R-Multiples: Normalize Outcomes by Initial Risk

Define the denominator first

In the convention popularized by the Van Tharp Institute's explanation of R, the reference amount is initial planned risk. A realized result divided by that amount gives its R-multiple. This is a measurement convention, not proof that the initial stop would limit every possible loss.

Write down whether initial R includes estimated costs or only the entry-to-stop price distance times exposure. Either convention needs an explicit label. Moving a stop later should not silently rewrite the original denominator, because that would change the measurement after seeing part of the trade's path.

Calculate an original trade example

Assume a hypothetical linear position of 30 units entered at 50 with an initial stop scenario at 48. Using a price-only definition, initial R is (50 − 48) × 30 = 60 monetary units. An exit at 55 produces 150 gross. If all attributable costs total 6, the net outcome is 144 and the net result against price-only initial R is 144 ÷ 60 = +2.4R.

If instead the initial convention included the same estimated 6 in the denominator, initial R would be 66 and the result would be approximately +2.18182R. These are different labeled definitions, not contradictory arithmetic. Retain the chosen convention throughout a comparison.

Why −1R is not a guaranteed floor

Suppose the first example exits at 47 because the market gaps past the intended stop. Gross loss is 90; adding 6 in costs makes the signed net result −96. Dividing by the original price-only R of 60 yields −1.6R. Reporting −1R simply because that was the plan would erase the execution difference the journal should preserve.

Partial exits can be handled by summing the complete campaign's realized net outcome and dividing by its documented initial R. If additional entries materially change exposure, define a campaign-level allocation rule in advance or report the legs separately. Do not compare a fill-level numerator with an unrelated whole-position denominator.

Compare normalized and monetary summaries

Trade A risks 100 and returns +200, or +2R. Trade B risks 1,000 and returns −1,000, or −1R. The equally weighted mean is +0.5R, while combined money is −800. Neither arithmetic result is wrong. The difference shows that an equal-trade R average does not encode how much capital each observation received.

Keep the original monetary amounts, initial risk and dates alongside R. A zero initial-risk amount makes division undefined, and a retrospectively chosen tiny denominator can inflate the statistic. R also does not capture time held, capital tied up or the dependence between simultaneous positions. For copied accounts, preserve each account's actual costs and fills; an identical planned stop does not ensure identical realized R.