Payoff Ratio: Average Win vs Average Loss, Formula and Examples

The payoff ratio is the average winning trade divided by the average losing trade, with the loss taken as a positive amount: payoff ratio = average win ÷ average loss. A payoff ratio of 3.0 means the average win was three times the average loss. The ratio measures how large wins are compared with losses, not how often trades win, so it describes a strategy only when read with the win rate.

The payoff ratio is one of the per-trade measures in the trading metrics library. Check the formula when a figure is called an “expectancy ratio”: expectancy is an amount per trade, and the payoff ratio ignores how often trades win. All numbers below are illustrative.

Payoff ratio at a glance

• Definition: average win ÷ average loss over closed trades, with the average loss as a positive number.
• Formula: AW ÷ AL, where AW = gross profit ÷ winning trades and AL = absolute gross loss ÷ losing trades.
• Written as: a number such as 3.0, or a ratio such as 3:1.
• Break-even point: none on its own. With win rate W and no breakevens, expectancy is zero when the payoff ratio equals (1 − W) ÷ W, before costs.
• Other names: average win/loss ratio and realised reward-to-risk. Some pages also say “win/loss ratio” or “expectancy ratio”.
• Where to find it: “Average profit / average loss” in the TradingView strategy report; “Average profit trade” ÷ the absolute value of “Average loss trade” in MT5; PineConnector Analytics shows a related “Reward-to-risk” figure.
• Undefined when: there are no winning trades or no losing trades, so one of the required averages is missing.

What is the payoff ratio formula?

Payoff ratio = average win ÷ average loss = AW ÷ AL

• AW, average win: gross profit ÷ number of winning trades.
• AL, average loss: gross loss ÷ number of losing trades, written as a positive number.
• W and L: the win rate and loss rate, each as a share of all closed trades.

TradingView's Average profit / average loss metric compares the monetary size of the average winner with the average loser.^[1] MetaQuotes defines the MT5 tester's Average profit trade as total winning profit divided by winning-trade count, with Average loss trade calculated over losing trades.^[2]

This page uses five conventions:

• Closed trades only. TradingView's ratio excludes open positions.^[1]
• Breakevens sit in neither average. TradingView's Average profit and Average loss both leave out trades that ended at break-even.^[3][4] Breakeven trades change the win rate but not the payoff ratio.
• The loss as a magnitude. MQL5's testing statistics note that gross loss “is always less than or equal to zero”, so divide by its absolute value.^[5] A negative payoff ratio means the sign was never removed.
• One cost basis. Take AW and AL from results that include the same costs, or compare both before costs.
• Currency or R. Dividing each result by its positive initial risk first gives separate winning and losing averages in R-multiples. Divide those averages to obtain an R-based payoff ratio. With the same planned risk on every trade, it matches the currency-based ratio; changing risk can make them differ.

Worked example: payoff ratio from a 100-trade log

Illustrative numbers, not a recommendation or a strategy result. A log before costs holds 100 closed trades in a USD account: 40 winners with a gross profit of 12,000 USD, 60 losers with a gross loss of 6,000 USD, and no breakevens.

1. Average win AW = 12,000 ÷ 40 = 300 USD.
2. Average loss AL = 6,000 ÷ 60 = 100 USD.
3. Payoff ratio = 300 ÷ 100 = 3.0, often written 3:1.
4. Win rate W = 40 ÷ 100 = 0.40. Loss rate L = 60 ÷ 100 = 0.60.
5. Expectancy = (0.40 × 300) − (0.60 × 100) = 120 − 60 = 60 USD per trade. Check: (12,000 − 6,000) ÷ 100 = 60 USD.
6. Profit factor = 12,000 ÷ 6,000 = 2.0.
7. Break-even win rate for a payoff of 3.0, before costs = 1 ÷ (1 + 3.0) = 25%.

With unchanged win/loss classification, costs shrink the average win and enlarge the average loss, lowering the ratio. Deduct an illustrative 25 USD per trade for spread and commission, assuming no trade changes sign and those costs are absent from the starting results. Do not deduct costs already reflected in fills again. If a small winner becomes a loser, rebuild both groups; the payoff ratio need not decrease even though net profit falls. AW falls to 275 USD and AL grows to 125 USD, so the payoff ratio falls from 3.0 to 275 ÷ 125 = 2.2.

After that cost, expectancy is (0.40 × 275) − (0.60 × 125) = 110 − 75 = 35 USD per trade. The break-even win rate moves from 25% to (100 + 25) ÷ (300 + 100) = 31.25%, which matches 1 ÷ (1 + 2.2).

