Friday, January 04, 2008
december jobs indicate recession
Until now soft landing optimists could dismiss other recessionary signals in the economy – a much worsening housing recession, faltering capex spending by the corporate sector, a severe liquidity and credit crunch, oil at $100, forward looking indicators of supply (ISM) showing contraction, a weakened consumer that was saving-less and debt burdened – based on the argument that as long as there was job generation the consumer would keep on spending and – with consumption being 72% of aggregate demand – a recession could be thus avoided.
Until now the US consumer had negative savings, was debt burdened and was being buffeted by many shocks: falling home values, falling home equity withdrawal, rising debt servicing given resetting ARMs, rising delinquencies on mortgages, credit cards and auto loans, falling consumer confidence, high and rising gasoline prices and, more recently, falling stock market wealth. But as long as income and jobs were generated at a satisfactory rate the optimists argued that all these shocks did not matter as the consumer would keep on consuming.
But the dismal employment report today – 18K jobs created, private jobs falling by 13K and unemployment rate up to 5% from 4.7% - confirms that even job and income generation is now faltering making a recession an almost sure outcome.
to be certain, BLS job weakness has been revised away once already. barry ritholtz at big picture has done a good job pointing out how the birth-death adjustment model has forced the jobs number to miss the turn down and made recent job creation more of a statistical artefact.
It has been my position that the increasing proportion of Birth Death jobs as a total percentage of NFP has rendered the BLS data meaningless. The straight line extrapolation of the B/D adjustment is, by design, unable to catch any change in economic direction. Hence, the absurd determination that 199,000 small business construction jobs have been created since January 2007.
Consider: The B/D generated 1,239,000 jobs from February thru November 2007. That's rather surprising, since the total NFP jobs created since January 2007 was 1,208,000. In other words, the Net Birth/Death jobs created over 10 months was actually greater than the total NFP jobs created in all of 2007. That's rather odd, don't you think?.
... The BLS NFP Birth/Death adjustment is the statistical equivalent of CDO pricing via mark-to-model. The idea is to approximate an unknown variables value as close to possible via measurements. The mark-to-model method is deeply, inherently flawed.
his chart on the birth-death component contribution makes it abundantly clear. this is another government-issued statistic which has been thoroughly gamed to hide reality. so surprising post hoc revisions are not at all out of the question.
but keeping an eye on employment ratio eliminates some of the distracting noise. an uptick in november to 63.0 (from the december 2006 cycle high of 63.4) seemed out of sequence, and the december report confirms as much as EMRATIO fell back to 62.7.
markets, which i consider primed for a bounce, gapped down 1-1.5% on the news and then extended losses.