Position Sizing Calculators – The 2026 Guide with Worked Examples

Before any method, framework, or calculator, there is one equation. Every tool you will ever use is just a specific implementation of this:

Position size = (Account × Risk%) ÷ (Stop distance × Value per unit)

The variables:

• Account: the equity you are trading, in account currency.
• Risk%: the fraction you are willing to lose if the stop is hit (typically 0.25%-2%).
• Stop distance: how far the stop sits from the entry, in pips / points / currency units of the asset.
• Value per unit: how much 1 pip / 1 point / 1 coin is worth in account currency, per unit of position size.

The output is position size – in lots, contracts, shares, coins – such that a full stop-out equals exactly your risk budget.

Calculators exist because the "value per unit" input requires asset-specific arithmetic, and getting it wrong silently (rather than catastrophically) is one of the most common error modes. But the equation itself is always the same.

Fixed-fractional sizing means risking a constant percentage of the account on every trade. Named formally in Ralph Vince's Mathematics of Money Management (1992), it has three properties that make it the default choice:

• Auto-adjusting: as the account grows, the absolute dollar risk grows; as it shrinks, the dollar risk shrinks – compounding on the way up, decompounding on the way down. This is geometric growth when the edge is positive.
• Scale-free: the sizing logic does not change when the account size changes.
• Psychologically tolerable: at 0.5-1% risk per trade, a normal 5-10 trade losing streak is emotionally survivable.

Worked example – FX

• Account: $100,000
• Risk: 0.5% = $500
• Pair: EUR/USD
• Stop: 20 pips from entry
• Pip value: $10 per pip per standard lot
• Position: $500 ÷ (20 pips × $10) = 2.5 standard lots

If the account drops to $95,000 after a drawdown, the same 0.5% is now $475 – and the calculated position size adjusts down automatically. This is the feature: the sizing shrinks when you most need protection.

Worked example – futures (MES)

• Account: $25,000
• Risk: 1% = $250
• Contract: CME Micro E-mini S&P 500 (MES)
• Stop: 8 points from entry
• Point value: $5 per point per contract
• Position: $250 ÷ (8 × $5) = 6.25 → round down to 6 contracts

Futures are integer-contract instruments – always round down, never up. The rounding error is a small reduction in risk, not an increase.

Fixed-ratio sizing, introduced by Ryan Jones in The Trading Game (1999), increases the number of contracts/lots only after the account grows by a fixed dollar amount (the "delta"). Formally:

Contracts = floor(sqrt(1 + 2 × (equity − starting) ÷ delta))

This grows more slowly than fixed-fractional on the way up and less aggressively on the way down. It is favoured by some systematic futures traders because it avoids the rapid position-size inflation that fixed-fractional produces during strong equity runs.

For most discretionary and evaluation-stage traders, fixed-ratio is unnecessary complexity – fixed-fractional with a sensible risk% is cleaner. But fixed-ratio is worth knowing for the conceptual point: sizing rules have different variance profiles, and the rule that maximizes expected geometric return (Kelly) is different from the rule that minimizes drawdown variance (fixed-ratio with small delta).

Fixed-fractional with a fixed pip stop breaks when you trade multiple assets, because a "20-pip stop" means something very different on EUR/USD (intraday noise) than on GBP/JPY (roughly 1.5x larger moves) than on BTC (completely different scale). Solution: let the stop distance scale with realized volatility.

The Average True Range (ATR) is a 20-day (or 14-day) moving average of daily high-low ranges, developed by J. Welles Wilder in New Concepts in Technical Trading Systems (1978). It produces a per-asset volatility number in the same units as the price.

The formula

Stop distance = N × ATR(k)

Where N is a multiplier (typically 1-3) and k is the ATR lookback. Then plug into the main sizing equation:

Size = (Account × Risk%) ÷ (N × ATR × Value per unit)

Worked example – multi-asset

• Account: $100,000, Risk: 1% = $1,000
• N = 2 (i.e. stop at 2× ATR)
• EUR/USD: ATR(14) = 60 pips → stop = 120 pips → size = $1,000 ÷ (120 × $10) = 0.83 standard lots
• GBP/JPY: ATR(14) = 90 pips → stop = 180 pips → size = $1,000 ÷ (180 × $6.70) = 0.83 standard lots
• BTC/USD: ATR(14) = $1,200 → stop = $2,400 → size = $1,000 ÷ ($2,400 × 1) = 0.42 BTC

All three trades carry the same dollar risk, but the position sizes adjust naturally to each asset's volatility. ATR is the single most important adjustment for traders running multi-asset portfolios.

John L. Kelly Jr. published "A New Interpretation of Information Rate" in the Bell System Technical Journal in 1956. The Kelly criterion gives the bet fraction that maximizes expected log-growth of capital for a given edge:

f* = (p × b − q) ÷ b

Where p = probability of winning, q = 1 − p, and b = odds (win size ÷ loss size). For trades with R-multiple payoffs:

f* ≈ p − (q / R)

Example: a strategy with 50% win rate and 2R average winners:

f* = 0.50 − (0.50 / 2) = 0.25

Full Kelly would size at 25% of capital per trade. This is mathematically correct if your estimates of p and R are perfect. They never are, for three reasons:

1. Sample size: 100 trades is not enough to estimate a 50% win rate to 5% accuracy.
2. Regime change: historical p and R do not necessarily predict future p and R.
3. Fat tails: Kelly assumes a finite variance; real trading has occasional extreme losers.

