How should the Fed respond to asset bubbles?

In the Wall Street Journal, Gerald Driscoll traces the current troubles in financial markets to the Fed's unstated guarantee that it will bail the markets out in a crunch:

In recent years, monetary policy has created an expectation that the Federal Reserve will bail out investors when asset bubbles deflate. The recent crisis in the subprime mortgage market is at least partly the outcome of this new approach to monetary policy. That crisis has already had widespread ramifications for homeowners and investors...

Today, monetary policy is fostering moral hazard. Monetary policy can generate moral hazard if it is conducted so as to bail investors out of risky and otherwise ill-advised financial commitments. If investors come to expect that the policy will persist, then they will deliberately take on additional risk without demanding commensurately higher returns. In effect, they will lend at the risk-free interest rate on risky projects, or at least at a lower rate than would otherwise be the case. Too much risky lending and investment will take place, and capital will be misallocated.

The new moral hazard in financial markets has its source in what can be best described as the Greenspan Doctrine. The doctrine was clearly enunciated by Alan Greenspan in his December 19, 2002 speech. Mr. Greenspan argued that asset bubbles cannot be detected and monetary policy ought not to in any case be used to offset them. The collapse of bubbles can be detected, however, and monetary policy ought to be used to offset the fallout.

Two months earlier, Mr. Bernanke endorsed the Greenspan Doctrine, arguing against the use of monetary policy to prevent asset bubbles: "First, the Fed cannot reliably identify bubbles in asset prices. Second, even if it could identify bubbles, monetary policy is far too blunt a tool for effective use against them." Since Mr. Bernanke is now Fed chairman, it is reasonable for market participants to assume that the Greenspan Doctrine still governs current Fed policy.

The two men were surely asking and answering the wrong question. They were implicitly treating bubbles as solely the consequences of real shocks or disturbances. (An example of a real shock is a technological innovation leading to productivity gains and higher future expected profits in a sector.) They asked whether monetary policy should be used to offset the effects of real shocks, and concluded that it should not. The latter is the correct answer to the question they each posed.

A different question would be to ask whether monetary policy should be conducted so as to create or exacerbate asset bubbles. The answer to that question is surely "no." Consider Mr. Bernanke's apt characterization of moral hazard in the context of the deposit insurance crisis: "When this moral hazard is present, credit flows rapidly into inelastically supplied assets, such as real estate. Rapid appreciation is the result, until the inevitable albeit belated regulatory crackdown stops the flow of credit and leads to an asset-price crash."

He could have been talking about the subprime mortgage market. The Fed pre-announced that it will take no action against bubbles, but will act aggressively to offset the consequences of their collapse. In effect, the central bank is promising at least a partial bailout of bad investments. The logic of the old deposit insurance system is at work: Policymakers should protect investors against losses, no matter their folly. Or, in Mr. Greenspan's own words: Monetary policy should "mitigate the fallout [of an asset bubble] when it occurs and, hopefully, ease the transition to the next expansion."

In the present context, the "next expansion" could also be rendered as "the next asset bubble." If the Fed promises to "mitigate the fallout" from "irrational exuberance," then it is rational for investors to be exuberant. Investors may be at risk for some loss, as with a deductible on a conventional insurance policy, but losses are still being mitigated.

The Bernanke Fed has confused matters for investors by not yet cutting interest rates in the face of the recent crisis. There are two possible (not mutually exclusive) reasons for its not doing so. First, it may not view the current crisis as serious enough. Second, current price inflation is above its comfort zone, and the Fed may feel it has no room to maneuver. Time will only tell which is at work...

A monetary policy of substantial stimulus will have a number of real consequences, including asset bubbles. These asset bubbles have real costs and involve misallocations of capital. For example, by the peak of the tech and telecom boom in March 2000, too much capital had been invested in high-tech companies and too little in "old economy firms." Too much fiber-optic cable was laid and too few miles of railroad track were laid.
By 2002, worried about the possibility of price deflation, the Fed introduced a strong anti-deflationary bias. A tilt to stimulus was understandable at the time. A continued bias against deflation at any cost, however, will produce a continued bias upward in price inflation. With the bursting of each asset bubble and the fear of deflationary pressure, Fed policy must ease. The Greenspan Doctrine prescribes a stimulative overkill that begins the cycle anew. The Greenspan-era gains against inflation will then prove to be only temporary. His doctrine will be the death of his legacy, a legacy that already includes a housing bubble and its aftermath.

Driscoll is right to warn about the perils of overreacting to the current crisis. But he makes what seems to me some basic errors in his analysis of what the Fed is actually doing right now. First of all, Greenspan and Bernanke's argument that the Fed should not respond to asset bubbles is based on their belief that economists at the Fed (and elsewhere) are usually unable to pinpoint the source of the bubble. That is, Greenspan and Bernanke's analysis does not assume that the bubble is driven by real factors (if they knew that, they would know that there was no bubble at all, just a rise in the fundamentals-driven price of the assets) - they assume that the Fed cannot tell whether the source is real factors (in which case the Fed should do nothing) or expectational factors (in which case the Fed should act to offset the bubble). Not knowing why prices are rising, Greenspan and Bernanke say the Fed should do nothing unless and until the economy - GDP, unemployment, inflation - is affected. This is a perfectly sensible rule. The Fed's legal mandate, after all, is to stabilize the economy, not financial markets.

Does the Greenspan doctrine constitute a promise to bail financial markets out in a pinch? Not at all - the Fed will let markets fall as long as there is no economic fallout. After the crash in 2000 the Fed waited almost a year to cut rates because it wasn't clear until December 2000 that the economy was headed for recession. The Fed is not cutting rates now because there is no evidence (yet) that the financial problems will lead to a slowdown in growth. The Fed seems to me to be following a symmetrical Greenspan rule - don't raise rates to pop an asset (in this case housing) bubble, don't cut them during a crash.