Be Careful, Mr. Greenspan

Central banking is the last refuge of central planning -- the notion that a group of experts can meet in secret and plan the economy from the top down. But this is a game played without any rules. We speak of the "art of central banking," as though it is akin to a magic show. So long as a central bank doesn't mess up too badly, we tend to almost deify central bankers. When they do mess up, many then argue that we must have deserved the pain and suffering as penance for the good times.

One well-known problem with this magic show is called "recognition lag," and it often results in "overshooting" -- pushing interest rates too high or too low for too long. Look carefully at the reasons given for Fed decisions, and you will find they always refer to something that happened in the past. Economic growth looked fine last year or last quarter, for example. This is like speeding down the highway while trying to steer your car by gazing in the rearview mirror -- to see if you're staying inside the lines.

For mysterious reasons, these speeding, backward-gazing magicians prefer to focus on past news about the real economy, rather than growth of inflation or nominal spending. The inference is that vigorous growth must be inflationary and that weak growth ensures weak inflation. Yet inflation has always gone up whenever economic growth turned flat or negative -- in 1974-75, 1979-81, 1990 and even 2001.

Nobody knows what the Fed will do next, or why. But I do know what happened in the past under eerily similar circumstances. Every recession in the past 30 years has been preceded by a confluence of four events, three of which may have been avoidable:

First of all, energy prices were rising rapidly before every recession. Within the consumer price index (CPI), energy prices rose 8.1 percent in 1973 and 29.6 percent in 1974; 25.1 percent in 1979 and 30.9 percent in 1980; 5.6 percent in 1989 and 8.3 percent in 1990; and 16.9 percent in 2000.

Aside from 2000, inflation in general was terribly high even before those previous energy price spikes -- the non-energy CPI rose 7.8 percent in 1978 and 4.4 percent in 1988. After oil prices spiked, however, the non-energy CPI always slowed rather than accelerated for a few years. There is no evidence that energy price spikes have ever led to higher non-energy inflation. That's a dangerous myth.

Second, as a result of Fed tightening, the yield curve became flat (as in 2000) or inverted. The Fed pushed the fed funds rate above the yield on long-term Treasury bonds in 1969, 1973-74,1979-81 and 1989.

Third, the fed funds rate was relatively high in real terms -- at least two or three percentage points higher than inflation in the CPI less energy.

These same interest rate patterns work well in reverse. Before inflation accelerated, the yield curve steepened and the real fed funds rate was near zero or negative. The funds rate was well below bond yields in 1971-72, 1975-77, 1987-88 and 2003-2004. And the funds rate was equal to or lower than non-energy CPI inflation in 1974-77, 1992-93 and 2003-2004. This is not what is happening now. The opposite, in fact.

If we compare the funds rate with long bond yields or non-energy inflation, the Fed was indeed too easy in 2003-2004. But not this year. Those advocating several more increases in the fed funds rate appear to be steering with the rearview mirror.

Although I believe the Fed pushed the fed funds rate too low in 2003-2004, that does not require or justify overshooting in the other direction in 2006. If that happens, the Fed might once again overreact to the resulting recession by easing too aggressively long after the recession ended (as in 1992-93 and 2003-2004). To avoid lurching back and forth between brake and accelerator requires great caution and foresight. Rules would be better, but there are none. Monetary policy is a euphemism for whim and caprice.

The fourth potentially troublesome development is this: Whenever energy prices and U.S. interest rates were simultaneously rising in the past, foreign central banks turned that into a synchronized global squeeze by raising their interest rates, too. That copycat behavior was partly a consequence of confusing energy price spikes with a general inflation. The dollar was rising, as it has been this year, which makes dollar-based oil prices rise even faster in terms of sinking yen or euros.

"Banks From Japan to Europe Ponder Tightening With Fed," notes a front-page headline in The Wall Street Journal. The writer's only hope was that "a reversal in oil prices or renewed economic slump could put them on hold." That confirms my concern that central banks at home and abroad are focusing on the one relative price they are least able to control without precipitating recession -- namely, the world price of oil. A higher fed funds rate is a perverse way to persuade the Chinese to buy less oil, or the Texans to produce more.

As for the ominously familiar idea that the biggest central banks will push rates up until there is clear evidence of a "renewed economic slump," that requires a trial-and-error process that can only end one way -- in error. By definition, rates would keep rising until it's too late.

The non-energy CPI has averaged 2.2 percent since 1996 -- compared with 5.5 percent from 1967 to 1995. On a year-to-year basis, non-energy inflation was still 2.2 percent the last time I checked. It was even lower if you look at a more accurate chain-weighted index, and lower still in Japan and Europe.

The futures market has lately been expecting the U.S. fed funds rate to reach 4.5 percent early next year and keep heading higher. Unless something unexpected happens to lift the non-energy CPI above its stubborn 2.2 percent trend, a 4.5 percent funds rate suggests a real interest rate on cash of about 2.3 percent.

It could be worse. The comparable real interest rate hit 3.5 percent in 1990 and 2000 before the economy began to unravel. Unless bond yields rise substantially, however, a funds rate of 4.5 percent or more would nonetheless be higher than the yield on 10-year Treasury bonds. By this measure, the yield curve would be at best flat and probably inverted. That has never happened in the midst of an energy price spike without the economy slipping into the tank. Never.