The distribution of guesses for tomorrow's "most important payrolls print ever" or at least until next month, skews modestly to the upside after the biggest spike in ISM employment ever this week jarred some economists to become more optimistic, and side with Goldman Sachs expecting a Fed-inspiring drop in the unemployment rate, rise in average hourly earnings, and better than expected payrolls of 190K. As a result, while consensus expects a NFP rebound from 151K to 172K, the whisper number is around 200k. Anything above this would send December rate hike odds surging to the all important 70% or above, bond yields spiking and equities at the mercy of whichever way the risk parity machines were calibrated tomorrow.
Others disagree: Southbay Research is leaning on the bearish side, nothing the following positive and negative factors ahead of tomorrow's report:
- ISM Non-Mftg: Surges to 57.1 from 51.4
- ISM Mftg: Up to 51.5 from 49.4
- PMI Services: Up to 52.3 from 51
- Chicago PMI: Up 54.2 from 51.5
- ADP: Payrolls drop (Forecast September 155K)
- PMI Mftng: Slips to 51.5 from 52
- Construction: drops -0.7% versus -0.3% prior month
- Corporate Profits 2Q: -1.7% y/y versus -2.2% prior quarter
- Durable Goods Orders (ex Trans): -0.4%
- ISM Non-Mftg: Employment Index surges to 57.2 from 50.7
- ISM Mftg: Employment Index ticks up to 49.7 from 48.3
- PMI Services: Weakest employment levels in 4.5 years
- PMI Mftg: Employment consistent with 115K jobs
It is worth noting that conflicting data is the hallmark of inflection points. Furthermore, as Southbay notes, "Forget Manufacturing, Focus on Services." Here is the full bearish case:
Look past manufacturing data for two reasons: (1) no real news here – manufacturing continues to be stuck in a low gear and (2) manufacturing doesn't factor much into September payrolls.
Services will drive the September payrolls. Specifically Leisure & Hospitality jobs.
As Summer vacation ends, restaurants, hotels and recreational hot spots wind down and layoff seasonal workers. The nominal level of layoffs is driven by exactly two things: the number of workers added and the amount of customer foot traffic.
A Good reason to Expect Heavier Layoffs: On a trailing 12 month basis, the Leisure Hospitality sector has added about the same number of jobs as the prior year.
What is different this year is that Restaurant operators have lost confidence.
According to the National Restaurant Association, the Restaurant Performance Index turned negative in August. The last time it was negative was February: that kicked off the weakest level of Restaurant payrolls since 2012.
Sales expectations remain weak and barely expansionary. Consequently staffing expectations have shifted into contractionary territory
Expect layoffs to at least equal, if not exceed, last year's.
Bottom line: The biggest driver for September Payrolls is Restaurant payrolls, and they are looking weak.
Whatever the actual number, expect the consensus to be vastly wrong as the last few months have been "volatile" outliers to say the least:
Going back to the bullish, "whsiper" outlier, Goldman expects a 190k increase in nonfarm payroll employment in September, above consensus expectations for a 172k gain, and up from their preliminary forecast of 175k. Although payroll growth slowed to 155k last month, subdued employment gains are not uncommon in August, and the trend growth rate in payrolls still looks solid, with the trailing 3- and 6-month averages at 232k and 175k, respectively.
Some observations from GS:
Their above-consensus payroll forecast primarily reflects improving underlying labor market fundamentals during the course of the month. Initial jobless claims continued trending down towards post-crisis lows, and nearly all other employment indicators from the various regional and national manufacturing and service sector surveys turned up. The ISM non-manufacturing survey’s employment index had its largest gain on record. In addition, we expect a modest rebound in employment in Louisiana, which fell by nearly 8k last month, likely due to adverse weather conditions. Offsetting these improvements, we look for a decline in employment in the education sector (specifically, private education services and state & local government education employment), which has been growing at a roughly 30k pace over the last three months, well above the +8k monthly average since 2011. Employment in the education sector has been highly volatile during September in recent years (Exhibit 1), and monthly gains have averaged -7k since 2011.