Because of this, practitioners use fractional Kelly: typically 1/4 to 1/2 of full-Kelly. For the 0.25 full-Kelly above, 1/4 Kelly = 6.25% per trade – still too high for most strategies in practice.

Where Kelly is useful

• As a sanity check: if your strategy's Kelly is negative, your edge estimate is wrong or negative. Stop trading it.
• As an upper bound: never risk more than full Kelly; even 1/2 Kelly produces uncomfortable drawdowns.
• For systematic strategies with long, stable live histories and tight regime assumptions.

Why not use it directly

Because drawdowns at fractional Kelly sizing are still severe. At 1/2 Kelly on a strategy with Sharpe 1.0, the expected maximum drawdown over a typical trading career easily exceeds 50%. Most traders cannot psychologically or operationally survive this, regardless of whether the math says they should.

The "value per unit" input is where most position-sizing errors originate. A quick reference:

Forex pip value (for USD account)

• USD-quoted pairs (EUR/USD, GBP/USD, AUD/USD): pip value = $10 per standard lot (100,000 base currency units).
• JPY-quoted pairs (USD/JPY, EUR/JPY, GBP/JPY): pip value ≈ $1,000 ÷ current rate. At 150.00, that's ≈ $6.67 per standard lot.
• Non-USD base (GBP/USD, AUD/USD): pip value in account currency = pip value in quote currency × (account_currency ÷ quote_currency rate).
• Cross pairs without USD (EUR/GBP, EUR/CHF): requires a two-step conversion through either USD or EUR.

Futures point values (common CME)

• ES (E-mini S&P 500): $50 per index point per contract.
• MES (Micro E-mini S&P 500): $5 per point.
• NQ (E-mini Nasdaq): $20 per point.
• MNQ (Micro E-mini Nasdaq): $2 per point.
• CL (WTI Crude Oil): $10 per handle ($1 per 0.01 per contract of 1,000 barrels).
• GC (Gold): $100 per dollar move per contract of 100 troy oz.
• 6E (Euro FX): $12.50 per 0.0001 move per contract of 125,000 EUR.

Crypto

• Spot: value per unit = current price in quote currency.
• Perpetual: same, but with the funding rate as an additional holding cost.
• CFD: check the broker's contract spec – some CFDs have $1 per pip, some have $10, some have variable multipliers.

The single most common mistake

Using a EUR/USD position-size calculator for a USD/JPY trade. The formula looks the same, but pip value is different, so the output is wrong by a factor of ~1.5x. Always check the quote currency before clicking enter.

Ten errors, each of which costs someone a blown account every week.

1. Pasting a EUR/USD pip value into a USD/JPY trade. Always recalculate pip value per quote currency.
2. Confusing "0.1 lot" on MT4 (mini lot = 10,000 units) with 0.1 lot on some brokers that mean "fractional standard lot." Check the broker's lot definition.
3. Ignoring commission in the risk budget. Round-trip commission on 2 standard lots at $5/lot is $20 – subtract from the stop-loss budget.
4. Using the calculator output without rounding correctly for the instrument. Futures: round down. Forex: to 2 decimals (0.01 lot). Crypto: to the exchange's min-order size.
5. Updating the account value only weekly, not after each trade. Compounding errors drift the real risk% over time.
6. Using leverage-based sizing instead of risk-based sizing. Leverage is an output, not an input.
7. Forgetting that multiple correlated positions share risk. EUR/USD long + GBP/USD long is largely the same risk twice; size each at half the budget.
8. Scaling down risk during a drawdown but forgetting to scale it back up. After recovery, re-enter at normal sizing – the drawdown-sized positions drag the recovery.
9. Applying ATR sizing without accounting for news-induced spikes. A 2x ATR stop on a news day is still inside the release spread.
10. Over-fitting the "optimal" risk% to historical backtests. The curve that maximizes past returns is usually not the one that survives future drawdowns.

A quick sanity check that catches most of these: before clicking Buy/Sell, say out loud: "if my stop is hit, I lose exactly X dollars, which is Y% of my account." If you can't answer cleanly, recalculate.

A compact workflow for running fixed-fractional + ATR sizing through an evaluation:

1. Choose a base risk%: 0.5% for evaluations with tight drawdown caps (5-10%), 1% for looser caps.
2. Choose an ATR multiplier: N=2 for trend-following, N=1 for mean-reversion, N=3 for swing.
3. Open the platform's built-in calculator or a reputable third-party calculator (Myfxbook, Babypips, ForexFactory). Verify the calculator matches your account currency and broker lot definition.
4. Input: account, risk%, stop pips/points, pip/point value for the specific instrument.
5. Round the output down to the instrument's minimum increment.
6. Say the sanity check out loud: "if stopped out, I lose $X, which is Y% of account."
7. Place the entry order with the stop and take-profit attached simultaneously. No "I'll set it later."
8. Journal the intended R and the realized R. Deviation is the training signal for the next session.

This process takes 60-90 seconds per trade. That's acceptable overhead for active swing / day trading. For scalpers placing dozens of trades per hour, pre-compute the sizing for 4-5 standard scenarios (typical stop distance × instrument) and use that table during the session – recomputing on each trade is infeasible.