The Kelly Criterion

The position-sizing lesson said how much you bet is a matter of survival, and promised a way to make the optimal size mathematically precise. The Kelly criterion is that answer. Developed by John Kelly at Bell Labs and used by well-known investors and gamblers, it gives a formula for sizing a series of bets to maximize the long-term growth of capital, tying together expected value, compounding, and risk of ruin.

What Kelly optimizes

The Kelly criterion prescribes betting an amount proportional to your edge: more when your advantage is larger, less when it is smaller, and nothing when you have no edge. Crucially, it does not maximize the expected value of any single bet. It maximizes the long-term geometric growth rate of wealth, the rate at which capital compounds over many repeated bets. A strategy that maximizes single-bet expected value can still ruin you through over-betting, whereas Kelly optimizes the growth of wealth across a long sequence, accounting for how gains and losses compound.

The formula in symbols

For a simple even-money bet, where b equals 1, this simplifies to f = p − q, which is the same as 2p minus 1. So with a 60 percent chance on an even-money bet, you would stake 2 times 0.6 minus 1, which is 0.2, or 20 percent of your capital.

Why over-betting destroys growth

The main insight connects straight to position sizing and risk of ruin. Even with a genuine edge, betting too much on each opportunity leads to ruin rather than riches, because losses along the way compound destructively and can wipe out your capital before the edge expresses itself. Betting too little leaves growth on the table. Kelly finds the precise fraction that balances these two failures, aggressive enough to harness the edge but conservative enough to survive the losses. It is the mathematical form of the survival-enables-compounding principle.

Bet too much and even a winning edge ruins you, bet too little and you crawl. Kelly finds the fraction that compounds fastest while surviving.

Full Kelly and fractional Kelly

Betting the full Kelly amount is aggressive and produces large, stomach-churning swings with deep drawdowns. So many practitioners use fractional Kelly, betting some portion, such as half. Fractional Kelly gives up a little of the theoretical maximum growth in exchange for substantially smoother performance and shallower drawdowns, a trade most investors gladly make, especially given the drawdown asymmetry from earlier. A smaller bet also provides a margin of safety against the estimation errors discussed next.

The critical limitation: you must know your edge

Here is the catch that makes Kelly risky to apply naively in markets. The formula needs accurate estimates of your edge and the odds, and in investing these are rarely known with any precision. Your true edge is usually unknown and easy to overestimate, and overestimating it leads directly to over-betting, which Kelly itself shows causes ruin. The math is precise, but only as reliable as its inputs, and market inputs are deeply uncertain. This is the garbage-in, garbage-out principle from Unit 5 applied to bet sizing, and it is a central reason practitioners favor conservative, fractional Kelly over trusting an overconfident estimate of their advantage.

Kelly formalizes bet sizing, but its inputs, and every model in this unit, can be defeated by one thing: the investor's own psychology. The next lesson confronts the behavioral pitfalls that undo good risk management.