"The Fed versus the financiers"

Below are some excerpts from a good piece by Kenneth Rogoff on the Guardian’s website, which discusses why Bernanke thinks the Fed has no business worrying about asset prices, but argues that the lousy quality of macroeconomic data means policy makers ought to consider the information embedded in market prices:

A bit of intellectual history is helpful in putting today’s debate in context. Bernanke, who took over at the Fed in 2006, launched his policy career in 1999 with a brilliant paper presented to the same Jackson Hole conference. As an academic, Bernanke argued that central banks should be wary of second-guessing massive global securities markets. They should ignore fluctuations in equity and housing prices, unless there is clear and compelling evidence of dangerous feedback into output and inflation….

On the surface, Bernanke’s view seems intellectually unassailable. Central bankers cannot explain equity or housing prices and their movements any better than investors can. And Bernanke knows as well as anyone that none of the vast academic literature suggests a large role for asset prices in setting monetary policy, except in the face of extraordinary shocks that influence output and inflation, such as the Great Depression of the 1930s.

In short, no central banker can be the Oracle of Delphi. Indeed, many academic economists believe that central bankers could perfectly well be replaced with a computer programmed to implement a simple rule that adjusts interest rates mechanically in response to output and inflation.

But, while Bernanke’s view is theoretically rigorous, reality is not. One problem is that academic models assume that central banks actually know what output and inflation are in real time. In fact, central banks typically only have very fuzzy measures. Just a month ago, for example, the US statistical authorities significantly downgraded their estimate of national output for 2004!

The problem is worse in most other countries. Brazil, for example, uses visits to doctors to measure health-sector output, regardless of what happens to the patient. China’s statistical agency is still mired in communist input-output accounting.

Even inflation can be very hard to measure precisely. What can price stability possibly mean in an era when new goods and services are constantly being introduced, and at a faster rate than ever before? US statisticians have tried to “fix” the consumer price index to account for new products, but many experts believe that measured US inflation is still at least one percentage point too high, and the margin of error can be more volatile than conventional CPI inflation itself.

So, while monetary policy can in theory be automated, as computer programmers say, “garbage in, garbage out”. Stock and housing prices may be volatile, but the data are much cleaner and timelier than anything available for output and inflation. This is why central bankers must think about the information embedded in asset prices.

In fact, this summer’s asset price correction reinforced a view many of us already had that the US economy was slowing, led by sagging productivity and a deteriorating housing market. I foresee a series of interest rate cuts by the Fed, which should not be viewed as a concession to asset markets, but rather as recognition that the real economy needs help.

In a sense, a central bank’s relationship with asset markets is like that of a man who claims he is going to the ballet to make himself happy, not to make his wife happy. But then he sheepishly adds that if his wife is not happy, he cannot be happy. Perhaps Bernanke will soon come to feel the same way, now that his honeymoon as Fed chairman is over.