Kelly criterion
A formula for sizing bets to maximize long-run growth while avoiding ruin; intuitively, never risk so much that a loss wipes you out.
As defined in Nassim Taleb - The #1 Rule to Not Be a Sucker (The Ergodicity Principle)
Other definitions
A betting strategy that sizes wagers as a fraction of current wealth—increasing bets when winning and decreasing them when losing—to grow capital while avoiding ruin.
Nassim Nicholas Taleb on Rationality, Risk, and Skin in the Game
Where it comes up · 12
- #93 Ed Thorp (A Man for All Markets)5 claims
- Nassim Taleb - The #1 Rule to Not Be a Sucker (The Ergodicity Principle)5 claims
- #222 Ed Thorp (My personal blueprint)3 claims
- 285 | Nate Silver on Prediction, Risk, and Rationality2 claims
- A conversation between Nassim Nicholas Taleb and Stephen Wolfram at the Wolfram Summer School 20211 claim
- #365 Nick Sleep’s Letters: The Full Collection of the Nomad Investment Partnership Letters to Partners1 claim
- #92 Ed Thorp and Claude Shannon1 claim
- #108 Jim Simons (Money Printer)1 claim
and 4 more
Most important claims · 10 of 23Search all claims for “Kelly criterion” →
- If one only maximizes expected value, then a bet that doubles the world's utility with 50%+epsilon probability while destroying it with 50%-epsilon is positive-EV and should be taken—a conclusion SBF endorsed (telling Caroline Ellison and Tyler Cowen he'd take such a bet)—but almost anyone recognizes this as irrational, showing rationality must value lowering the risk of complete annihilation.285 | Nate Silver on Prediction, Risk, and Rationality
- Thorp treated hedge bets like blackjack hands with a known advantage: he estimated expected return and risk, then chose how much of his bankroll to bet, and with no betting limit on Wall Street he made risk reduction central to his investing.#93 Ed Thorp (A Man for All Markets)
- Given a favorable bet, the Kelly system scales each wager to the current bankroll, so the bettor can never run out of money, and because bets grow with a growing bankroll, it produces the fastest growth of any betting system when there is a genuine edge.#92 Ed Thorp and Claude Shannon
- Finding an edge is the easy part; the hard part is capturing it by dosing bets correctly, neither too little nor too much, which determines outcomes.#93 Ed Thorp (A Man for All Markets)
- The Kelly criterion is a mathematical formulation of the simple concept 'don't risk everything' — bet carefully each time so you never lose the whole kitty — and it helps avoid ruin, making risk-taking calculated across large populations not always rational for an individual.Nassim Taleb - The #1 Rule to Not Be a Sucker (The Ergodicity Principle)
- Being right in economic analysis is not enough; without proper evaluation of leverage risk, even a correct thesis can lose money, which Thorp made a central theme of his life, and the 2008 financial establishment failed the same lesson by over-leveraging.#93 Ed Thorp (A Man for All Markets)
- Arithmetic averaging does not work for compounding processes: when one gambler goes bankrupt (returns zero), adding that zero to an arithmetic average barely impacts the result, but in a time-sequential process with the same person, that one bankruptcy event eliminates all future returns. For arithmetic averages to match continuous compounding behavior, one must bet in log-sizes according to the Kelly criterion.Nassim Taleb - The #1 Rule to Not Be a Sucker (The Ergodicity Principle)
- Using the Kelly Criterion, if one is certain of being right one should invest the entire portfolio in that idea, and since almost no investors do this, Nomad's portfolio concentration was likely too low.#365 Nick Sleep’s Letters: The Full Collection of the Nomad Investment Partnership Letters to Partners
- A bet with favorable odds (e.g., 55/45) is not necessarily rational to keep taking: if you repeatedly take such a bet over a lifetime without following a survival strategy like the Kelly criterion, you will eventually go bankrupt even with fair odds, because you are subject to a time average and cannot survive a ruin event.Nassim Taleb - The #1 Rule to Not Be a Sucker (The Ergodicity Principle)
- You should never go all in on any bet no matter how good it appears; the Kelly Criterion formalizes betting only a fraction of wealth to keep playing long enough to reap the expected average, but Harry Crane's critique is that Kelly is too aggressive because it assumes you truly know the odds, which in real life requires playing many times to verify, justifying even smaller fractions.Luca Dellanna on Risk, Ruin, and Ergodicity