Blackjack, Banking, and the Risk of Not Taking Enough Risk

I was in Canada this week giving a talk and I must have had a half-dozen Canadians ask, nay demand, that I flatter them immediately about the solidity of their banking system. Contrary to the crazy banks in the U.S., lending millions of dollars to random homeless people despite only teensy pieces of quartzite as collateral, Canadian banks oozed conservatism, not even lending money to people who could pay it back, I was repeatedly reminded.

And it’s true, of course. The Canadian banking system didn’t take anywhere near as many risks as the U.S. one did over the last few years, and that did lead to grotesque excesses in the U.S. market. It was partly structural, partly psychological, and partly cultural, but the effect was the same: The Canadian banking system left this episode stronger than how it entered it.

But let’s not get things backwards. Well-timed conservatism is a fine thing, but there is something to be said for taking risks. Among the reasons why Bill Gates dropped out of university and started Microsoft and made billions when you know you’re the same age as he is and you know he’s not really that smart and Windows sucks and you can’t stand the guy, is that he took a risk and you didn’t. Elevating uber-conservatism into the highest virtue is no path to growth and wealth and all good things capitalism.

There is a new paper out on the subject from NBER, and its abstract follows:

Fear and Loathing in Las Vegas: Evidence from Blackjack Tables

Bruce Ian Carlin, David T. Robinson

Psychologists study regret primarily by measuring subjects’ attitudes in laboratory experiments. This does not shed light on how expected regret affects economic actions in market settings. To address this, we use proprietary data from a blackjack table in Las Vegas to analyze how expected regret affects peoples’ decisions during gambles. Even among a group of people who choose to participate in a risk-taking activity, we find strong evidence of an economically significant omission bias: players incur substantial losses by playing too conservatively. This behavior is prevalent even among large stakes gamblers, and becomes more severe following previous aggressive play, suggesting a rebound effect after aggressive play. [Emphasis added]