How Finance Executives Aren’t Like Poker Players, and Why it Means We Should Be More Skeptical of Their Expertise

Imagine nine poker players seated at the final table of a major tournament. Although they posses a variety of different skillsets and backgrounds, there is one thing they have in common. They have all had the luck necessary to win the majority of the many 50-50 hands that arise during a tournament. Nobody reaches a final table by only gambling as an overwhelming 80-20 favorite.

Now imagine nine random Presidents/CEOs of large finance firms or banks. In some sense, they are also finalists in a tournament. Hordes of smart young people enter the finance industry each year, but only a select few reach the top. Like the pokers players at the final table, the CEOs have benefitted from a perfect combination of skill and luck. Nobody is repeatedly finding risk-free piles of money lying around in a competitive market, and that means each CEO must have successfully gambled on at least a few big ideas. Even if the CEOs repeatedly found major opportunities with a 70% chance of success, they never would have reached the top without beating the odds that they would fail.

Although poker players and finance CEOs rely to some degree on luck, the perception exists that we have identified the ones who are the most skilled. This is where there is a big difference between the two groups. Evaluations of poker players are based on hundreds of thousands of hands, not the outcome of a few tournaments. I won’t venture a guess as to how many risky decisions a finance CEO makes during his first 25 years of employment, but I’m confident the number has substantially less descriptive and predictive power than a poker player’s vast history. Finance CEOs were clearly wise with many gambles — that’s why they are where they are — but the small sample size means we can’t ignore the fact that some of them are CEOs instead of senior managers because they lucked out on bad gambles or a long string of coin flips.

All of this is to say that when we think about the expertise of finance bigwigs we should be more mindful of “outcome bias,” the tendency to focus on the outcome of a decision rather than the decision itself. Imagine LeBron passes up an uncontested dunk but seconds later makes a contested three pointer. Did he make a good decision? The answer is obviously no, but many people might say yes because they fail to understand that good outcomes don’t retroactively create good decisions.1  For example, one recent study found that ethically questionable behavior produces less condemnation when it leads to a positive outcome, even if the outcome is entirely due to chance. In their pioneering work on the issue, Jonathan Baron and John Hershey found that when people were asked about the quality of a medical decision, they said the thinking behind the decision was better when they were told the decision led to a positive outcome.

The lesson is that if a Wall Street CEO rises to his position by making a string of profit-growing decisions, the decisions may not have all been wise given the conditions at the time. That doesn’t mean Jamie Dimon isn’t a smart man who knows a lot about the American financial system, but it does imply that even in a vacuum (i.e. no consideration of political or financial motivation) we should be marginally more skeptical of the the expertise of influential finance people.

In the early days of the internet there was a classic scam where somebody claiming to be an NFL gambling expert would send 800 people a free pick during the first week of the season. The catch is that half the people were told a certain team would cover the spread and the other half were told the team’s opponent would cover the spread. The same thing was repeated the following week among the 400 people who initially received the winning pick. After four weeks 50 people had received four consecutive winning picks, and at that point the scammer asked them to pony up $49.95 to subscribe to the picks for the rest of the season.

The beauty of the scam is that 1/16th of your initial sample will always get the four good picks. Similarly, in a poker tournament nine people will always get lucky enough to make the final table. When it comes to finance, some set of people will always rise to the top of our largest institutions — after all, the banks can’t be leaderless. But when these leaders are annointed, it’s important to be mindful that in an arena full of unpredictability, a rise to the top is not always as indicative of ability as we might think.

1One highlight of my never-heralded sportswriting career was finding a Washington Redskins message board thread that used the outcome bias to take-way-too-seriously a joke I made about Joe Gibbs’ poor decision making. Skins fans need to learn that when Gibbs plays it too conservatively in overtime by kicking a 39-yard field goal on 1st down, the fact that the kick went in doesn’t make it good coaching.
Gino, F., Moore, D.A. & Bazerman, M. (2009). No Harm, No Foul: The Outcome Bias in Ethical Judgments SSRN Electronic Journal DOI: 10.2139/ssrn.1099464

Baron, J. & Hershey, C. (1988). Outcome bias in decision evaluation Journal of Personality and Social Psychology DOI: 10.1037/0022-3514.54.4.569

4 Responses to How Finance Executives Aren’t Like Poker Players, and Why it Means We Should Be More Skeptical of Their Expertise

  1. You’re very mistaken here:

    “They have all had the luck necessary to win the majority of the many 50-50 hands that arise during a tournament.”

    There are pivotal hands in poker tournaments like these, but because they win or lose these 50/50 hands doesn’t guarantee that they will appear or not appear at the final table. You are confusing probability and variance with luck. There are a lot more factors involved other than 50/50 flips. For example, I player with 20,000 chip can lose a 50/50 flip to a player that has 5,000.

    In this statement:

    “Nobody reaches a final table by only gambling as an overwhelming 80-20 favorite.”

    One is not gambling when they are ahead “80-20 favorite” they would be gambling if they were the 20% player of that statement.

    Like great CEOs, great poker players know when to ‘gamble’ and when not to… the preponderance of luck should not be misconstrued for for probability. Good decisions will prevail for both CEOs and poker players in the long run. A CEO and a poker player will never be able to consistently make ‘that final table’ if they didn’t possess knowledge.

    • erichorowitz says:

      Obviously not everybody who wins important coinflips will make the final table, but if you happen to make a final table you’ve probably won more than your EV on important coinflip hands. Probably a better poker analogy to use would have been one where you’re blind to everything but the outcome. For example, say you put somebody on a range of relatively weak hands and decide to bluff. It turns out they have a stronger hand beyond that range, but for whatever reason they decide to fold anyway. Is that skill or luck?

      My broader point is that there’s enough data on poker that in the long run our evaluations of players mitigate the influence of outcomes not based on skill. I think in finance there just aren’t enough major decisions being made to do as good a job isolating that influence. It’s clear, and I say so in the post, that people at the tops of finance institutions are talented. But even if somebody got there by finding a bunch of 80-20 propositions, I would bet that they ended up winning 85 or 90 percent of them.

  2. Pingback: Thursday 7atSeven: losing their nerve | Abnormal Returns

  3. Misaki says:

    People don’t vote for finance executives… I think the reasoning works just as nicely with politicans though.

    It is difficult to know what to do with this information, but as a way of giving an incentive to perform well I don’t think it’s that bad for the amount of information and processing it requires.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s