FED FOCUS-Central banks toil to glean inflation expectations

DALLAS May 25 (Reuters) - Measuring inflation expectations is fraught with pitfalls and monetary policy-makers must work hard to extract clear information from the various tools they have, central bankers and academics said on Friday.

The task has become easier, but is still a very imprecise science, according to papers presented at a conference hosted here by the Dallas and Cleveland Federal Reserve banks.

The Federal Reserve, European Central Bank and other authorities care about expectations for inflation because it helps them gauge how households view price shocks, like a jump in the cost of oil, and compare it with their own forecasts.

"If your objective is to stabilize prices, you want to know what that index is, and what the risks are in reaching that objective," said Stephen Cecchetti, a professor at Brandeis University and former research chief at the New York Fed.

In particular, inflation expectations provide feedback so policy-makers can keep tabs on whether households and investors see their actions as credible.

If inflation expectations rose, it might signal monetary policy was too loose. This would expose a lack of confidence in the central bank's ability to keep inflation at bay, while also shedding light on its communication strategy.

"It helps us to understand how people are reacting to misperceptions about how the economy will unfold ... and that might be different from the view of policy-makers," said Mark Sniderman, head of research at the Cleveland Fed.

"If we understand how that view is formed it helps us with our own view, and might signal a lack of clarity from policymakers," he said.

The Federal Reserve is internally debating its communications strategy, including whether it should adopt an explicit inflation target. It is expected to refresh this discussion at its next policy meeting, on June 27 and 28.

There are three main ways to measure inflation expectations: surveys of households and of professional forecasters, and movements in index-linked bonds.

TIPS

One paper examined the Treasury Inflation Protected Securities (TIPS) market. Launched in 1997 to compensate investors for changes in the headline consumer price index, it now accounts for 10 percent of outstanding Treasury debt.

A feature of the market is that subtracting the yield of a TIPS security from the yield of a normal coupon-bearing Treasury note with a corresponding maturity exposes what the market expects inflation to be.

This helps the Federal Reserve get a sense of the inflation financial markets are expecting, which it can combine with consumer surveys to form its own judgments.

However, through economic modeling, the authors identified significant mis-pricing in this sector until 2005, which they put down to an unexpectedly high liquidity premium that investors demanded for buying these instruments.

This premium narrowed as the market grew and investors became more comfortable with these securities. By 2005 it had all but vanished, improving the information which can be extracted from their behavior.

"That the early mis-pricing has now disappeared is good news ... for a central bank that wants to extract inflation information," said George Pennacchi from the University of Illinois, who discussed the paper during the conference.

There is some scepticism about how much households are really able to judge credibility, although professional forecasts should be better equipped. But volatile oil prices since 2001 have wrongfooted this community quite badly, according to work by the ECB.

It has surveyed around 75 forecasters consistently over the last eight years, but noted in a paper presented here that their track record was far from stellar.

"The Survey of Professional Forecasters systematically under-forecast inflation," the paper noted, adding that "SPF respondents do not seem to fully capture the overall level of macroeconomic uncertainty,".

On the other hand, ECB economist Aidan Meyler, who presented the paper, admitted this performance was "not much worse than ourselves", given that the last eight years had been a difficult period in which to forecast inflation.