How do payoff ratio and win rate combine?

Payoff ratio and win rate describe complementary parts of a trade log. Monetary expectancy also needs the average loss and the loss rate. TradingView notes that its average profit to average loss ratio “does not factor in the frequency of trades (Win rate)”.^[1]

Expectancy = AL × (W × payoff ratio − L)

Profit factor = (W ÷ L) × payoff ratio

Both follow from the library's definitions. Expectancy is (W × AW) − (L × AL), and AW equals payoff ratio × AL. Profit factor is (W × AW) ÷ (L × AL). In the example, 100 × (0.40 × 3.0 − 0.60) = 60 USD, and (0.40 ÷ 0.60) × 3.0 = 2.0. W ÷ L also equals winning trades ÷ losing trades, because both rates share one denominator.

Setting expectancy to zero gives the break-even line. Before costs and with no breakevens, the break-even win rate is 1 ÷ (1 + payoff ratio). PineConnector's Analytics guide gives the same formula for reward-to-risk in a simplified case where every win is the same size and every loss is the same size, with 25% at 3:1.^[6] Read the other way, a positive win rate W breaks even at a payoff ratio of (1 − W) ÷ W. Break-even payoff ratios before costs, with no breakeven trades:

• 20% win rate: 0.80 ÷ 0.20 = 4.0
• 25% win rate: 0.75 ÷ 0.25 = 3.0
• 40% win rate: 0.60 ÷ 0.40 = 1.5
• 50% win rate: 0.50 ÷ 0.50 = 1.0
• 60% win rate: 0.40 ÷ 0.60 ≈ 0.67
• 75% win rate: 0.25 ÷ 0.75 ≈ 0.33

These values are arithmetic, not targets, and say nothing about the next trade.

With a payoff ratio of 3.0, a 20% win rate loses 20 USD per trade, 25% breaks even and 40% makes 60 USD. A payoff ratio of 0.6, below 1, still gives +12 USD per trade at a 70% win rate, because 70% sits above its 62.5% break-even point. The win rate guide covers the other half of the pair.

The two-outcome, no-breakeven Kelly form, f* = W − (1 − W) ÷ b, uses the fixed win/loss payoff as b. Substituting averages for a variable outcome distribution is an approximation, not a full Kelly calculation. The Kelly criterion page explains why an estimated b gives only an estimated fraction.

Is the win/loss ratio the same as the payoff ratio?

Not always. Sources use “win/loss ratio” for two different numbers, so check which one a figure means before comparing it with another.

• Count ratio: winning trades ÷ losing trades. Corporate Finance Institute defines win/loss ratio using profitable-trade count divided by unprofitable-trade count, without weighting by monetary size.^[7] In the example, 40 ÷ 60 ≈ 0.67.
• Size ratio: average win ÷ average loss, which is the payoff ratio. TurtleTrader's page on the Kelly formula defines its R as the “Historical Average Win/Loss ratio”.^[8] In the example, 3.0.
• Win rate: winning trades ÷ all closed trades, 40% in the example. TradingView reports it as Percent profitable and keeps even trades in the denominator.^[9]

The count ratio and the payoff ratio multiply to the profit factor: (40 ÷ 60) × 3.0 = 2.0. A count ratio of 1.0 equals a 50% win rate only when there are no breakeven trades.

The “expectancy ratio” mix-up

If a figure labelled “expectancy ratio” is calculated as AW ÷ AL, it is the payoff ratio. Monetary expectancy is (W × AW) − (L × AL), an amount per trade. It can change sign with win rate while the payoff ratio stays unchanged. Use the formula, rather than the label, to identify the metric.

Realised reward-to-risk

Realised reward-to-risk over closed trades is the payoff ratio. TradingView's help says comparing Average profit with Average loss “is essential for determining Risk/Reward ratio”.^[3] Planned reward-to-risk is a different number: one order's target distance ÷ its stop distance, fixed before the trade exists. TechTarget notes that “some traders use the ratio in reverse”, so state the direction of division.^[10] The risk-reward ratio guide covers the planned version.

