Bankers vs. Athletes

By James Kwak

Bill George, a director of Goldman Sachs (GS), defending the bank’s compensation practices, said this:

“The shareholder value is made up in people and you need the people there to do the job. If you don’t pay them for their performance, you’ll lose them. It’s much like professional athletes and movie stars.”

The idea that the level of inborn talent, hard work, dedication, and intelligence you need to be a banker is even remotely comparable to that of, say, NBA basketball players is ridiculous. But leaving aside the scale, there are some similarities. Most obviously, athletes on the free market–those eligible for free agency–are overpaid. John Vrooman in “The Baseball Players’ Labor Market Reconsidered” (JSTOR access required) goes over the basic reasons, but they should be familiar to any sports fan. There is the lemons problem made famous by George Akerlof: if a team gives up a player to the free agent market, it probably has a reason for doing so. There is the winner’s curse common to all auctions: estimates of the value of players follow some distribution around the actual value, and the person who is willing to bid the most is probably making a mistake on the high side.

Another common factor is the mistake general managers make in overpaying for luck. Take any group of .265 hitters, give them 450 at bats, and a handful will hit .300. On the free agent market, they will be paid like .300 hitters, especially if they are young and they do not have a long history of hitting .265 behind them. This is the exact same as one bank making a huge offer to a trader from another bank who just had a great year.

For another, the apparent productivity of a player in a team sport is largely due to his team and cannot simply be reproduced individually. John Vrooman in “The Baseball Players’ Labor Market Reconsidered” cites the sad case (for New York Mets fans, of which I am one) of Bobby Bonilla, who racked up spectacularly numbers hitting ahead of Barry Bonds for the Pirates, but flopped with the Mets. The same trader who makes big profits at Goldman based on its low cost of funding, sterling reputation, and tremendous client connections will not necessarily do nearly as well on his own.

Finally, while there is a strong relationship between pay and past performance, there is only a loose relationship between pay and future performance. Look at the teams that won the World Series between 2000 and 2009: Arizona Diamondbacks, Anaheim Angels, Florida Marlins, Boston Red Sox, Chicago White Sox, St. Louis Cardinals, Boston Red Sox, and Philadelphia Phillies. Only the Red Sox were among the sport’s traditional big-market teams. Yes, there is a correlation between payroll and number of wins (the Yankees do win the World Series more than other teams), but the random factors play a big role as well.

So yes, bankers are like athletes. Their individual contributions are overrated relative to their supporting environments; they are overpaid; they are paid based on where they randomly fall in the probability distribution in a given year; and paying a lot for bankers is no guarantee that your bank will be successful in the future. Team sports, like banking, are an industry where the employees capture a large proportion of the revenues. And one with negative externalities, like upsurges in domestic violence around major sporting events. Neither one should be a model for our economy.