I’m expecting 130K on total payrolls and 135K on private, so I’m below consensus, which Bloomberg puts at 153K for the total. I expect the unemployment rate to maybe tick up a tenth.
And that’s all the time we should spend forecasting these volatile monthly numbers. More interesting, I think, is the question of why everyone gets so wrapped up in the monthly numbers when the actual signal-to-noise ratio they yield is not that high. The 90% confidence interval around the payroll number, for example, is about 100,000, meaning there’s a 90% chance that the actual change in payrolls in a given month is that much higher or lower than the reported change. So, given last month’s initial print of 88,000, we can’t rule out the possibility we actually lost jobs.
Well, one reason for all the attention, courtesy of GS analysts, is that the jobs report doesn’t just move markets. It moves markets far more than any other economic indicator. The figure below plots how equity futures and the yield on the 10-year T-bill respond to upside surprises in the various indicators we all pour over.
“NFP”—nonfarm payrolls—is a huge outlier. A stronger-than-expected payroll number has a much larger effect on market prices than anything else, especially on the fixed-income side of the market.
So, given the relatively weak signal from the monthly jobs report, does this imply market irrationality? Not necessarily. It shows that the jobs report is important because people making trades think it’s important.
More substantively, even a weak signal contains information, and when you combine that signal with others—GDP, prices, retail sales, industry reports, trade data—you get more explaining power. Right now, most of the variables on that list are disappointing, not to mention last month’s weak jobs number itself. That’s why I’m a bit under consensus.
Back to you on this in the AM, shortly after the 8:30 release.