Wednesday, April 19, 2006

The Kelly Criterion

Investing has often been compared with Poker, in that in both environments it is key to only play when the odds are in your favor.

The Kelly Criterion, explained here by My 1st Million at 33, is an equation that defines how much to risk for each level of winning probability and magnitude of returns.

Its the first time I have seen someone quantify this principle. Essentially the amount you should bet is based both on your odds of having a positive result and the average return that you can expect.

It speaks directly to the second rule of successful investing, which is: Focus your efforts. Put all your eggs in the few best baskets, and watch those baskets VERY CAREFULLY.

Generally, I try to follow the advice of Charlie Munger, and place large bets on high probability events. How much is a "large bet"? Buffett and Munger suggest that you need to be prepared to put at least 10% of your capital on a high probability event. Less than 10% and the positive impact of a terrific investment will be diluted by your next best ideas. Like a pitcher - you don't want to get beaten on your 2nd or 3rd best pitch. When the game is on the line - you want your best ideas working hard for you.

Philip Fisher, in Common Stocks, Uncommon Profits, suggests that holding more than 15 common stocks is a sign of financial incompetence - it is a sign that you cannot make investment decisions, for holding a smorgasbord of stocks is a way of avoiding decision making.

But what about diversification? Most evidence suggests that the vast majority of the benefit that diversification offers for investing returns comes from the first five or ten stocks. Thus it is possible to be both FOCUSED and DIVERSIFIED at the same time.


  1. Great post! I'm a big believer in only holding a small basket of stocks, because the last thing I want to do is diversify myself out of gains.

    The argument I can see for diversification though, is more for people that don't really understand investing. When someone who has no real financial knowledge asks me what to invest in, I'll always recommend looking into managed funds.

    The small number of holdings works when you know what you're doing. I guess the way to diversify that is to have each of your companies in different sectors.

  2. I agree with you, that it is important to have stock in companies with low correlation in returns.

    When people ask me about investment selection (and it is obvious that they only want a prescription, and not understanding), I also tell them to invest as broadly as possible.

    I generally do suggest more financial education as well, so they can make better decisions, but I have never been comfortable with people who really don't want to know what they are doing. Is helping them find a short-cut, albeit a poor one better than no advice at all? I am not sure.