Where to find average win, average loss and the payoff ratio

Labels as of 25 September 2026:

• TradingView strategy report (formerly the Strategy Tester^[11]): Average profit, Average loss, and Average profit / average loss. Closed trades only; breakevens are left out of both averages.^[1][3][4]
• MT5 Strategy Tester report: Average profit trade and Average loss trade. No ratio field is documented; divide one by the absolute value of the other.^[2]
• MT5 Trading Report, Risks tab: Avg. Profit and Avg. Loss, described as average actual profit or loss on closing positions.^[12]
• PineConnector Portal Analytics: Reward-to-risk, described as the size of a win compared with a loss. Analytics needs MT5 EA v3.41 or later and is not available for MT4 in the published guide.^[6]

TradingView's Strategies example divides strategy.avg_winning_trade by strategy.avg_losing_trade and labels the result Avg. win / Avg. loss.^[13] After a test, MQL5's TesterStatistics() function returns the same inputs: gross profit, gross loss and the counts of profitable and losing trades.^[5] The TradingView strategy report guide and the MT5 Strategy Tester report guide map the neighbouring fields.

Check the cost basis before comparing two figures. A TradingView strategy without commission arguments applies no commission by default; users can override that default in Settings/Properties.^[13] MT5's Trading Report calculates its Summary gross profit and loss “excluding swaps and commissions”, while its Profit/Loss tab includes them. Its help does not state the cost basis of Avg. Profit and Avg. Loss.^[12]

Whether PineConnector Analytics calculates Reward-to-risk as an average over the period, and which costs it includes, is not documented in the Analytics guide; test on demo by checking it against AW ÷ AL from the same account's closed trades. Read it as history: the guide states that “A historical average is not a promised result”.^[6]

Planned versus realised payoff: from TradingView alert to MT5 order

A PineConnector workflow runs from an idea to a measurable rule, a TradingView alert and an MT5 order. An alert can request a stop and a target, which fixes a planned reward-to-risk; the payoff ratio is measured afterwards, on trades that closed. Illustrative message for the TradingView alert's Message field. Replace LicenseID with the ID from Portal Connections and EURUSD with the broker's exact symbol:

LicenseID,buy,EURUSD,vol_lots=0.01,sl_pips=20,tp_pips=40

From an illustrative entry of 1.10000, the syntax reference calculates a requested stop at 1.09800 and a requested target at 1.10400: a planned 2:1 in pips. The broker must accept the requested levels, volume increments and minimum distances.^[14] Try any message on a demo account first.

The realised payoff ratio of those trades can still differ from 2.0:

• Spread and costs. For this forex example, MT5 buys at Ask and sells at Bid, and checks a long position's stop against Bid. Exchange-traded symbols can use different trigger rules.^[15] Commission and swap also change results.
• Fills and gaps. PineConnector's syntax reference warns that costs, fills, gaps and broker limits can change the actual loss.^[14]
• Exits before the stop or target. A strategy close signal, a breakeven move or a trailing stop ends trades at other distances. A native stop accepted by the broker is stored and executed on its server.^[16] A stop or target that exists only in the TradingView strategy is different: PineConnector's converter guide says the integration does not place simulated strategy stops and targets as protective broker orders.^[17]
• Trade counting. MT5's Trading Report counts operations that realise a result, including exits, partial exits and reversals, as trades.^[12] One position closed in two parts can add two trades to deal-based averages.
• Missing trades. The stages are separate: the alert condition becoming true and the alert triggering; webhook delivery; PineConnector processing the signal; the EA's order request; broker acceptance, the deal and the position. A missed entry can leave a trade absent from MT5; a missed exit can leave an existing position open.
• Different prices. TradingView and MetaTrader can use “different data feeds, bid/ask quotes, spreads and timestamps”.^[18]

Measure the realised payoff ratio from the MT5 account's closed trades, and compare it with the TradingView report over matching trades. The checklist from a TradingView backtest to a verified demo trade and why TradingView and MT5 results differ cover that comparison.

What the payoff ratio hides: limitations and common mistakes

• Frequency. The ratio ignores how often trades win.^[1] In the profiles above, a payoff ratio of 3.0 works out to −20 USD per trade at a 20% win rate.
• One large winner. Illustrative: if the largest of the example's 40 winners made 2,250 USD, the other 39 made 9,750 USD, an average of 250 USD. Without that trade, the payoff ratio is 2.5 rather than 3.0. TradingView's “Largest profit as % of gross profit”, here 2,250 ÷ 12,000 = 18.75%, shows how much a single outlier carries.^[19]
• Averages versus typical trades. A mean can sit far from most results. TradingView's Trades analysis draws the average profit and loss on a histogram of trade returns.^[13]
• Few losing trades. Illustrative: with two losses of 100 USD each, AL is 100 USD. A third loss of 400 USD makes AL (200 + 400) ÷ 3 = 200 USD, which halves the payoff ratio if AW is unchanged. With no losses or no wins, one required average is missing and the ratio is undefined. See how many trades a backtest needs.
• Changing position size. Larger positions dominate currency averages. R-multiples put trades of different sizes on one scale.
• Tuning exits to the ratio. Moving a target or a stop can change both the payoff ratio and the win rate. Adjusting either until history looks attractive risks overfitting.
• History, not a forecast. A payoff ratio summarises past or simulated trades. Automation executes the rules as written; it does not create an edge.

