The other half of the truth: reward-to-risk

Ask a beginner what makes a good trader and you will hear one answer: a high win rate. It sounds obvious. It is also the single most expensive misunderstanding in retail trading — because a 90% win rate can go bankrupt, and a 35% win rate can get rich. This column is about the other half of the truth: the reward:risk ratio.

The 90% winner who blew up

Imagine a strategy that takes profit at +2 pips and holds losses until they come back. Most trades come back. The scoreboard says 90% wins, and the account grows in tidy little steps — until the one trade that doesn't come back, which returns six months of profits in an afternoon. High win rate, catastrophic average loss: the most common shape of a blown account.

Win rate alone tells you how often you're right. It says nothing about what being wrong costs. Trading outcomes are decided by the pair, never by either number alone.

R — the unit that makes trades comparable

Define your risk per trade (from your sizing formula) as 1R. A trade that risks ¥6,000 and makes ¥12,000 is +2R; one that loses its budget is −1R. Denominating everything in R does two things: it makes trades of different sizes comparable, and it turns your journal into a dataset. From here on, your strategy has exactly two vital signs: win rate and average R:R.

The seesaw — break-even win rates

For every reward:risk ratio there is a win rate at which you break even. The arithmetic is one line: required win rate = 1 ÷ (1 + R:R). What matters is the shape of the table it produces.

Read the two ends. A scalper harvesting +0.5R wins must be right 67% of the time just to stand still — every mistake costs two wins. A trend follower taking +3R wins can be wrong three times out of four and still profit. Neither is "better"; they are different bargains with probability. But beginners overwhelmingly drift to the top row — quick small wins, occasional deep losses — which is the bargain with the harshest terms.

Expected value — the only number that decides

Combine the two vital signs and you get the verdict: EV = (win rate × average win) − (loss rate × average loss). A 40% win rate with 2R wins gives (0.4×2) − (0.6×1) = +0.2R per trade. Thirty trades a month at +0.2R is +6R — with 2% risk, roughly +12% before compounding. From a strategy that loses more often than it wins.

This is the number your backtest (VOL.001) exists to estimate, and the number your journal keeps updating. If it is positive after costs, you have a business. If it is negative, no amount of discipline, psychology or motivation will save it — you would simply be executing a losing business plan very well.

Two levers, one realistic

To raise EV you can raise the win rate or raise the R:R. The win rate lever is mostly an illusion — it is capped by market randomness, and chasing it (wider stops, earlier exits) usually just trades R:R away in disguise. The R:R lever is mechanical and available today: cut losers at the planned −1R without negotiation, and stop harvesting winners at +0.4R out of fear. Both are order-placement habits (VOL.014 covers why fear truncates winners, and the pre-ordered exits that fix it).

One honest caveat: R:R cannot be raised infinitely — targets further away are hit less often, so pushing R:R lowers the win rate. The point is not "maximize R:R"; it is that the two numbers must be designed together, and verified as a pair — never admired one at a time.

Your own numbers — the ten-minute audit

Pull your last 30 trades. Compute three things: win rate, average win in R, average loss in R. Most struggling traders discover the same silhouette: a decent win rate, average wins around +0.6R, average losses around −1.3R (losses past the stop, "just this once"). That silhouette loses money at almost any win rate — and it is fixed not by a new strategy, but by making the exits mechanical. The numbers will tell you exactly which half of the truth you've been neglecting.

Frequently asked questions