Fed's forecasts tend to be accurate, alas


WASHINGTON -- When Federal Reserve Board Chairman Alan Greenspan began warning in February that he was seeing early signs that inflation might be intensifying, many private analysts wondered what his crystal ball had that their own forecasting models did not.

"I just don't see any serious indication that inflation is on the rise again," said one Wall Street economist at the time. "Of course," he added, "I'm always very cautious in contradicting the Fed. Maybe they know something we don't."

As it turns out, they probably did.

A new survey by the respected National Bureau of Economic Research shows that the Federal Reserve's confidential internal prognostications are consistently far more accurate than those of private economists, both in predicting inflation trends and in forecasting economic growth.

The study suggests that the Fed's predictions are so much better that anyone who had access to both sets would be best-advised to "put essentially no weight on the commercial forecast."

Indeed, the document says archly, if private analysts ever were given timely access to the Fed's internal calculations, the best thing they could do would be simply "to discard their [own] forecasts and adopt that of the Federal Reserve."

The study, compiled by the husband-wife team of David H. and Christina D. Romer at the University of California, Berkeley, suggests some fundamental reasons for the disparity, and it is not that the Fed has any special secret sources or inside information.

Rather, it says, the central bank's superior insight comes because its large, well-equipped staff is "simply better at processing and interpreting information." The Fed "commits far more resources to forecasting" than do big private firms, the document notes.

The problem is, under long-standing practice, the Fed keeps its internal forecasts secret for five full years before making them public -- a delay that makes it too late for anyone but historians (and researchers such as the Romers) to profit from their contents.

The Romers' study refers to the internal forecasts in the so-called "Green Book" that the Fed staff provides top policy-makers just before they decide whether to alter interest rates. It does not apply to the watered-down version that Greenspan uses in public statements.

To compile their comparison with predictions by private forecasters, the Romers tracked now-public Green Book forecasts for the period between November 1965 and December 1991. The Fed won hands down.

Some private forecasters have been aware of this disparity for years, and regularly nudge their own predictions to what they anticipate will be the Fed's position, based on public signals sent by the reserve board. But the Romers say the private analysts do not usually go far enough, and the gap -- inevitably -- remains large.

The Romers say the superiority of the Fed's forecasting ability probably helps explain at least part of the reason that the central bank has been so successful in managing the economy.

In fact, not long after the Fed raised interest rates by a quarter percentage point on March 25, the statistics began proving that the central bank had had a point in worrying that the economy was growing too strongly. The stock market quickly took that cue and plunged.

Although the National Bureau of Economic Research study does not cover the period since the latest Fed action, it suggests that one reason for the kind of market reaction that is occurring now is that investors are adjusting their own perceptions of the economy to bring them into line with the Fed's.

One question might be: If the Fed's forecasts really are that good, why does the central bank not make them public immediately, so that everyone can get a look?

Joseph R. Coyne, the Fed's spokesman, will not say, but it is clear that no such policy change is likely anytime soon.

Coyne contends that, despite all the secrecy, "the flavor" of the Fed forecasts comes out in minutes published six weeks after every meeting.

Jay Hancock's column will return in two weeks.

Pub Date: 4/14/97

Copyright © 2019, The Baltimore Sun, a Baltimore Sun Media Group publication | Place an Ad