Edward L. Glaeser

Too risky to regulate? Not with proper verification

By Edward L. Glaeser
December 3, 2009

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

Your article has been sent.

  • E-mail|
  • Print|
  • Reprints|
  • |
Text size +

IN THE 1912 presidential election, Woodrow Wilson fought for a “New Freedom’’ and favored breaking up over-mighty businesses. One of his opponents, Theodore Roosevelt, wanted a “New Nationalism,’’ where big business would be controlled by robust regulation and a powerful public sector. Today, America is again faced with the Morton’s Fork of either regulating or dismantling financial firms that are “too big to fail.’’ For Rooseveltian regulators to succeed, they need enough information to restrain excessive risk-taking and that requires a system where financial firms reveal each other’s risks.

In the halcyon days before 2007, libertarians could argue that financial failures impose few social costs, so there was little point to either regulation or size limits. But the wrenching market turmoil that followed Lehman Brothers’s collapse shattered that illusion. The bailouts that ensued showed that taxpayer dollars would be used to stop failures of financial firms. However, it is a bad idea to have unregulated firms betting with taxpayers’ money.

Alan Greenspan, former chairman of the Federal Reserve Board, recently said that if financial firms are “too big to fail, then they’re too big’’ and suggested that they may need to be split up the way Standard Oil was split up in 1911. His reasonable view is that only the threat of failure can discipline financial firms. Other new Wilsonians have correctly emphasized that sophisticated firms will game any regulations.

But bank-busting won’t solve the problems. Lehman Brothers was no giant, and the government bailed out even the smaller Bear Stearns. Since the public sector seems unable to let even modestly-sized financial firms go belly up, hard limits on their growth will restrict creative expansion and the gains from diversification without reducing the need for regulation.

In an ideal world, the government would differentiate between entities that would receive public support under a limited set of circumstances and unprotected firms that would never be bailed out. These smaller, unprotected entities could be lightly regulated and would enjoy the advantages that accrue from unfettered innovation. The harder question is how to limit inordinate risk-taking by the protected financial firms, especially given the difficulties in evaluating the risk in their portfolios. Ideally, bigger firms that enjoy some sort of public guarantee should pay a “public insurance premium,’’ which would be a function of capital levels and risk.

But the daunting problem in either charging such a premium, or enforcing plain old capital requirements, is measuring risk-weighted assets. Each new transaction means a change in a firm’s risk profile. Regulators can’t assess the risk transferred in each credit default swap, where one firm insures another against the default of a third. Certainly, firms have little incentive to accurately report the risks that they take.

This problem is not new. Every tax and regulation has to deal with measurement. For centuries, governments taxed imports, property values, and even the windows in homes because these things were more observable than income.

One way for regulators to better evaluate financial risk is to borrow a trick from the income tax system. Our incomes are reported more or less correctly because employers have a strong incentive to state our earnings. Since stating our full income reduces their tax liability, firms rarely help hide their workers’ incomes. Employers and employees that report different numbers are raising a giant red flag.

Banking regulation can use the same rat-on-your-partner structure. Every credit default swap involves two parties - one taking on more risk and one shedding risk. The regulatory system should treat this as a transfer of risk where the insurer is taking on exactly as much danger as the insured party is shedding. If each firm’s public insurance payments or capital requirements depend on their risk, then the insured entity has an incentive to accurately report the risk the insurer is taking on. With the right system, every firm has an incentive to report on each other and to make sure their numbers match.

Mistakes will still be made but forcing every financial firm to be small won’t solve everything. Better regulation is the best way forward, and effective regulation depends upon the information flows that come from a system where these firms inform on each other and work for the rest of us.

Edward L. Glaeser, a professor of economics at Harvard University, is director of the Rappaport Institute for Greater Boston.

More opinions

Find the latest columns from: