StockWatch

Double Dip Recession?

By Mark Hulbert,
Editor of the Hulbert Financial Digest, a service of MarketWatch.com

"A double dip recession is all but certain."

That's what a growing number of investment advisors I monitor are saying -- which, in turn, makes me suspicious. Contrarian analysis teaches me to be skeptical of any claim of certainty in the investment arena, especially if and when more and more advisors begin repeating it.

So, for this column, I decided to take a critical look at the evidence that supposedly guarantees that the economy is about to embark on another leg down.

The piece of data that seems to be advanced most in support of this prediction is the Economic Cycle Research Institute's Weekly Leading Indicator (ECRI WLI). The advisors I monitor are falling over themselves in praising this indicator, with one calling it, "the most prescient statistical guide to the health of the U.S. economy".

Another wrote that the ECRI WLI has "an almost 'holy allegiance'."

Wow. If comments like those aren't enough to get one's contrarian juices flowing, then nothing will…

What has so exercised the advisors about the ECRI WLI is not its level but its growth rate. That's because the WLI itself, at 122.5 currently, remains relatively high -- though down from its 134.7 all-time high set on April 30 -- yet still 16 percent above its bear market low set in March 2009.

But, the growth rate is most definitely coming down and is currently rated minus 5.7 by ECRI.

One analyst I monitor wrote that a negative growth rate has been a reliable predictor of impending recession -- getting it right for all but two recessions over the last four decades. In addition, the analyst continued, a recession would become virtually certain if the growth rate were to drop to minus 10 percent -- something he thinks is likely to come to pass any week now.

The analyst continues: "There is an indisputable fact that when this indicator reaches minus 10 percent, there has been a 100-percent correlation to a developing new recession. Call this a 'Guaranteed 2010 Double Dip'."

This, for sure, doesn't paint a pretty picture for the economy's prospects.

The problem with these conclusions, however, is that they are based on very few data points. During the time the ECRI has been producing its WLI, there have been just seven recessions -- as judged by the National Bureau of Economic Research, the unofficial arbiter of when recessions begin and end in this country. It's difficult, if not impossible, to draw statistical conclusions with confidence based on a sample this small, no matter how impressive the correlations otherwise appear.

But, how impressive is the ECRI WLI in predicting a recession? You be the judge. The accompanying table lists the seven NBER-defined recessions over the last four decades, along with when the ECRI WLI's growth rate fell to at least minus 10 percent.

Date when ECRI WLI growth rate first fell to minus 10%	NBER-defined recession
12/19/1969	Dec. 1969 to Nov. 1970
12/21/1973	Nov. 1973 to Mar. 1975
4/11/1980	Jan. 1980 to Jul. 1980
Not before or during	Jul. 1981 to Nov. 1982
Not before or during	Jul. 1990 to Mar. 1991
10/12/2001	Mar. 2001 to Nov. 2001
2/15/2008	Dec. 2007 - ???

Notice that in the case of two of the seven recessions, the ECRI WLI growth rate never fell to minus 10 percent, either before or during the recession itself. In two more cases, it didn't do so until after the recession was already half over -- and, in one of those cases, far closer to the end of the recession than the beginning. Only in three of the seven cases did a minus 10 percent growth rate for the ECRI WLI actually succeed in being a leading indicator.

Let me hasten to add that this recitation of the facts is not intended as a criticism of the ECRI WLI itself. Instead, it is intended as a criticism of how some advisors' use the indicator. In contrast to them, the ECRI is very careful to stress that no one indicator should be used to predict a recession.

As fellow MarketWatch columnist Jon Markman pointed out in a recent column, quoting the head of the ECRI: "The recent drop in the WLI growth rate means that economic ‘growth is about to slow noticeably -- not collapse.’"

With regard to whether the slow-down in growth will translate into a recession, Markman quotes ECRI's head as saying that, "the answer is not knowable yet, and won't be known for a few months, because while the downturn so far is pronounced, it is not yet persistent."

None of this guarantees that the economy won't soon slip into another recession, of course. It may very well. The slow-down in the growth rate of the ECRI WLI definitely bears watching.

My point, instead, is the contrarian advice to be skeptical of any emerging consensus, because -- to the extent it is widely held -- investors begin to act on it unthinkingly. And, that, in turn, creates opportunities for those investors who question that consensus.