Looking for Recession Signals Where There Are None

Reading the economic tea leaves to divine the start date for the current recession (if there even is one) has become to economy watchers what box scores and injury reports are to sports gamblers.

Putting aside the very valid question of whether such hand-wringing even makes sense, the conventional wisdom points to December as the first month of the recession -- right when payrolls started their decline.

For his part, Merrill Lynch's David Rosenberg, the gloomiest of Wall St. economists (save perhaps for the team at BNP Paribas), has said in the past that the U.S. economy turned recessionary in January. This morning, he writes in a note that the recession in G.D.P. is finally here:

but it's in the monthly data, not the quarterly data. The MacroEconomic Advisers monthly database shows that real GDP dipped at a 0.5% annual rate in April and has contracted now in two of the past three months.  Note that since January, the month we had been saying for some time that represented the peak of the business cycle, real GDP has declined at a 2.2% annual rate.  So, do not be fooled by that 0.9% first quarter GDP print - it is masking an erosion in activity beneath the veneer of quarterly averages.

But there are many problems with this statement.

First, it's not at all unusual for the Macroeconomic Advisers monthly GDP figures to turn negative. Since April 1992, there've been 131 months with positive economic growth, but also 61 negative-growth months (56  ex-recession).

Second, the number of times outside of a recession that the monthly GDP numbers have been negative "two of the past three months" since 1992 is 25.

And third, four-month 2 percent declines (at an annual rate) of GDP outside of recessions have happened 4 times since 1992.

All this isn't to say that we're not in a recession, other analysis suggests we are. But reading into monthly GDP figures like this isn't the way to find out.