Incentives

Incentives get at the heart of the current economic crisis.  In an interview yesterday for the WSJ, Roubini said

“I believe that people react to incentives, that incentives matter, and that prices reflect the way things should be allocated. But I also believe that market economies sometimes have market failures, and when these occur, there’s a role for prudential — not excessive — regulation of the financial system. The two things that Greenspan got totally wrong were his beliefs that, one, markets self-regulate, and two, that there’s no market failure.”

via Nouriel Roubini Says Nationalizing the Banks Is the Market-Friendly Solution – WSJ.com.

Roubini makes the point that incentives matter.  And indeed, incentives get at the heart of the current fiasco, and there may lie the long term solution as well.  Paradoxically, the short term fixes that might be needed to claw our way out of the crisis might end up creating all kinds of perverse incentives (i.e., the problem of moral hazard that occurs by rewarding destructive behavior through bailouts).  But for the moment, let’s step back though and think through a few of the perverse incentives that helped cause this problem.

Executives at large financial firms (and indeed, at many companies in general) are largely compensated through stock options.  The goal of stock options is well intentioned — they’re meant to align the incentives of the executives with the interests of the company.  If the company does well, then the executives profit, which seems like a good thing.  But as they say, the road to Hell is paved with good intentions.  The reason that some scheme needs to be setup to align incentives is because of what’s known as the principal-agent problem.  In this context, the company (which is essentially the collection of shareholders) is the principal, which hires the executives, who are the agents, to run the firm.  The executives want to act in their own self-interest, so we need a scheme to align their self interests with those of the company.  Stock options seem like a good way to do that.  Unfortunately, any incentive scheme is imperfect, and it is generally not possible to perfectly align the interests of the agent with that of the principal.  What are the imperfections of stock options?  Well there are many, and they’re not exactly secret; for instance:

• The agent is incented to increase short term performance of the company to drive up  the stock price,  just long enough for him to sell his shares.  Since options generally vest within 4 years, the long term health of the company is irrelevant to the agent.
• The agent is incented to be risk seeking because of the asymmetry of his exposure to the company’s performance — if the company does well, the agent gets lucratively rewarded; but if the company does poorly, the agent doesn’t suffer losses (assuming our starting condition is when the strike price of his option is the market value of the share).

Just these two fairly obvious flaws in the incentive schemes for Wall street executives goes a long way towards explaining the current financial crisis.  The execs at Fannie Mae, Freddie Mac, AIG, Countrywide, … ad naseum were all acting in their self interests by taking on undue risk, and profiting heavily.  But the principals (namely the companies themselves) suffered.  And how can you really blame the executives, when they were doing exactly what the company was incenting them to do?  (By the way, railing at the current Wall Street execs is somewhat pointless; the people behind this mess cashed out long ago.)

As a side note, consider the fact that stock options are heavily used in the tech industry as well.  It’s interesting that the largest and most prominent of the tech companies don’t suffer the same kinds of problems.  Isn’t it curious though that at highly successful companies like Microsoft, Amazon, and Google, the “principal” is to a large extent the founder(s) of the respective companies?  In other words, if Jeff Bezos is running the show at Amazon, he does care about the long term health of the company, since not only is he the founder, he’s by far still the largest stakeholder.

It seems clear that companies should develop incentive schemes that reward executives and employees for long-term rather than short term performance.  In some sense, pre-IPO companies are like that, in that it could be several years before an employee’s shares are even tradable.

Game theory provides a framework for understanding incentives and how people react to them.  Several other key aspects of the current crisis can be understood in that context, which I’ll post about later.  But bailouts, restructuring, and masses of regulation won’t really fix the problem until we think about better ways to improve the incentive structures at the key points of our economy.