Yellen's "Favorite" Labor Indicator Tumbles: Job Openings Drop Most Since 2009

One month ago, when the JOLTS data showed the highest number of job openings in history, rising by a near record 430,000 to 5.8 million vacant positions, most speculated that this data - considered to be Janet Yellen's favorite indicator of slack in the labor market now that the unemployment rate has become utterly unreliable due to the 94 million Americans out of the labor force - had sealed the fate of a September rate hike.

It did not, as instead the Fed decided to shift to its 4th mandate (the 3rd being the stock market), namely the "global environment" as the reason not to tighten monetary conditions.

But while the JOLTs data turned out to be useful to "confirm" a stronger economy, even if it was roundly rejected when it was expected to be the fulcrum catalyst for monetary policy change, we are confident it will be completely ignored this month when moments ago the BLS revealed that in August, the number of job openings tumbled back down by 298,000 to 5.37 million: far below expectations and the biggest monthly drop since the 301,000 slide in March of 2009.


So was this the job openings peak of the current cycle?

The answer will have to wait until next month as the JOLTS series is consistently revised and remarkably unreliable (and frequently made up: recall it was this site that caught the BLS fabricating data back in 2013), however another series that may serve as a guide is the JOLTS total hires vs the cumulative change in payrolls over the past year. Both of these series have topped out, even as the number of hires this cycle remains far below the peak from the previous cycle, plateauing just above 5 million compared to 5.4 million during the last bubble burst, while the 12 month cumulative change in payrolls recently slipped back under 3 million after peaking at 3.2 million in February, and well above the prior cycle peak.



Worse yet, looking at the JOLTS net turnover (hires less separations) and comparing it to the monthly change in Payrolls - two series which should otherwise overlap - one of two things will have to happen: either hires will be revised sharply lower (or separations higher) or payrolls in the past few months will need to be pushed well higher. Most likely, the estimated rate of hiring will end up being revised far lower, once again confirming the labor market is slowing down substantially.


Finally, looking at the separations end of the equation, what we see is that slowly but surely even the BLS is catching up to the private sector, which recently announced the highest number of terminations among energy companies since the crisis, and sure enough, a chart of the discharges in the Midwest shows that while the trendline is now pointing decidedly higher, it will go far higher before any substantial recovery in the US shale space.


Bottom line: despite being two months backward looking, today's JOLTS report provide the latest batch of weak news coming from the US labor market, which likely explains the ongoing levitation in stocks which have officially given up on the "rate hike is good for stocks" narrative, and fallen back to what has worked for the past 7 years, namely terrible economic news being great news for risk assets.