R-Multiple Explained: Why Pros Size by Risk, Not Dollars

Most retail traders judge a trade by its dollar P&L. A $200 winner feels good. A $400 winner feels great. A $1,000 winner feels like genius. The problem with that scoring system is that it tells you nothing about whether the trade was actually a good trade.

A $200 winner where you risked $50 to make it is a 4R outcome. A $200 winner where you risked $500 to make it is a 0.4R outcome. Same dollar P&L. Wildly different setup quality. Without a unit that normalizes by risk, you cannot tell the two trades apart, and your journal becomes a record of how much money you made, not which of your setups actually works.

That unit is R.

What R actually is

R-multiple is a unit of measurement. The R stands for "risk" — specifically, the risk you committed at the moment of entry. Every trade is expressed as a multiple of that risk: +1R means you made exactly what you risked, +2R means you made twice what you risked, −1R means you lost exactly what you risked, and −1.5R means your stop blew through and you lost more than you planned.

The formula for a long trade is:

R = (exit - entry) / (entry - stop)

For a short trade you flip the sign of both the numerator and the denominator:

R_short = (entry - exit) / (stop - entry)

The denominator — the difference between entry and stop — is your initial risk per share. The numerator is your actual P&L per share. The ratio is the trade's outcome in units of the risk you took.

Worked example: a clean winner

Long entry $50, stop $48, exit $54.

R = (54 - 50) / (50 - 48)
  = 4 / 2
  = +2R

You risked $2/share. You made $4/share. The setup paid 2-to-1 on the risk you committed. That's a +2R trade.

Notice what's not in the formula: position size, stock price, percent gain. None of them. R strips all of those out. A 100-share trade and a 10,000-share trade on the same setup are both +2R if both used the same stop. That's the point. R lets you compare your $50 stock breakouts to your $400 stock fades on the same scale, because both are denominated in their own risk.

Worked example: a loser that's worse than -1R

Long entry $50, stop $48, exit $47.50.

R = (47.50 - 50) / (50 - 48)
  = -2.50 / 2
  = -1.25R

The stop was at $48, but the actual exit was $47.50 — maybe the stop blew through on a gap, maybe a fast-moving tape filled you wide of the stop price, maybe you panicked and got out below the level. The realized loss is 1.25 units of risk, not the 1.0 units you originally planned.

This is the most useful diagnostic R-multiple gives you. A book of −1R losses is healthy risk management. A book of −1.5R losses is a stop-discipline problem. The dollar P&L on the two books can be identical if position size compensates; the R-multiple tells you which book is actually well-managed.

Why R beats percent and dollars

Imagine two trades. Trade A: long a $10 stock, exits up 20% for a +$2 move per share. Trade B: long a $400 stock, exits up 0.5% for a +$2 move per share. By percent, A looks like a much better trade. By dollars, they're identical.

By R, the question is: what did each trade risk? If trade A used a 10% stop ($1 risk per share) and trade B used a 0.25% stop ($1 risk per share), they're both +2R. The percent gain was a function of the underlying stock's price, not the trader's skill. R strips that out.

This is why every serious quant fund tracks expectancy in R, not in dollars. Expectancy in dollars is dominated by which positions happened to be sized large. Expectancy in R is dominated by which setups actually have edge.

The one pitfall that breaks the framework

R is meaningless if there was never a stop.

A trader who enters without a stop, sets a stop after entry "based on where the chart says no," or moves the stop around as the trade unfolds is not generating R-multiples. They're generating P&L outcomes that someone retrofitted a denominator onto. The R numbers will look mathematically valid, but they encode nothing about the trader's actual risk discipline.

This is why Tapeboard's R-multiple methodology only computes R when a stop was set at entry — either via a bracket order's stop leg, or via a manually-entered stop field. Trades with no original stop are excluded from R-based analytics. A trader can still review the trade by P&L or by percent; the R column just stays empty.

If you want R-multiples to mean anything, set the stop before you press buy, and don't move it once you have. That's not a Tapeboard rule. That's how the unit is defined.

Where R shows up in practice

The Analytics tab in the Tapeboard trade journal surfaces three R-based views: average R per setup tag (so you can see which named setups produce positive R on average), win rate by R bucket (how often each outcome occurs), and an R-distribution histogram across all closed trades.

The first one is the most useful. Tag your trades by setup — breakout, fade, vwap-reclaim, whatever your taxonomy is. After 100 trades the average-R-per-tag view will tell you which setups have positive expectancy and which are noise dressed up as a system. That's a different question from "which setups made the most money," and it's the one that matters if you want the next 100 trades to be better than the last 100.

What R won't tell you

R is a unit, not a strategy. It will not tell you:

• Whether your setup has edge. (Run 100+ trades and check the average R per tag.)
• Whether you sized correctly. (R is per-share; position-size sets dollar exposure.)
• Whether the market regime favors your style. (R averages over too long a window hide regime shifts.)

What R will tell you, ruthlessly, is whether the trades you took matched the risk you said you were taking. If your average loser is worse than −1R, you have a stop-discipline problem regardless of how the wins look. If your average winner is below +1R, your setups are not paying enough to overcome the losers regardless of how often they hit.

Both of those questions are invisible in a P&L column. They're the first thing R shows you.