Regulating Risk Systemically, But Not Systematically

Kevin Drum responds to Tyler Cowen's post on the Fed as systemic risk regulator:

It's true that the Fed is the agency with the brute force to make things happen in an emergency.  But I'm not sure that's the relevant thing to think about.  What we want is some kind of body that works to prevent emergencies.  That requires credibility and influence, but it doesn't necessarily require a trillion dollar balance sheet.

I guess the model I have in mind here is the Congressional Budget Office.  The CBO is unknown to most people, but despite its small size and low public profile it has a remarkable amount of power.  This power comes from two sources.  First, it has institutional credibility.  I honestly don't know how it's managed to keep this credibility in the face of what must be enormous partisan pressure, but it has.  It's widely considered an honest broker and its budget estimates are taken seriously by everyone.

Second, although the CBO itself doesn't have a huge staff or control of a huge budget, Congress has agreed to abide by its cost estimates for legislative programs.  This means that CBO analysts have considerable indirect control over a lot of money.  And in Washington, money equals power.

So my question is: could we create an agency like the CBO, but charged with monitoring systemic risk in the financial system?  It would have to be nonpartisan and independent.  It would need to have risk management baked into its DNA as its primary mission, rather than being #7 on a list of ten goals -- with everyone knowing that only the top three get any real attention anyway.  Its director would need the kind of credibility that makes people listen when he warns that other agencies are allowing too much giddiness on Wall Street.  And, finally, it would need the right mix of authority, either direct or indirect, that's enough to force people to take it seriously when its mere credibility isn't quite enough.

But here's the incoherent part: I'm not quite sure how you'd construct such an agency or what authority might be sufficient for it to do its job without getting it hopelessly at odds with other regulatory agencies.  One way or another, though, I feel that giving this mission to the Fed is simply a waste of time.  Right now, virtually every impulse -- both at the Fed and in the private sector -- works in the direction of either ignoring credit bubbles or actively cheering them on.  If we're going to put a brake on this, we need to think about institutional priorities and balances of power, and figure out what it would take to get systemic risk established as a bureaucratic turf with a built-in constituency dedicated to protecting it over the long term.

My thoughts:

  • In finance, the power to stop crises is often the power to prevent them.  The FDIC stops a hell of a lot of bank runs from ever happening because people know their deposits are insured.
  • More broadly, attempting to build systems that can't fail have generally . . . failed.  Such systems tend to be brittle--a single failure is catastrophic.  In general, we should seek ways to let systems fail gracefully.  I.e.  no matter how strongly you try to ensure that nuclear power plants can't overheat, you also build them so that if they do, there is no critical mass to create a bomb-like nuclear blast.
  • A systemic risk regulator outside the Fed would spend a lot of time at war with the Fed.  It would probably lose.  And if it did not lose, the result might be ugly.  A regulator that is only focused on preventing downside will be too conservative.  Right now, that seems like a good thing.  But inefficient capital allocation also has a high cost.  Whether the systemic risk regulator won or lost, the battle would create great instability in financial markets at a time when we don't need more of it.
  • A systemic risk regulator will probably have a harder time staying independent of Congress than the Fed, because the inflationary history of fiat currency puts a high cost on Congressional meddling.  The systemic risk regulator will have no such insulation, making it less effective at its job.