Back in September, courtesy of an unprecedented discrepancy between the JOLTS "net turnovers" (or hires less separations) print, which traditionally has been the equivalent of the NFP's establishment survey monthly job additions, we highlighted just what happens [7]when the BLS has caught itself in a estimation lie, and is forced to adjusted the data set both concurrently and retroactively to correct for cumulative error.
We suggested that as a result of this public humiliation, the BLS would have no choice but to ramp up its monthly net turnovers print in order to "catch up" to what the monthly payrolls survey indicated is America's "improving" jobs picture.
Sure enough, when moments ago the latest October JOLTS survey was released, the October "net turnovers" number soared from 155K in September to a whopping 260K in October, more than eclipsing the revised NFP print of 200K job gains in October, and leading to the second highest JOLTS turnover print since February's 271K, and before that - going back all the way to the 287K in February of 2012. And yes, this was in the month when the government had shut down and the result was supposedly major, if temporary, job losses.
[8]
Today's number also means that the YTD monthly average job gain based on either the payroll data or the JOLTS survey has declined to just 24K (160K for JOLTS, 184K for NFP), the lowest average difference in 2013.
Finally, this is how the difference between the two time series on a monthly snapshost basis looks:
[9]
Why is any of this important? Because to Janet Yellen, the JOLTS survey has traditionally been an important secondary metric for the jobs market, and judging by the huge jump in implied job gains, if indeed the Fed was in a tapering mood, the December FOMC meeting looks increasingly like the day when a Taper may be announced. Of course, that ignores how a very illiquid market would react, and is perhaps the reason why December's final, massive double POMO is on the day just after the FOMC announcement.