Edgoro
Risk · Free tool

Kelly Criterion Calculator

Get the Kelly fraction from your win rate and payoff ratio, compare full, half and quarter Kelly, and cap risk per trade.

Fraction of Kelly to apply
Full Kelly: share of equity to risk
16.67%
Payoff ratio (avg win ÷ avg loss)1.50
Half Kelly8.33%
Quarter Kelly4.17%

Your risk per trade

Suggested (half Kelly, capped)2.00%
Risk amount$200.00
Long-run growth kept vs full Kelly23%
Drawdown after 10 losses in a row18.3%

Edge

Expectancy per trade+0.25R
Break-even win rate40.0%

Kelly = W − (1 − W) ÷ payoff ratio. It treats your win rate and payoff as exact; they are estimates, so most traders apply a fraction and a hard cap.

Trading involves substantial risk of loss and is not suitable for every investor. Edgoro is a software tool, not a broker or financial adviser: nothing it produces is investment advice, and past or backtested performance does not guarantee future results.

What the Kelly criterion calculator does

The Kelly criterion, published by John Kelly at Bell Labs in 1956, gives the share of capital to stake on a repeated bet with a known edge so that the bankroll grows as fast as possible over the long run. In trading, the stake is what you lose if the stop-loss is hit: your risk per trade. The calculator turns your win rate and payoff ratio into the Kelly fraction, shows half and quarter Kelly beside it, and converts the fraction you choose into a risk per trade and a dollar amount, with an optional cap.

How to use it

  1. Enter your win rate in percent, taken from a journal or a backtest with a meaningful number of trades.
  2. Enter your average win and average loss in the same unit — dollars, pips or R. Only their ratio, the payoff ratio, matters.
  3. Enter your account balance and the maximum risk per trade you accept. The cap protects you when Kelly asks for more than you would ever risk.
  4. Choose the fraction of Kelly to apply: full, half or quarter.

The panel shows full, half and quarter Kelly, your suggested risk per trade after the cap, the amount in dollars, the share of the full-Kelly growth rate you keep, and the drawdown ten losses in a row would cause at that risk.

The formula

payoff ratio   b  = average win ÷ average loss
Kelly fraction f* = W − (1 − W) ÷ b
growth per trade  g(f) = W × ln(1 + f × b) + (1 − W) × ln(1 − f)

W is the win rate as a decimal. f* is the risk per trade, as a share of equity, that maximises g(f), the expected logarithmic growth of the account per trade. If f* is zero or negative, the strategy has no edge on these numbers and Kelly says to stay out.

Worked example

A strategy wins 50% of the time and its average win is 1.5 times its average loss:

  • Payoff ratio b = 1.5
  • Full Kelly: f* = 0.50 − 0.50 ÷ 1.5 = 0.50 − 0.333 = 16.67% of equity per trade
  • Half Kelly: 8.33%; quarter Kelly: 4.17%
  • Expectancy: 0.5 × 1.5 − 0.5 = +0.25R per trade; break-even win rate 1 ÷ 2.5 = 40%

On a $10,000 account, half Kelly means risking $833 on the next trade. With a 2% cap, the suggested risk becomes $200. What each level keeps of the maximum growth rate, and what ten losses in a row would cost:

Risk per tradeShare of full-Kelly growthDrawdown after 10 straight losses
16.67% (full Kelly)100%83.8%
8.33% (half Kelly)75%58.1%
4.17% (quarter Kelly)44%34.7%
2% (cap)23%18.3%
33.3% (double Kelly)0%98.3%

At double Kelly the growth rate falls to zero: the account goes nowhere in the long run despite a positive edge, and anything above that shrinks it.

Why most traders use fractional Kelly

Full Kelly is growth-optimal only when its inputs are exact, and in trading they never are:

  • Estimates are noisy. A win rate measured on 100 trades can easily be five to ten points away from the true one. Overestimate the edge and full Kelly becomes overbetting.
  • The penalty is asymmetric. Half Kelly keeps about three quarters of the growth rate; double Kelly loses all of it. Erring on the small side costs little, erring on the large side costs a lot.
  • Drawdowns are deep. In the classic continuous-time model, a full-Kelly bettor has a 50% chance of seeing the account halve at some point; at half Kelly that falls to 12.5%, at quarter Kelly to under 1%.
  • Trades are not independent draws. Correlated positions, changing market regimes and losses beyond the stop (gaps, slippage) all make real risk larger than the formula sees.

This is why Kelly is best treated as an upper bound rather than a target. Many traders size from a fraction of it and cap the result, often somewhere between 0.5% and 2% per trade. See what your chosen risk does over a long series of trades with the risk of ruin calculator and the Monte Carlo trading simulator, then turn it into lots with the position size calculator.

Common mistakes

  • Treating backtest numbers as exact. Costs, slippage and curve-fitting inflate a backtested edge. Feed Kelly conservative estimates, or the averages from a live journal.
  • Mixing units. The average win and average loss must use the same unit, and the loss is entered as a positive number.
  • Applying Kelly to each of several correlated trades. Three positions on the same theme behave like one large bet, so their combined risk should respect the fraction.
  • Forgetting it is a share of current equity. After a loss the dollar risk shrinks with the account — that is part of the protection, not a mistake to correct.

Frequently asked questions

What is the Kelly criterion in trading?

It is a formula for the share of capital to risk on each trade so that the account grows as fast as possible in the long run: win rate minus loss rate divided by the payoff ratio. In trading, the share is the amount lost at the stop, as a fraction of equity.

Is full Kelly too aggressive for trading?

For most traders, yes. It assumes your win rate and payoff are exact, it produces deep drawdowns, and it turns into overbetting as soon as the edge is overestimated. Half or quarter Kelly with a hard cap is the more common choice.

Why does half Kelly keep about 75% of the growth?

The growth rate is a curve that peaks at the Kelly fraction and falls slowly on the low side. Halving the stake keeps roughly three quarters of the maximum growth rate while cutting volatility and drawdowns far more.

What does a negative Kelly percentage mean?

The strategy has no edge on the numbers you entered: the expected result per trade is zero or negative, so the growth-optimal stake is nothing. Improve the win rate or the payoff before sizing anything.

Can I use the Kelly criterion on a prop firm account?

Kelly ignores daily and maximum loss limits, which usually force a much smaller risk per trade than it suggests. Size from the firm's rules first, then use Kelly only as a check that your strategy has an edge.

Related guides

All guides
All free tools

Let AI build the indicator behind your numbers.

Turn the rules you just calculated into a TradingView indicator, an MT5 bot or a crypto bot. Instant access, with a 7-day money-back guarantee.