Contact

News

 

Misleading Economic Indicators


The fortune tellers of the economic data world have had their crystal balls clouded by Federal Reserve chief Alan Greenspan's so-called "conundrum".

The peculiar decline in long-term interest rates as the Fed pushes up short rates has rendered one oft-used component of leading economic indices slightly dysfunctional. The shrinking of the gap between short and long-term interest rates -- a so-called flattening of the yield curve -- has typically been seen as a harbinger of a slowing economy, ebbing inflation and lower official interest rates. As a result, many business cycle researchers compiling their oracles use the yield spread along with other early economic signals... .

But these leading indices -- which aim to warn about future peaks and troughs in the economic cycle -- have leaned negative over the past year about future U.S. economic growth even as national output continued to grow well above trend. Now, questions raised by the Fed and others about whether the yield curve has been distorted by factors such as global savings gluts and demographic trends is leading many to review their use of this measure.

The Conference Board, the firm which publishes the U.S. Leading Economic Indicators series, said its index has declined 1.9 percent in the year to May, with about half of the drop due to the flattening of the yield curve. This would typically signal a sharp economic slowdown this year. But it is a lonely signal. Top forecasters polled by Blue Chip Economic Indicators this week raised their outlook for U.S. growth in 2005 for the second straight month, now focussing on a robust 3.6 percent.

Perplexed, the Conference Board announced last month that it is changing how the yield spread affects its reading of where the economy is heading.
It said declines in the yield spread will no longer act as an outright negative on its leading index unless the curve inverts, or long rates fall below short rates.
"You won't find a full year in history where the Fed steadily raised interest rates by more than 200 basis points and the bond market not only didn't raise 10-year rates, but dropped them," said Conference Board economist Ken Goldstein.

Since last June the Fed has raised its key rates to 3.25 percent from 1 percent. But 10-year yields have declined to little over 4 percent from about 4.60 percent in the interim.

The Organisation for Economic Cooperation and Development also uses the yield spread in its closely watched Composite Leading Indicators series, which aim to predict industrial cycles in OECD member countries. Its U.S. index fell to 101.6 in May from 102.2 in May 2004 and a six-month rate of change is down 1 percent.

Ronny Nilsson, an economist at the OECD in Paris, said there is likely to be a review of the components of indices. The yield spread was only recently introduced in 2002, and nominal short-term and long-term rates were separate inputs until then, he said. But Nilsson acknowledged a problem with the spread is gauging whether its movements are a positive or negative economic influence. Are falling long-term rates a plus for the economy because they mean cheap borrowing or a negative signal of future activity? What is more, financial data tend to have a longer lead time on changes in economic activity but they are often very "noisy", Nilsson said. "There are a lot of minor fluctuations that never get picked up in GDP and that in itself is a good argument to play down their influence in the index," he said.

Others think use of the yield spread is inherently flawed.

Lakshman Achuthan, managing director at the forecasting group Economic Cycle Research Institute, said his firm's weekly leading index does not use the yield spread. "It just doesn't pass muster," he said, adding the curve did not invert before the 1990/1991 recession. Achuthan said that since the mid-1950s there were three occassions when the yield curve did not invert ahead of a recession and one time when it inverted and there wasn't a recession. "That's three misses and one false signal."

ECRI, Track Record, to be one of the few to flag an impending U.S. recession early in 2001. It was set up by Geoffrey Moore, who developed the U.S. government's first leading indicator in the 1960s -- the one now published by the Conference Board. ECRI's weekly leading index for the United States, which is up about a point on a year ago and whose annualized growth rate rose to a seven week high last month,... .

But Achuthan said another problem with financial data (yield curve) is that while they have a long lead time the variability of the lead is also very large.
"If I was to tell you there's a risk of a turning point a year and half out -- give or take a year -- what are you going to do? Run out and buy stocks?"