As Ben Bernanke took over from Alan Greenspan this week at the Fed -- and let's hear it for financial markets for not freaking out that much about Greenspan's departure -- it seems only fitting to link to Adam Posen's Institute for International Economics brief about what central banks should do when there's an asset price bubble.

Basically, they should do nothing:

Central banks should not be in the business of trying to prick asset price bubbles. Bubbles generally arise out of some combination of irrational exuberance, technological jumps, and financial deregulation (with more of the second in equity price bubbles and more of the third in real estate booms). Accordingly, the connection between monetary conditions and the rise of bubbles is rather tenuous, and anything short of inducing a recession by tightening credit conditions prohibitively is unlikely to stem their rise. Even if a central bank were willing to take that one-in-three or less shot at cutting off a bubble, the cost-benefit analysis hardly justifies such preemptive action. The macroeconomic harm from a bubble bursting is generally a function of the financial system’s structure and stability—in modern economies with satisfactory bank supervision, the transmission of a negative shock from an asset price bust is relatively limited, as was seen in the United States in 2002. However, where financial fragility does exist, as in Japan in the 1990s, the costs of inducing a recession go up significantly, so the relative disadvantages of monetary preemption over letting the bubble run its course mount. In the end, there is no monetary substitute for financial stability, and no market substitute for monetary ease during severe credit crunch. These two realities imply that the central bank should not take asset prices directly into account in monetary policymaking but should be anything but laissez-faire in responding to sharp movements in inflation and output, even if asset price swings are their source.