Little good can be said about the August jobs report. The 156k growth in jobs disappointed and is well below the recent averages. The back two months were revised lower for a total of 41k jobs. The unemployment rate ticked up to 4.4% even though the participation rates were unchanged at 62.9%. 

As if that was not sufficiently disappointing, the work week slipped a to 34.4 hours from 34.5 hours and average hourly earnings did not rise as much as expected (0.1% instead of 0.2%). One of the few bright spots was manufacturing, which added 36k jobs. This is the most since August 2013. The July manufacturing payrolls were revised to 26k from 16k. 

The market has responded as one would imagine. The dollar was sold, and Treasuries bought. That is what markets do, but for policymakers, it is a different story. The employment data is important, but it is noisy, and the broader indications are of a labor market has fared better than the Fed had expected at the end of last year. This year’s average jobs growth slipped to 176k from 184k. It averaged 187k last year.   

The Fed has tried to position it balance sheet operations as, once they start, being on the almost automatic pilot, not subject to the vagaries of high-frequency data. This means that expectations for the Fed to announce that it will begin allowing its balance sheet to shrink starting in Q4 should not be impacted by today’s employment report.  

The dollar’s sell-off in responses to the data was reversed in response to reports that suggested that the ECB may not decide on its balance sheet until December.  Many had expected an announcement next week, though some thought October. We had seen a September announcement as the most likely venue given the desire to 1) prepare the market and 2) have new staff forecasts in hand. We had thought that a dovish tapering could be signaled by a cut in next year’s CPI forecast. We thought having the purchases to 30 bln would maximize the ECB’s flexibility to end the program in the middle of next year.