Frequently asked questions

What is the payoff ratio in trading?

The payoff ratio in trading is the average winning trade divided by the average losing trade, with the loss taken as a positive amount. A payoff ratio of 3.0 means the average win was three times the average loss. The ratio ignores how often trades win, so read it together with the win rate.

How do you calculate the average win to average loss ratio?

Divide gross profit by the number of winning trades to get the average win. Divide the absolute gross loss by the number of losing trades to get the average loss. Then divide the average win by the average loss; both winning and losing trades are required. Illustrative: 12,000 USD over 40 winners and 6,000 USD over 60 losers give 300 ÷ 100 = 3.0.

Is the win/loss ratio the same as the payoff ratio?

Not always. Some sources use win/loss ratio for the number of winning trades divided by the number of losing trades, which is a ratio of counts. Others use it for average win divided by average loss, which is the payoff ratio. With 40 winners averaging 300 USD and 60 losers averaging 100 USD, the count ratio is about 0.67 and the payoff ratio is 3.0.

Is the realised reward-to-risk ratio the same as the payoff ratio?

Yes. Realised reward-to-risk over closed trades is the average win divided by the average loss, which is the payoff ratio. Planned reward-to-risk is different: it divides one order's target distance by its stop distance before the trade exists. Costs, fills and exits before the stop or target can make the realised figure differ from the plan.

What is a good payoff ratio?

There is no universal good payoff ratio, because the ratio needed depends on the win rate. Before costs, with no breakeven trades and a positive win rate, expectancy is zero when the payoff ratio equals (1 − win rate) ÷ win rate: 1.5 at a 40% win rate and about 0.67 at 60%. Costs move that threshold, and a small sample makes any measured ratio unreliable.

Reviewed 25 September 2026. Facts were checked against the linked sources on that date. Nothing in this article was tested on a trading account and no code was compiled.

Related reading

• Trading metrics library: every formula in one place
• Trading expectancy: the average result per trade
• Profit factor: gross profit divided by gross loss
• Break-even win rate, before and after costs
• Win rate: what it measures and what it misses

Sources

1. TradingView – Average profit / average loss, accessed 25 September 2026.
2. MetaQuotes – MetaTrader 5 Help: Testing Report, accessed 25 September 2026.
3. TradingView – Average profit, accessed 25 September 2026.
4. TradingView – Average loss, accessed 25 September 2026.
5. MetaQuotes – MQL5 Reference: Testing Statistics, accessed 25 September 2026.
6. PineConnector – Trading analytics: performance metrics, accessed 25 September 2026.
7. Corporate Finance Institute – Win/Loss Ratio, accessed 25 September 2026.
8. TurtleTrader (Michael Covel) – The Kelly Formula: How Bell Labs Research Shaped Trend Following Money Management, accessed 25 September 2026.
9. TradingView – Percent profitable, accessed 25 September 2026.
10. TechTarget – What is the risk-reward ratio?, accessed 25 September 2026.
11. TradingView – Pine Script release notes, accessed 25 September 2026.
12. MetaQuotes – MetaTrader 5 Help: Trading Report, accessed 25 September 2026.
13. TradingView – Pine Script v6 User Manual: Strategies, accessed 25 September 2026.
14. PineConnector – Syntax: take profit and loss-based sizing, accessed 25 September 2026.
15. MetaQuotes – MetaTrader 5 Help: Basic principles, accessed 25 September 2026.
16. MetaQuotes – MetaTrader 5 Help: Performing deals, accessed 25 September 2026.
17. PineConnector – Pine Script converter, accessed 25 September 2026.
18. PineConnector – Frequently asked questions: price differences, accessed 25 September 2026.
19. TradingView – Largest profit as % of gross profit, accessed 25 September 2026.

PineConnector executes the instructions you send it. It does not select trades, manage money, or hold funds. Trading carries risk, and past performance of any strategy does not indicate future results.