R-Multiples: The Single Metric That Tells You If Your Strategy Works

You took 50 trades this quarter. Some won, some lost. Some were big positions, some small. Some had tight stops, some wide. Looking at "$3,200 net profit" tells you almost nothing about whether you're trading well, it could be one giant winner masking 49 losers, or 49 small winners and one disaster. The R-multiple framework is how you actually answer "did I trade well?"

The R definition

R = the dollar amount you risked on a trade. It's defined by your stop-loss at the moment of entry.

Long 1 BTC at $66,000, stop at $65,000 = $1,000 of risk = 1R. Long 0.5 ETH at $2,400, stop at $2,300 = $50 of risk = 1R.

The R for each trade is whatever number falls out of (entry − stop) × position size. It varies trade by trade because positions and stops vary. That's the whole point, R is the unit that makes disparate trades comparable.

Trade outcomes as R-multiples

Once R is set at entry, every outcome is expressed as a multiple of that R.

Trade exit	Outcome
Stop fires (full loss)	-1R
Closed at half the planned loss	-0.5R
Break-even exit	0R
Hit 1× target distance	+1R
Hit 2× target distance	+2R
Trade ran 5× target before exit	+5R

Now your trade ledger transforms from "$200, -$80, $340, -$120, $50, -$150..." into "+2R, -1R, +3R, -1R, +0.5R, -1R..." The R-multiples are directly comparable across trades, position sizes, and time periods in a way the raw dollars aren't.

Average R, the strategy's expectancy

Expectancy = average R-multiple per trade.

Compute it as: (sum of all R-multiples) / (number of trades).

If your last 100 trades sum to +30R, your expectancy is +0.3R per trade. That means on average, every trade you take adds 0.3R to your account. With 1% risk per trade, that's 0.3% expected gain per trade, or 30% expected per 100 trades, before fees.

Expectancy is the single most important number for evaluating a strategy. Win rate alone is meaningless. PnL alone is misleading. Expectancy in R captures both win rate AND average win/loss size in one number.

The relationship between win rate, average R, and expectancy

Two strategies, both with the same expectancy:

Strategy A: 70% win rate, average winner +0.5R, average loser -1R. Expectancy = 0.7 × 0.5 − 0.3 × 1 = +0.05R per trade.

Strategy B: 30% win rate, average winner +3R, average loser -1R. Expectancy = 0.3 × 3 − 0.7 × 1 = +0.2R per trade.

A wins more often. B is 4x more profitable per trade. Most beginners gravitate to A because the frequency of being right feels good. The math doesn't care how you feel, B is the better strategy.

The biggest single edge in trading is letting winners run while cutting losses fast. That trades win rate for average R, often a worthwhile trade. Most psychologically difficult strategies (trend following, breakout trading) have low win rates and high average R; most psychologically easy ones (mean reversion in chop, scaling into losers) have high win rates and low average R.

Worked example, evaluating two real-looking quarters

Trader X (Q1): 100 trades, 65 winners, 35 losers. Average winner +0.6R, average loser -0.9R.

• Expectancy = 0.65 × 0.6 − 0.35 × 0.9 = 0.39 − 0.315 = +0.075R
• "65% win rate!" sounds great. Per-trade edge is barely positive.

Trader Y (Q1): 100 trades, 38 winners, 62 losers. Average winner +2.4R, average loser -1.0R.

• Expectancy = 0.38 × 2.4 − 0.62 × 1.0 = 0.912 − 0.62 = +0.29R
• "38% win rate" sounds bad. Per-trade edge is 4x Trader X's.

After 100 more trades, Trader Y is +29R total ahead of where they started; Trader X is +7.5R. Trader Y compounded 4x faster on the same trade volume. The win rate was the wrong number to optimize.

R-multiples for trade-level discipline

R-multiples force you to think about exits in terms of risk taken, not dollars made. Examples:

"Should I take partial profits at +0.5R?" Almost always no. Cutting winners at half their planned target collapses your average R toward zero. The asymmetry that makes you money is letting them run to full target.

"Should I move my stop closer after the trade goes my way?" Maybe. But understand that you've now changed the R for the trade: if you originally risked 1R and tighten the stop so you're now risking 0.3R, your max loss is now 0.3R, but your potential winner in R-multiples shifts proportionally smaller too. The decision should be conscious, not "I want to lock in some profit."

"Should I add to a winner?" Possibly, but each addition has its own R based on its own entry and stop. Sloppy adds blow up the trade-by-trade R math because you can't tell which add was profitable.

A common mistake: not recording R for every trade

People keep PnL journals. Few keep R journals. Without R, you can't compute expectancy, you can't see whether your strategy is deteriorating, and you can't tell whether a 6-trade losing streak is normal noise or a sign your edge is gone.

The fix: in your trade log, record entry, stop, position size, and final exit. Compute R at entry. Compute R-multiple at exit. Plot your average R over rolling 30-trade windows. When that line falls, you have a data-driven reason to investigate. When it stays positive across regimes, you have evidence your edge is real.

A common mistake: cherry-picking R-multiples

"My average R is +1.2, except for these three trades I shouldn't have taken." Once you start excluding trades from your stats, your stats are fiction. Every trade you took counts. Every trade your strategy would take counts. If your strategy includes trades you "shouldn't take," then your strategy is wrong, and you need to change the strategy or accept the lower expectancy.

The discipline is brutal honesty in measurement. Cooked stats lead directly to cooked confidence, which leads to position sizes you can't justify, which leads to drawdowns.

Mental model, R as the scale that makes trades comparable

Imagine measuring the speeds of a car, a bicycle, and a runner. You need a common unit (mph) or comparison is impossible. R is that unit for trade outcomes. A 200-pip win on EURUSD and a 5% gain on SOL aren't directly comparable, different markets, different volatilities, different position sizes. But if both were +2R for your account on those trades, they had identical impact on your strategy's expectancy. R abstracts away everything that doesn't matter and isolates what does.

Why this matters for trading

Hex37's journal page (/app/journal) computes R-multiples per trade, average R per breakdown bucket (instrument, day-of-week, hour, session), and rolling expectancy. Use it. The breakdowns will surface things your gut won't notice, that you make money in Asia hours and lose it in NY hours, that your alts trades are +0.4R but your BTC trades are -0.1R, that your "high conviction" setups are actually below average. The data is brutal and useful.

Takeaway

R is the dollar risk per trade, set at entry by your stop. R- multiples normalize every outcome to that risk so trades become comparable. Average R per trade (expectancy) is the cleanest number for telling whether a strategy works. Win rate without average R is meaningless. Track R for every trade you take. Most of the discipline in professional trading reduces to: do whatever keeps your average R positive across thousands of trades.