Kelly Criterion Calculator (with Half & Quarter Kelly)

The Kelly Criterion tells you the mathematically optimal fraction of your account to risk on each trade. Enter your win rate and average win/loss size — we'll calculate the full Kelly percentage, the more conservative half-Kelly and quarter-Kelly fractions most pros actually use, and your recommended dollar risk per trade. The formula is f = W − ((1 − W) ÷ R).

Your inputs

%

Percentage of trades that finish profitable. Be honest — most retail strategies are 40-55%.

$

Average $ amount you make on winning trades (or any unit — we use the ratio, not the absolute number).

$

Average $ amount you lose on losing trades. Enter as a positive number.

$

Total trading capital. Enter 0 to skip the dollar-risk outputs.

Results

Kelly fraction

32.50%

The mathematically optimal % of account to risk per trade. Theoretical maximum growth rate.

Half-Kelly

16.25%

What most pros actually use — same expected return, ~75% less drawdown variance.

Quarter-Kelly

8.13%

Conservative. Slower growth but psychologically survivable during losing streaks.

Recommended risk per trade (full Kelly)

$8,125.00

Based on your account size + full Kelly fraction.

Recommended risk per trade (half-Kelly)

$4,062.50

The more livable number for most retail traders.

Expected value per trade

$2,640.63

Avg per-trade profit at half-Kelly sizing. Negative = your edge is too thin to size up.

Results update live as you change inputs. This calculator runs entirely in your browser — your numbers are never sent to a server.

Worked example

You win 55% of the time, your average winner is $200, your average loser is $100. Win/loss ratio R = 2. Plugging in: Kelly = 0.55 − (0.45 ÷ 2) = 0.325 = 32.5%. That's the theoretical max — but full Kelly is brutally volatile (you'd routinely see 50%+ drawdowns). Half-Kelly = 16.25% is what most pros use: same long-run expected return, far smaller drawdowns. At a $25,000 account, half-Kelly means risking ~$4,062 per trade.

Frequently asked questions

What is the Kelly Criterion?

The Kelly Criterion is a formula developed by John Kelly at Bell Labs in 1956 that tells you the mathematically optimal fraction of your bankroll to bet (or risk on a trade) to maximize long-term geometric growth. It accounts for both your edge (win rate) and your odds (win/loss ratio).

Why do most traders use half-Kelly instead of full Kelly?

Full Kelly maximizes expected growth but also maximizes volatility. It routinely produces 50%+ drawdowns that are psychologically (and often financially) unsurvivable. Half-Kelly captures about 75% of full Kelly's expected return with roughly 25% of the variance — a much better risk-adjusted outcome for most traders.

What if Kelly tells me to risk 50% of my account per trade?

That means your inputs are either very optimistic or your edge is genuinely huge. In practice, never risk more than 1-2% of your account per trade regardless of what Kelly says — even half-Kelly on a great strategy rarely exceeds 5%. Use Kelly as a directional signal (am I undersizing?), not an absolute prescription.

What if my Kelly result is 0% or negative?

A 0% or negative Kelly means your win rate and win/loss ratio combine to give you no statistical edge — you'd lose money in the long run trading this strategy. Either your inputs are wrong (small sample size?), or the strategy isn't profitable and you should stop trading it.

Should I use Kelly for swing trades, day trades, or both?

Kelly works for any series of binary win/loss bets, including both swing and day trades — but only if your inputs are reliable. You need at least 30-50 closed trades to have a meaningful win rate estimate, and your avg win / avg loss should be measured on similar setup types, not pooled across strategies.

How is Kelly different from fixed-percent position sizing?

Fixed-percent (e.g., 'always risk 1%') treats every trade as equally important. Kelly adjusts position size based on your statistical edge — bigger when your edge is bigger, smaller (or zero) when it isn't. Kelly is more aggressive when you have a real edge and more defensive when you don't, but it requires accurate inputs.