R-multiples: measuring a trade in units of its own risk

How R is computed

One R is the per-unit risk multiplied by the quantity held: |entry − stop| × size × the instrument's point value. A trade's R-multiple is then its result divided by that number.

A stop at 98 on a 100 entry, 125 shares, is a risk of $250 — one R. Closing at 104 is +$750, or +3R. Closing at 98 is −1R. Closing at 99 is −0.5R: the stop was not reached, and half the risk was returned.

Why the unit matters

Dollar results are not comparable across sizes, and position size usually varies with the stop. A trader who risks a fixed percentage takes a large position on a tight stop and a small one on a wide stop; in dollars those two trades look unrelated, and in R they are on the same scale.

That is what makes an average meaningful. Expectancy in R — the mean result per trade, in units of risk — is the same number whether the account is $5,000 or $500,000, and it is the figure the sizing arithmetic actually consumes.

Where R stops working

R depends on a stop that was decided in advance and recorded. Three things break it:

• No stop recorded. There is no denominator, so the trade has no R, and excluding it quietly changes every average it would have been part of.
• A stop moved after entry. The risk that was actually taken is the original one; measuring against the moved stop reports a different trade than the one that was placed.
• A gap through the stop. The realised loss can exceed 1R, and a book that assumes losses cap at −1R will understate its own tail.

R and the break-even win rate

Because R normalises size out, a reward-to-risk ratio can be converted directly into the win rate it needs to break even: 100 ÷ (1 + R:R). A 2R target breaks even at a 33.3% win rate, a 1R target at 50%, a 5R target at 16.7%.

That conversion is what makes a ratio comparable to anything. A high reward-to-risk figure on its own says nothing about whether an approach is above water, because it says nothing about how often the target is reached.