Nov 1st 2016, 16:10 by Buttonwood
THERE are few sure things in investing. But the chance to bet on a rigged coin sounds like a good one. Alas, a paper by two fund managers, Victor Haghani of Elm Partners (and co-founder of the collapsed Long-Term Capital Management) and Rich Dewey of Pimco, shows that it is possible to get even that wrong.
The paper invited 61 people, a combination of college-age students in finance and economics and some young professionals at finance firms (including 14 who worked for fund managers), to take a test. They were each given a stake of $25 and then asked to bet on a coin that would land heads 60% of the time. The prizes were real, although capped at $250.
Remarkably, 28% of the participants went bust, and the average payout was just $91. Only 21% of the participants reached the maximum. 18 of the 61 participants bet everything on one toss, while two-thirds gambled on tails at some stage in the experiment. Neither approach is in the least bit optimal.
Apparently the right strategy is to use the Kelly criterion, named after a researcher at Bell labs. Based on the odds in this experiment, the right approach would be to bet 20% of your pot on each throw. So if you lose, you cut the size of your bet; if you win, your stake increases.
Your blogger had a go himself, and being a cautious Yorkshireman, never bet more than 10% of the pot. But I got to the $250 with a few minutes to spare. It is a frustrating process at first; literally three steps forward and two back. There was more than one run of four tails. You can try it yourself here (there is no prize money; this is just for intellectual interest).
Is this a fair test of the real world? Well, real money was involved and the people being tested should have had a greater knowledge of finance than most. But even smart people can be blindsided by dumb luck.