Last week contained very little to alter our view that a global economic downturn is likely here. While we recognize the modest, low-level improvement in a variety of indicators (see Dodging a Bullet, from a Machine Gun ), and also estimate that recession risk is something less than 100%, this is far from a suspension of our recession concerns. To the contrary, a concerted global downturn that includes the U.S. remains the most likely outcome.
Last week, the Conference Board released its revised version of Leading Economic Indicators, which shows a sharply weaker trajectory than the former version if the LEI. Indeed, the revised LEI has already turned down, though to a lesser degree than just before previous recessions.
In early 2010, we examined the seasonal adjustment factors used by the Bureau of Labor Statistics in the monthly employment report (see Notes on a Difficult Employment Outlook ). While we didn’t observe any striking divergences between the BLS adjustment factors and our own estimates, I noted that the effect of those seasonal adjustments typically amounted to anywhere between +1.9 and -1.3 million jobs, depending on the month. Presently, we estimate that the effect of these adjustments range between +2.1 million and -1.1 million jobs in any given month (see When Positive Surprises are Surprisingly Meaningless ). These are strikingly large numbers compared with the typical range of forecasts that often surround the monthly employment numbers.
Think of it this way – if there is typically a great deal of temporary job creation in the fourth quarter of the year (and there is), the effect of seasonal adjustment will be to subtract off a certain proportion of actualemployment in order to smooth that bulge down. Accordingly the October-December adjustment factors range between -0.6% and -0.8% of total non-seasonally adjusted employment. In contrast, if there is a great deal of job destruction in January and February (and there is), the effect of seasonal adjustment will be to add back some amount of phantom employment, amounting to between 1.1% and 1.6% of total nonfarm payroll jobs.
Given that virtually all economic series undergoes some amount of seasonal adjustment, it isn’t difficult to see how the extraordinarily weak economic data in late 2008 and early 2009 may have produced anupward bump in a wide variety of seasonal adjustment factors for data around the turn of the year, adding to the short-term noise we’re already observing in various economic series. In any event, even without any skewed seasonal factors, the broad ensemble of leading economic evidence remains unfavorable here.
Finally, while we typically discourage drawing inferences from any single indicator, it’s at least worth noting that with the release of Q4 GDP figures, the year-over-year growth rate of real U.S. GDP remains below 1.6% (denoted by the red line below). A decline in GDP growth to this level has always been associated with recession, usually coincident with that decline, though with a two-quarter lag in two instances (1956 and 2007), and with one post-recession dip in growth during the first quarter of 2003. As it happens, the GDP growth rate dropped below 1.6% in the third quarter of 2011.
Given the strong and rather obvious relationship between the most recent year-over-year rate of GDP growth and the prospect of oncoming recession, it’s difficult to understand why Wall Street so completely rejects the likelihood of an economic downturn. Then again, that’s exactly why we’re expecting a Goat Rodeo.