Arguing for a stronger report:
- Job availability: The Conference Board’s labor differential—the difference between the share of consumers saying jobs are plentiful vs. hard to get—rose 2.3pt to +6.3, reaching a new post-crisis high
- Jobless claims: The four-week moving average of initial jobless claims during the survey week fell by 7k to 258k. The continued downtrend in initial claims suggests that layoff activity across the economy remains minimal. Insured unemployment, or continuing claims, fell by 81k between the August and September survey weeks, and currently sits at post-crisis lows.
- Service sector surveys: On balance, the employment components of service sector surveys improved in September. The ISM non-manufacturing report’s employment index registered its largest increase on record, rising 6.5pt to 57.2. The NY Fed (+5.3pt to +8.1, not seasonally adjusted) and Philly Fed (+9.3pt to +19.3) surveys also rose. Offsetting these increases were declines in the Richmond Fed (-7pt to +6) and Dallas Fed (-1.4pt to +4.4) surveys. Service sector employment rose 150k last month, and has increased 164k on average over the last six months.
- Manufacturing surveys: Almost all of the employment components of the various monthly manufacturing surveys improved in September. The ISM manufacturing (+1.4pt to 49.7), Philly Fed (+14.7pt to -5.3), Dallas Fed (+7.3pt to +2.3) and Kansas City Fed (+7 to -3) rose, while the Richmond Fed (-20pt to -13) declined. Manufacturing employment fell by 14k in August, and has declined by 7k on average over the last six months.
Arguing for a weaker report:
- ADP: ADP reported a 154k gain in private payroll employment in September, below a revised +175k increase in August. Service-sector job gains softened a bit to 183k, manufacturing employment was flat, and construction employment fell 2k. Despite the softening in ADP employment, we have found this to have somewhat mixed value as an input towards forecasting nonfarm payrolls.
- Online job ads: The Conference Board’s Help Wanted Online (HWOL) report showed a small 0.4% increase in new online job ads in September, while total jobs listed fell by 1.9%. However, we put only limited weight on this indicator at the moment in light of recent research by Fed economists that argued that the HWOL ad count—which has departed significantly from the job openings figures in the official Job Openings and Labor Turnover Survey (JOLTS)—has been influenced by price increases for online job ads.
- Job cuts: According to the Challenger, Gray & Christmas report, job cuts rose 17.4% on a seasonally adjusted basis in September, although they remain roughly 25% lower than their year ago levels.
We expect the unemployment rate to decline to 4.8% in August from an unrounded 4.922% in August. The headline U3 rate was unchanged in August but is up from a low of 4.7% in May, while the broader U6 underemployment rate held steady at 9.7%. The household survey showed a modest 97k increase in employment versus a 176k increase in the labor force. The labor force participation rate remained at 62.8%, close to where it has been since the beginning of the year.
Average hourly earnings for all workers are likely to rise 0.3% in August, in large part reflecting favorable calendar effects. As a result, we expect the year-on-year rate to rise to 2.7% from 2.4%. The broader wage data remain encouraging: our wage tracker—which aggregates four measures of wage growth—stands at 2.6% year-on-year, a sign that diminishing slack is boosting wage growth.
Separately, many commentators have highlighted this year’s decline in average weekly hours worked in the establishment survey, fearing that it signals weakness in labor demand. The decline in average hours appears to be relatively widespread across states and industries (Exhibits 2 & 3).
Although it is hard to find a basis for dismissing the decline in hours worked, the series can be a bit noisy on a month-to-month basis. Furthermore, falling hours appear at odds with fundamental payroll growth figures and other labor market indicators over the same horizon, and we thus would hesitate to take an overly negative signal from these series as of yet.
* * *
And finally, if you needed any advice on trading the print, it's simple – no matter what the print – wait until US markets open and buy it all with both hands and feet: