Risk-Reward Ratio and the Win Rate You Actually Need

A win rate on its own is not information. It only becomes information when you know what a win is worth relative to a loss, and once you know both, a single line of arithmetic tells you whether the record in front of you makes money or quietly loses it. That is why a headline “90% win rate” should raise a question, not an eyebrow.

What R actually measures

R is shorthand for the amount you put at risk on a trade. If a trade risks 100 USD, then 1R is 100 USD for that trade, and every outcome can be described as a multiple of it. A loss at the stop is −1R. A win that returns 200 USD is +2R. The unit changes size from trade to trade; the arithmetic does not.

The risk-reward ratio is that multiple stated in advance. A trade described as 1:2 risks 1R to make 2R, so its R is 2. A 1:3 trade has an R of 3. Writing it this way strips out account size, lot size and instrument, which is precisely the point: two traders with different balances running the same 1:2 structure are running the same trade in R terms. How the size behind that 1R is set is a separate calculation, covered in Position Sizing and the 1% Rule.

Two things are worth being precise about before the arithmetic starts. First, R as planned and R as realised are different numbers. A trade that targets 3R but is closed early at 1.4R contributes 1.4R to the record, not 3R. Second, R is only honest if the stop is real. A trade that risks 1R on paper but is held through the stop has no defined R at all: it has an unknown loss wearing an R label.

The breakeven win rate for each R

For any given R there is exactly one win rate at which the account goes nowhere: wins and losses cancel and equity ends where it started, before costs. That figure is the breakeven win rate, and it comes from one formula.

Breakeven win rate = 1 ÷ (1 + R)

The logic is short enough to check by hand. Over 100 trades you win a share p of them for R each and lose the rest for 1 each. Breakeven means p × R = (1 − p) × 1, and solving for p gives 1 ÷ (1 + R). The table below is that formula applied, and every row is illustrative arithmetic rather than any observed result.

Risk-reward	R	Breakeven win rate	Check
1:0.5	0.5	66.7%	1 ÷ 1.5 = 0.667
1:1	1	50.0%	1 ÷ 2 = 0.500
1:2	2	33.3%	1 ÷ 3 = 0.333
1:3	3	25.0%	1 ÷ 4 = 0.250
1:4	4	20.0%	1 ÷ 5 = 0.200
1:5	5	16.7%	1 ÷ 6 = 0.167

Read the table twice, in both directions. Downward, it says a bigger reward multiple buys you the right to be wrong more often: at 1:3 you can lose three trades in four and still be flat. Upward, it says the opposite and says it more sharply: a strategy taking 0.5R wins needs to be right on two trades out of three simply to stand still. Neither direction is better. They are different shapes of the same requirement.

What the table does not say is that a high R is easier. Targets sit further away, so they are hit less often; that is the trade the market offers, and it is why the breakeven line moves in the direction it does.

Where costs move the line

Every figure above is the theoretical breakeven, calculated as though trading were free. It is not. Spread, commission and swap all sit between the arithmetic and the account, and they always move the required win rate up, never down.

Work an illustrative case. A trade risks 20 pips to a stop and targets 40 pips, so on paper R = 2 and the breakeven win rate is 33.3%. Now add a round-trip cost of one pip. The loss becomes 21 pips and the win becomes 39, so the realised R is 39 ÷ 21 = 1.857. The new breakeven is 21 ÷ (21 + 39) = 21 ÷ 60 = 35.0%.

One pip moved the requirement by 1.7 percentage points. Three observations follow from that, and they matter more than the specific number.

• Costs hit tight stops hardest. The same one-pip cost against a 200-pip stop is a rounding error; against a 5-pip stop it is 20% of the risk. Scalping strategies carry the heaviest cost burden per unit of R for exactly this reason.
• Swap is a clock, not a fee. Positions held overnight accrue a financing charge or credit each rollover. A strategy holding trades for days can see its effective R drift while nothing about the entry or the stop has changed.
• Costs are certain; the edge is not. You pay the spread on every trade, including the losers. Any comparison between a backtest and a live record should start here.

Expectancy: one number instead of two

The breakeven table tells you whether a record clears the bar. Expectancy tells you by how much. It is the average outcome of a trade, expressed in R.

Expectancy per trade = (win rate × R) − (loss rate × 1)

Illustratively, a strategy winning 40% of its trades at 3R returns (0.40 × 3) − (0.60 × 1) = 1.20 − 0.60 = +0.60R per trade. One winning 55% at 1R returns 0.55 − 0.45 = +0.10R. One winning 25% at 4R returns 1.00 − 0.75 = +0.25R. Three completely different-looking records, all positive, none comparable on win rate alone.

Positive expectancy is the correct thing to look for. It is also routinely over-read, so it is worth being blunt about what it does and does not mean.

What it means: if the win rate and R hold, the average trade adds equity, and the sum of many trades trends upward. That is a real and important property. It is what separates a method from a coin flip.

What it does not mean: it says nothing about the next trade, or the next twenty. Expectancy is an average, and averages arrive unevenly. Take the 45%-at-1.5R case: expectancy is (0.45 × 1.5) − 0.55 = 0.675 − 0.55 = +0.125R, comfortably positive. A stretch of 20 trades producing only 6 wins is entirely ordinary at that win rate, and it returns (6 × 1.5) − 14 = 9 − 14 = −5R. Nothing broke. The average simply had not arrived yet.

Two further limits. Expectancy is computed from a sample, so it inherits that sample’s size: a positive figure from 40 trades is a weak claim, and a positive figure from 40 trades in one market condition is weaker still. And expectancy is silent on the path: it tells you the average step, not how deep the account goes between those steps, which is a separate question with a separate answer. A positive expectancy is a necessary condition for a strategy worth following. It is not a guarantee of anything, over any horizon.

Why a headline win rate with no R is uninformative

Now the table earns its keep. Put “90% win rate” against it and the first question is no longer how but at what R, because 90% is a passing grade at some R values and a failing grade at others.

Illustratively: a strategy wins 90% of its trades, and the average win is 10 USD while the average loss is 120 USD. R is 10 ÷ 120 = 0.083. The breakeven win rate is 120 ÷ (120 + 10) = 92.3%. Expectancy is (0.90 × 10) − (0.10 × 120) = 9 − 12 = −3 USD per trade. A 90% win rate, and the account bleeds three dollars a trade before costs.

This is not an exotic construction. It is the natural arithmetic of any method that takes small profits and lets losers run: averaging into losing positions, removing stops, or holding until price comes back. Those methods produce a long, flattering row of small wins and a short, decisive column of large losses, and the win rate reports the row while ignoring the column.

So a win rate published without an R alongside it is not a strong claim being modestly stated. It is an incomplete claim, and the missing half is the half that determines the answer. Treat it as a prompt to ask for average win, average loss, and the verified record they came from, and treat a refusal to supply them as the answer itself. The wider set of things a presentation can leave out is catalogued in Verified vs Unverified Results: The Red Flags Checklist.

Applying it to a record you are evaluating

None of this is theoretical work. On a third-party-verified record you can extract both inputs directly and run the check in under a minute.

• Take the win rate from the won-versus-lost split, and note the total trade count next to it. The count is what tells you how much weight the percentage deserves.
• Compute R as average win divided by average loss, in the same unit: currency or pips, consistently. This is the realised R across the whole record, which is the number that matters, not the ratio the strategy aims for.
• Compare the win rate to 1 ÷ (1 + R). The gap between them is the margin the record is actually operating on.

An illustrative worked case. A record shows 1,000 trades, a 62% win rate, an average win of 84 USD and an average loss of 126 USD. R is 84 ÷ 126 = 0.667. Breakeven is 126 ÷ (126 + 84) = 126 ÷ 210 = 60.0%. The strategy is running two percentage points above its own breakeven line. Expectancy confirms it: (0.62 × 84) − (0.38 × 126) = 52.08 − 47.88 = +4.20 USD per trade, which is 0.033R.

That is a positive result and a thin one. A two-point margin is the kind of gap that ordinary variance, a wider spread, or a change in market conditions can close. Reading it as “62% of trades win” would have missed the entire point; reading it against the breakeven line is what turns a percentage into a judgement.

One cross-check is worth knowing, because most verified platforms display it. Profit factor (gross profit divided by gross loss) equals (win rate ÷ loss rate) × R, so it sits above 1.0 exactly when expectancy sits above zero. In the case above: (0.62 ÷ 0.38) × 0.667 = 1.63 × 0.667 = 1.09. A profit factor just above 1.0 is the same thin margin, said a different way. Where these fields live on a verified statement, and which ones can be gamed by filtering, is walked through in How to Read a Myfxbook Track Record.

A final caution on scope. Every figure here describes a record as it has already been traded. It is a measurement of the past, taken under whatever conditions the past contained, and a strategy’s realised R and win rate can both shift when conditions change. The arithmetic tells you what a record has been doing. It does not tell you what it will do next.

Key takeaways

• 1R is the amount risked on a trade, and every outcome is a multiple of it, which makes records comparable across account sizes and instruments. R on its own is the reward multiple: how many times the risk a win pays back.
• Breakeven win rate = 1 ÷ (1 + R). At 1:1 that is 50%, at 1:2 it is 33.3%, at 1:3 it is 25%, at 1:4 it is 20%.
• Spread, commission and swap always raise the real breakeven above the theoretical one, and they bite hardest on tight stops.
• Expectancy = (win rate × R) − loss rate, in R. Positive means the average trade adds equity; it promises nothing about the next twenty, and it is only as trustworthy as the sample size behind it.
• A win rate quoted without an R is an incomplete claim. Ask for average win and average loss, from a verified record.

This article is educational and is not investment advice, a recommendation, or an offer. Every figure above (win rates, risk-reward ratios, average wins and losses, pip values, expectancy figures, profit factors and trade counts) is illustrative and used solely to demonstrate the arithmetic; none are results, projections, or typical outcomes. Trading foreign exchange and CFDs carries a substantial risk of loss, including the loss of your entire investment. Positive expectancy describes an average across a past sample; it does not remove variance or market risk, and past performance is not a guarantee of future results. Consider your own circumstances and seek independent professional advice where appropriate.