minus costs, and realised R as net P&L divided by original cash risk. Retain open and unfilled statuses but do not include them as completed outcomes. The headers contain no measured results or automatic formulas.

For example, a strategy may have a 3:1 planned ratio but repeatedly exit winners early and incur losses beyond 1R. Using 3 and 1 in the formula would describe its plan rather than its results. Read the common mistakes guide alongside the actual fill history.

Average outcome does not describe the path

The same net average can occur with different drawdowns, losing streaks, exposures and concentration in a few trades. Keep the chronological equity record, costs and position overlap alongside expectancy. If the result depends on one exceptional winner, report that concentration rather than deleting the trade or assuming it will recur.

A sample estimate also changes when new trades arrive. Closely related trades may respond to the same market move, so the row count alone does not establish independent evidence. Compare continuous periods and disclose the number of model versions tested. An attractive sample average is a reason to investigate the process, not a forecast of the next outcome.

Use the calculator as a scenario worksheet

Enter the planned prices in the calculator, then add estimated cost and an explicitly assumed win rate. Stress a worse entry, a worse stop fill and a lower winning share. Record the assumptions with the result. To measure historical expectancy, use realised net outcomes from the complete ledger instead; the calculator does not supply a measured win rate.

Use one consistent unit, one declared denominator and the original risk budget. Those choices make expectancy reproducible and show exactly which assumption drives the conclusion.

Keep reading

Retain the cost lens and expose its branch assumption

The main same-cost lens still computes net winner as favourable distance less cost and net loser as loss distance plus cost. Its export now records the entered price units, optional frequency and educational classification. A cost exceeding the gross reward makes profitable two-branch break-even infeasible, not a prediction of real failure. The new asymmetric-cost guide changes the branch cost policy explicitly; the three-outcome guide adds a labelled extra outcome. Neither replaces the existing cost explanation or supplies an observed distribution. Complete actual payoffs and source-defined frequencies are still needed before an empirical expectancy claim.