The bet size that maximises long-run growth is proportional to edge divided by odds; betting more reduces growth.
Kelly's 1956 formula from information theory, adopted by blackjack teams and hedge funds. Its practical lesson is asymmetric: betting above the Kelly fraction reduces long-run growth and eventually guarantees ruin.
In practice
- A trader with a genuine 2% edge who stakes a quarter of capital per trade goes broke faster than one with no edge who stakes 1%.
- A firm with a real advantage in one market that commits its whole balance sheet to it does not survive the first bad year.
- Further reading
- Wikipedia
- Subject
- Economics & Incentives
- Essay
- Rules for choosing, and the paradoxes that humbled the theory