Thursday, July 01, 2004
Somebody check my math
...it came as a surprise to some economists and portfolio managers that Fed policy makers thumbed their nose at inflation worries in the statement accompanying the rate increase. "Although incoming inflation data are somewhat elevated, a portion of the increase in recent months appears to have been due to transitory factors," the policy makers said.The Mises Institute people recently put up this article by Frank Shostak discussing the troubles of the Taylor Rule, which is used as a forecasting tool for what Federal Reserve policy regarding the federal funds rate target will be. That's important to emphasize: Nobody should believe that the Fed is actually using this formula, only that the formula seems to mimic what rates they tend to set. I have created macro models for some ministries of finance in developing countries, and we sometimes have used a Taylor-esque rule in the model simply so the ministry can guess at what the central bank will do. The central bankers would tell us they don't use the Taylor rule. We would reply that it didn't matter what they used, the equation was a fair approximation of whatever the heck it was they were doing.
That is not the opinion of Alan W. Kral, portfolio manager at Trevor Stewart Burton & Jacobsen in New York. "We believe that inflation has returned," Mr. Kral said. "And the cause of it has been an overexpansive monetary policy for almost 10 years."..."I believe the Fed is behind the curve because the economy continues to be strong and the inflation rate is creeping up and will continue to creep up," said Henry Kaufman, an economist in New York. Its plan to raise interest rates gradually may be good for the economy, he said, "but is not designed to put the system back into balance in terms of constraining inflation itself."
The formula for the rule is target FF rate = 2% + current inflation + 1/2*(deviation of inflation from 2%) + 1/2*(percentage gap of GDP from its potential level)
So let's check the current 1.25% target set by the Fed.
- The initial 2% is a long-run estimate of what the rate would be with full employment and price stability (neoclassical economists would call it the "real equilibrium Fed Funds rate".)
- Because of the Fisher equation, the nominal or stated rate is going to be the real rate plus inflation. Most Fed watchers believe the Fed uses the inflation rate for personal consumption expenditures, which has run at 1.6% over the last twelve months. 2%+1.6%=3.6%.
- That 1.6% is currently 0.4% below the Fed's target inflation rate of 2%. Shostak has roundly criticized this feature; I'm quite sympathetic to his argument, but I'm in the prediction game right now so I have to go with it. So we're going to knock off 0.2% from the fed funds target, because right now the Federal Reserve believes that the 1.6% is a better representation of true inflation than the headline rate for CPI at 2.9%.
- The Fed uses the output gap for one or both of the following possible reasons. First, it might be trying to target GDP due to a belief of the Phillips Curve. I sure hope not because that is a thoroughly discredited notion in economics (even at the Fed?), but some people still believe it. The other possibility is that the output gap might actually help forecast expected inflation. You can get potential GDP figures from the Congressional Budget Office and actual GDP figures from the Bureau for Economic Analysis. Using real (2000 chained) figures of $10851.6 billion for potential and $10702 billion for real GDP for the first quarter of 2004, for an output gap of 1.4%. That means that GDP is currently 1.4% below potential, and we should move the target Fed Funds rate down to 2% + 1.6% - 0.2% - 0.7% = 2.7%.