ADP’s Job Data Reliability? Not Great . . .


Yesterday’s selloff was triggered in part by a weak ADP report, and fears of a broader economic slowdown.

Lets see if we can navigate the crosscurrents here to discern what, if anything, is happening.

First off, the economy is slowing. At least, the 2nd derivative rate of growth is throttling back, from over 3% GDP in several 2010 Qs to under 2% now. That much we know, and it lends some credence to the belief that QE3 is inevitable. I am far more uncertain about QE3; it really depends how slow things get and how much the Fed panics. But we do know QE2.5 — the reinvestment of the rolling maturities of QE2 — is a done deal, and that is somewhat supportive of equity prices.

Second, we also know the ADP numbers are suspect at best (See this, this and this). They seem to do better at forecasting BLS numbers later in the economic cycle when job creation is dominant at established, larger firms — their client base — than in the early parts of the cycle when it is the new start-ups driving job creation.

ADP also does better following the revisions and benchmarking — suggesting that they may be more accurate than they appear when the payroll numbers are first released.  That is almost a non-issue, as the anticipation of BLS numbers, warts and all, is what drives trading.

Last, we have two cyclical factors fighting it out: The negative “Sell in May” theme is up against the positive “Presidential 3rd year” factor.

The bottom line remains this is a challenging set of conditions to navigate. Our 30% cash position reflects that . . .


ADP: Stick to Your Knitting (July 10th, 2006)

ADP Goes Back to the Drawing Board (January 7th, 2009)

ADP Report Is Worse At Predicting Payrolls Than A Consensus Estimate (February 4th, 2011)

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:

Posted Under