Where forecasting went right

Alan Reynolds
Posted: Jan 05, 2006 12:05 AM
The first of the year brings surveys of economic forecasts, which soon become irresistible targets of journalistic derision. Washington Post columnist Steven Pearlstein just wrote about the "remarkably rosy scenario" of the "economic pundocracy" (though he's a member). On the same day, Robert Samuelson brushed off all the "sunny predictions" and offered a variation of Pearlstein's shopping list of things that might go wrong. The dollar might crash, for example, which they almost certainly said last year. Or we might be hit by a comet.

In my capacity as an economic journalist, I too have often succumbed to the temptation to make fun of forecasters. The difference is that I actually was a forecaster for many years.

In the late '70s, I found myself on a panel in Chicago at the American Economics Association with Bob Eggert, who founded Blue Chip Economic Indicators in 1976. I remarked that economic forecasts were an excellent lagging indicator of where the economy has been. That resulted in me being recruited to join the Blue Chip forecasters for many years, until it dawned on me that Eggert was getting the last laugh on forecasters by selling our work and not paying us even the minimum wage.

Since forecasting paid nothing, I rarely spent more than a few hours on it. Yet, I nonetheless continued to contribute to the Wall Street Journal survey from 1986 to 2000 as a hobby -- a sport similar to target practice. I was understandably nervous when, in 2002, the Federal Reserve Bank of Atlanta published a potentially embarrassing study --"Evaluating Wall Street Journal Survey Forecasters."

Larry Kudlow came in with an extraordinarily accurate score of 74.9, compared with mine of 65. We are old friends who worked together on the OMB transition team in 1981. As the study notes, however, "the people with the superior performance record tend to be those whose forecasts covered a short period of time." The longer the records are kept, the greater the risk of being caught stumbling over surprises. The highest score went to someone who had participated in only five surveys after the 1991 recession, the second highest made six forecasts. By contrast, Kudlow had accumulated 11 forecasts, while I had 27 under my belt -- including two recessions.

Among those with as dangerously long a forecasting record as I had, the two with a higher score were someone named Hoffman (67.7) and someone I knew -- David Resler of Nomura Securities (67.8) -- because we had worked together at the First National Bank of Chicago.

Forecasts tend to cluster close together, because it's risky to go out on a limb -- as I did on Jan. 1, 2001 with my sunny forecast that "we are precariously close to recession, or already in one." We are not always sunny or rosy. The expert judgment of 56 economic forecasters in The Wall Street Journal is not something that can be brushed off that easily.

This year depends on the forecasters being right about inflation, and I think they are. What happens to inflation should affect the one risk that Pearlstein, Samuelson and I agree is real -- namely, that the Fed might push the fed funds rate significantly above the yield on 10-year bonds, thus inverting the yield curve.

Wall Street Journal inflation forecasts are for the 12 months ending in May. The 56 forecasters expect the consumer price index (CPI) to be up by 3.1 percent between May 2005 and May 2006. But they already know what inflation was for six of those 12 months -- namely 3.8 percent (half of which was energy). To get that 12-month average down from 3.8 to 3.1 percent by May, inflation will have to average just 2.4 percent from December to May . Nearly half the forecasters expect it to be even lower.

Is that unrealistic? I don't think so. Aside from energy, which is last year's story, the CPI rose at only a 1.9 percent rate over the past six months. The odds of energy prices rising much further from today's high level are near zero. Even if energy remained as expensive as it is, that would add nothing
to future inflation. To get overall inflation up to 2.4 percent would therefore require much faster price hikes for everything except energy.

I keep reading that the Fed worries that last year's energy price spike will spill over into this year's nonenergy prices. Yet, that has never happened. Never.

The biggest previous spikes in CPI energy prices, with increases of 25-30 percent per year, occurred in 1974 and in 1979-1980. Yet, nonenergy inflation subsequently slowed from 9.8 percent in 1974 to 5.6 percent in 1976. The pace of nonenergy inflation likewise hit 11.6 percent, while energy prices spiked in 1980, but fell to 3.6 percent during the booming recovery of 1983 (more supply is not inflationary).

The next energy-price spike in 1989-90 was again followed by lower nonenergy inflation, which slowed from 5.2 percent in 1989 to 2.7 percent by 1993. Energy prices in the CPI rose by 16.9 percent in 2000, but that too was quickly followed by slower price increases for everything else.

Nonenergy inflation has always slowed rather than accelerated in the wake of previous energy price spikes.

This is where the yield curve comes in. Because inflation is going to be drifting down to 2.4 percent or less for the foreseeable future, there is no "conundrum" in explaining why 10-year bond yields remain low. And there is no plausible rationale for pushing the fed funds rate higher while inflation is heading lower . If the Fed does that, it will be a mistake. Maybe a big mistake.

The federal funds rate was deliberately pushed above the 10-year bond yield in 1969, 1973-74, 1979-81, 1989 and 2000. By no coincidence, the economy was in recession in 1970, 1974-75, 1980-82, 1990 and 2001. In the last four of those cases, energy prices alone appeared to mislead the Fed into overdoing "restraint." Yet that seems somewhat less likely now because oil prices have likely peaked.

The funds rate is not above the 10-year bond yield today. Even if we were, there has often been a lag of a year between yield curve inversion and recession. On the other hand, I might be worrying about nothing. For better or worse, however, my own batting average, like the Fed's, is a matter of public record.