How Big to Bet: Position Sizing and the Kelly Criterion

Most people put all their energy into "what to buy" and give almost no thought to "how much to buy." But the Kelly criterion and countless real-world experiences point to the same conclusion: position sizing affects long-term outcomes no less than stock selection itself.

What the Kelly Criterion Says

Given a known win probability and payoff odds, the Kelly criterion yields the bet fraction that maximizes long-term compound growth. Its intuition is very simple: the greater your conviction and the more favorable the odds, the heavier you should bet; otherwise, the lighter. It translates "confidence" into "position size."

• High win rate, good odds → large position (but still capped).
• Low win rate or poor odds → small position, or none at all.
• No edge (negative expectancy) → Kelly's answer is: bet 0, meaning don't play.

Why You Should Use a Fraction in Practice

Full Kelly is extremely volatile and highly dependent on your estimates of win rate and odds—and our estimates of the future are almost always overconfident. So professional investors typically use half-Kelly or an even more conservative fraction: sacrificing a little theoretical growth rate in exchange for a sharply lower drawdown and a higher probability of surviving.

“Rule number one: never lose money. Rule number two: never forget rule number one.”— Warren Buffett

Risk of Ruin: The Line You Must Never Cross

This is the most important psychological discipline in position sizing: no matter how attractive an opportunity looks, never bet a size where a single wrong call could permanently take you out. Because investing is a repeated game—as long as you remain in it, you have a chance to recover; once you go to zero, no win rate, however high, is any use to you. Controlling bet size is fundamentally about protecting your eligibility to keep participating in compounding.

How It Maps to Our Engine

Our platform uses the Return Spread (the gap between expected return and required return) to quantify the "odds" of an opportunity, and the report requires an explicit position size and scenario probabilities rather than a single, isolated target price. This is Kelly thinking in practice: first make clear how good the odds are and how strong your conviction is, so that the position size has a basis—rather than sizing up on gut feel.