Dollar, Yields Fall As Jobs Report Misses

And the numbers are in.

i love financial networks with this headline: JOBS REPORT IMMINENT

like it's a fucking incoming meteor

— Heisenberg Report (@heisenbergrpt) September 1, 2017

Estimates and priors:

• Change in Nonfarm Payrolls, est. 180,000, prior 209,000
• Change in Private Payrolls, est. 171,500, prior 205,000
• Change in Manufact. Payrolls, est. 7,500, prior 16,000
• Unemployment Rate, est. 4.3%, prior 4.3%
• Average Hourly Earnings MoM, est. 0.2%, prior 0.3%
• Average Hourly Earnings YoY, est. 2.6%, prior 2.5%
• Average Weekly Hours All Employees, est. 34.5, prior 34.5
• Labor Force Participation Rate, prior 62.9%
• Underemployment Rate, prior 8.6%

Actual:

• U.S. Aug. Nonfarm Payrolls Rose 156k; Unemp. Rate at 4.4%

• Nonfarm payrolls, net revisions, -41k from prior two months
• Participation rate 62.9% vs prior 62.9%
• Avg. hourly earnings 0.1% m/m, est. 0.2%, prior 0.3%; Y/y 2.5%, prior 2.5% est. 2.6%

• Nonfarm private payrolls rose 165k vs prior 202k; est. 172k, range 140k-228k from 38 economists surveyed
• Manufacturing payrolls rose 36k after rising 26k in the prior month; economists estimated 8k, range -10k to 15k from 22 economists surveyed
• Unemployment rate 4.4% vs prior 4.3%; est. 4.3%, range 4.1%-4.4% from 79 economists surveyed
• Underemployment rate 8.6% vs prior 8.6%
• Change in household employment -74k vs prior 345k

And the knee-jerk reaction is predictable, yields and the dollar down:

******************************************

Here is the context, excerpted from the piece we posted early this morning….

Well, it’s jobs day and there are a couple of ways you can approach the August print.

It does of course come at a particularly delicate time for yields and the dollar which have been beset by worries about the Trump agenda and about the looming debt ceiling debate.

10Y yields fell by the most since Brexit last month…

…and the dollar has fallen for six months in a row.

A big miss could take whatever wind was left in the sails out, and in that regard this a pretty important set of numbers. Especially as it comes just ahead of the Fed and the ECB, with the latter looking desperately for signs that the policy convergence theme that’s driven the euro to nosebleed levels will drift back towards the policy divergence theme that prevailed headed into 2017. Simply put: the last thing Draghi needs is another blockbuster Friday for the single currency and a big miss on the jobs number could catalyze just that.

On the other hand, Bloomberg’s Mark Cudmore thinks you might as well just stay in bed or, if you’re across the pond, go get drunk. To wit:

Ignore this Friday’s labor data from the U.S. Through years of trading the NFP number, I am firmly in the camp that going to the pub (if in Europe) or staying in bed (if in the U.S.) is the most profitable strategy around a complete lottery ticket of a data release. I have some evidence to back up this complacent attitude.

Looking through the eight previous releases this year, the market reaction seems to be completely independent of whether the data beats or misses. And it doesn’t matter whether you look at the wage data or the headline payrolls number. The S&P 500 has finished up on seven of the eight releases even though wages only beat twice and the headline missed on three of the five. For trading equities on payrolls day, the question is more about whether the average marginal investor is looking to add or reduce exposure. After a long bull market, the answer to that question may be starting to shift negatively as volatility picks up and credit spreads widen out. But whichever way you answer that question, you can do it from the pub/bed and you don’t need to come to work.

So obviously that’s an awesome assessment but I’m not sure Mark is as heavy a drinker as Heisenberg. Because when Heisenberg is at “the pub,” Heisenberg drinks like Richard Pryor – that is, “until the bartender says ‘we ain’t got no more fuckin’ liquor!”

Speaking of bartenders, they’ll be plenty of focus on how many jobs were added in food services. When you see that theme (“it’s all waiters and bartenders”) parroted do remember this: waiters and bartenders make a lot more money than you think they do and a non-trivial percentage of tips are paid in cash. If you think those waiters and bartenders are reporting those cash tips to the I.R.S. then you are out of your fucking mind. So before you go and malign “the waiters and bartenders economy”, make sure you know the difference between “Flo” at the Waffle House down the street and “Jack” the head mixologist at the trendy bar in the city who just bought a new Ghibli with money he’s not paying taxes on. I mean obviously that’s not an ideal situation for anyone other than “Jack,” but the point is, you can like it or not and the government might be getting the short end of the stick on tax receipts, but this is increasingly a services-based economy and a whole lot of the people providing those services are doing a lot better than you might imagine. So maybe it’s not the worst thing in the world that yesteryear’s assembly line workers are now pouring drinks (sorry Steve Bannon).

Everyone will of course be watching the AHE number for signs that the supposedly overheating labor market is finally leading to wage inflation. “A decent payrolls number today would be the icing on the cake in a week that has seen some positive signs that the U.S. economy may be in better shape than was previously thought prior to Jackson Hole,” analyst Michael Hewson at CMC Markets wrote in note out Thursday, adding that “annual hourly wage growth is currently 2.5%, a little on the weak side for an economy supposedly at full employment, so a strong number here could increase the odds of another rate rise this year, most likely in December.” Here’s Goldman on that:

We expect a weak reading for August average hourly earnings at +0.1% month-over-month. This would leave the year-over-year rate unchanged at +2.5%. Our forecast reflects a 0.1-0.2pp drag from unfavorable calendar effects, more than offsetting the continued firming in labor markets. Early payroll survey weeks are historically associated with weaker average hourly earnings growth. These effects may have to do with the intra-month timing of wage payments, which US Treasury Statement data suggest are elevated around the 1st and 15th of each month. Pay periods that are relatively early may miss some of the later payments. In months when the Saturday of the reference week falls on the 12th of the month–as is the case in August 2017– average hourly earnings growth tends to be particularly weak. August 2017 also had three additional workdays relative to July, which we believe is also a negative factor, because paid wages may be “stickier” than hours worked in the BLS data.

Speaking of Goldman, they think this is going to be a disappointment, and here’s why:

August residual seasonality. There has been a trend toward weak first-print payroll growth in August, which may reflect a recurring seasonal bias in the first vintages of the data. In Exhibit 1, we show first-print payroll growth in August relative to consensus estimates and relative to the previously published three-month moving average (i.e., the average in May, June, and July, as published in the July employment report). Payroll growth has been below consensus in each of the last six years and below the three-month average in each of the last seven (consensus for today’s report is 180k and the 3-month average is currently 195k). This August weakness has tended to occur in many of the same industries, including manufacturing, professional services, retail, and information. We estimate the magnitude of the potential bias at roughly 30-50k (for headline payroll growth).

Right. The old “residual seasonality” discussion. I’ll spare you any further color on that, because if you took Cudmore’s advice and went to the bar, talking about residual seasonality would be a real buzz kill.

Look, that’s really all you need to know. Stay tuned. The number will be out shortly.