Skip to main content

Position-Size Formula (Risk / Stop Distance)

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,240 words

The risk-based position-size formula answers the only question that matters once a swing trade is planned: how many shares? It inverts the usual order of thinking. Instead of deciding a dollar amount to deploy and then hoping the stop survives, you fix the dollar you're willing to lose if the trade is wrong, decide where the stop must sit (the chart level that invalidates the idea), and let those two numbers dictate the share count. The core tension it resolves: a trader's natural impulse is to size by conviction or by available buying power, but those have no relationship to the actual risk of the trade. This formula severs share count from emotion and ties it to a single, constant, pre-committed loss.

How it's calculated

Three quantities, two steps:

1. Dollar risk (R) = Account equity × Risk-per-trade % 2. Stop distance = Entry price − Stop price (for a long; reverse for a short) 3. Shares = Dollar risk ÷ Stop distance

In one line: Shares = (Equity × Risk%) ÷ (Entry − Stop), rounded down.

Worked example (common across calculators and Van Tharp's framework): $25,000 account, risking 1% = $250. Entry $40.00, stop $37.00, so stop distance = $3.00. Shares = $250 ÷ $3.00 = 83 (83.3 rounded down). Position notional = 83 × $40 = $3,320 — about 13% of the account, even though only 1% is at risk. Widen the stop to $5.00 and the same $250 risk buys only 50 shares; tighten it to $1.50 and it buys 166. Dollar risk stays flat; share count floats inversely with stop distance. This is Van Tharp's "CPR" relationship — Capital at risk = Position size × Risk-per-unit.

How it's used in practice

The discipline is in the order of operations. A master swing trader never reverses these steps:

  • Stop first, from the chart — not from the math. Place the stop where the thesis is wrong (below the breakout base, under the pivot low, beyond the ATR-defined noise band), then back into share count. Setting the stop to fit a desired position size is the cardinal sin this formula is meant to prevent.
  • Risk% is the trader's chosen constant. Widely cited professional ranges run 0.5%–2% of equity per trade, with 1% or less the common default for traders still proving consistency (Trade That Swing, Chart Guys, multiple calculator references). The percentage is a function of edge quality, win rate, and drawdown tolerance — not a universal truth.
  • Apply a notional / buying-power cap. Because tight stops can demand large positions, sizing purely on risk can return a notional that exceeds available capital or concentrates the book. The standard guardrail: take the smaller of the risk-based size and a hard cap (e.g. no single position above X% of equity, or no more than available cash/margin). A tight stop should never let one name dominate the account.
  • Add expected slippage to the stop distance. The formula assumes you exit at the stop. In reality, fast moves, illiquid names, and especially overnight gaps (earnings, news) blow through stops. Swing traders hold overnight, so gap risk is structural — pad the denominator, size down before catalysts, or avoid holding through earnings entirely.
  • Track portfolio heat. Per-trade sizing is necessary but not sufficient. Ten positions each risking 1% is 10% portfolio heat — the loss if a correlated selloff triggers every stop at once. Disciplined desks cap total open risk (Alexander Elder's Trading for a Living popularized a ~6% monthly cap; ~6–10% total-at-risk is the commonly cited band) and treat correlated positions as one risk unit.

Adoption, debate & evidence

The risk/stop-distance formula is close to universal among rule-based discretionary and systematic traders, and is the engine inside essentially every "position size calculator" online. It is most strongly associated with Van Tharp (Definitive Guide to Position Sizing), who argues position sizing is the part of trading that does the heavy lifting toward meeting objectives.

What deserves a caveat is the headline statistic Van Tharp and many educators repeat: that position sizing accounts for "91% of the variability in performance." This number is borrowed from Brinson, Hood & Beebower (1986/1991) — a study about asset-allocation policy across stocks/bonds/cash among pension managers, not per-trade position sizing. More importantly, the academic record (Ibbotson & Kaplan, 2000; CFA Institute commentary; the Asset Allocation Hoax critique) shows the figure was itself widely misinterpreted: Brinson measured time-series R² (how much a portfolio moves with markets over time), which says almost nothing about cross-sectional differences between portfolios in a given period. So the "91%/93.6%" stat does not validate the position-sizing formula at all — it's a misapplied number. Flag: treat any claim that "position sizing explains 90%+ of performance" as folklore, not evidence.

The defensible claims, which the math itself guarantees, are narrower and real: risk-based sizing bounds single-trade loss, normalizes risk across setups with different stop distances, and survives losing streaks by shrinking dollar exposure as equity falls (when risk% is applied to current equity). Those are properties of the formula, not empirical edges — they reduce ruin probability; they don't generate return.

Strengths & limitations

Strengths. Converts a vague intent ("don't lose too much") into an exact, repeatable share count; makes every trade risk the same regardless of stop width; mechanically de-risks during drawdowns; removes conviction-driven over-sizing — the single biggest account-killer.

Limitations & #1 misuse. The formula is only as honest as its stop. The most common misuse is using a stop too tight to give the math a small denominator, producing an oversized position that gets shaken out by normal noise — the trader took more real risk while the spreadsheet said 1%. Second failure mode: ignoring gap/slippage, so realized loss exceeds the planned R. Third: applying it per-trade while ignoring correlation — six "1% risk" positions in the same sector is one ~6% bet. Fourth: sizing off a static account number rather than current equity, which breaks the drawdown-protection property. The formula sizes risk; it does not create edge, choose the stop, or account for correlation — those remain the trader's job.

Sources

Disputed/flagged: The "position sizing explains ~91% of performance" claim is a misattributed/misinterpreted statistic and should not be cited as evidence for the formula's efficacy.