You are here

Payrolls Preview: Blame Weakness On Weather, Strength On Trump

With all eyes likely on wage growth indications in the subtext of tomorrow's payrolls report (following The Fed Minutes' comments on full employment), Goldman Sachs is forecasting a better-than-expected 0.3% rebound in average hourly earnings (helped by more favorable calendar effects) and a better-than-expected 180k payrolls print (albeit with a small rise in the unemployment rate). However, they are careful to note that any downside can be blamed on "a considerable drop in temperatures."

As Goldman Sachs details:

We forecast that nonfarm payroll employment increased 180k in December, after an increase of 178k in November and 142k in October. On balance, labor market indicators were moderately strong in December, with improvement in the employment components of many service-sector and manufacturing surveys and a rise in consumer confidence to a 15-year high. The key labor market subcomponent of consumer confidence also hovered near its post-crisis high, despite a modest pullback from November. We also expect above-trend payroll growth in the transportation and warehousing industry, driven by elevated hiring related to strong online holiday shopping. On the negative side, initial jobless claims drifted higher and continuing claims posted the largest survey-week-to-survey-week increase in nearly a year. December’s relatively cold payroll survey week could also constrain payroll growth in weather-sensitive industries such as construction, particularly after the warmer-than-usual November.

We do not expect the end of the election to be a major factor this month. If election-related uncertainty reduced or delayed hiring plans earlier in 2016, the end of the election could potentially catalyze a reacceleration in hiring, as we’ve documented in research studying past elections. However, the reacceleration historically tends to occur over several quarters, and perhaps more importantly, we would not characterize the post-election landscape as one of reduced uncertainty. We will study the industry composition of tomorrow’s reports for signs of an impact from the election, such as above-trend growth in energy or financial services, or below-trend growth in healthcare or import-reliant subindustries.

Arguing for a stronger report:

  • Service sector surveys. Most of the employment components of service sector surveys improved or remained at encouraging levels in December. The Philly Fed non-manufacturing employment index rose most dramatically to an 18-month high (+5.4pt to +19.7), and the New York Fed index increased to a 9-month high (+1pt to +12.0, SA by GS). The ISM non-manufacturing employment component was disappointing, falling to 53.8 from 58.2, but remains at a level consistent with moderate growth in service-sector employment. Meanwhile, the Richmond Fed (-1pt to +12.0) and Dallas Fed (-1.4pt to +7.8) measures pulled back modestly to levels also consistent with expansion. Service sector payroll employment increased 139k in November and has increased 177k on average over the last six months.
  • Manufacturing sector surveys. The employment components of manufacturing surveys were somewhat stronger in December. The ISM manufacturing employment component rose to an 18-month high (+0.8pt to 53.1), and the Philly Fed (+9.0pt to +6.4) and Kansas City Fed (+9pt to +10) employment components both improved sharply. On the negative side, the Dallas Fed (-7.4pt to -2.9), Richmond Fed (-6pt to -1), and Empire State (-1.3pt to -12.2) employment components all declined into contractionary territory, and the Chicago PMI employment component remained flat. Manufacturing payroll employment declined by 4k in the November report, and has declined by 3.5k on average over the last six months.
  • Job availability. The Conference Board labor differential—the difference between the percent of respondents saying jobs are plentiful and those saying jobs are hard to get—declined modestly to +4.4 from a cycle high +6.6 November. Despite the pullback, this measure has risen 5pt year-over-year and remains 8pt above its 2015 average, suggestive of a strong report. The Conference Board labor differential is historically correlated with economy-wide hiring rates (JOLTS), which slowed during 2016 and have yet to rebound as of October (see Exhibit 2).
  • Transportation Jobs. Transportation and warehousing payrolls have seen elevated readings in recent Decembers (Exhibit 1), as the seasonal adjustment factors have appeared to lag the secular shift toward online holiday sales. We also note the possibility that the seasonal factors have been anchored to some extent by a particularly weak reading in December 2009 (-52k, compared to an average of -10k in the preceding three months and +12k in the subsequent three months). Given the generally positive anecdotes around online holiday shopping, we expect another year of above-trend growth in this industry, which should provide a boost to headline payrolls of roughly 10k relative to its recent trend.

Exhibit 1: Online Holiday Shipments May Boost December Transportation Payrolls (Again)

Source: Department of Labor, Goldman Sachs Global Investment Research

  • Hurricane Matthew. We also see a small amount of upside risk from continued bounce-back from Hurricane Matthew, which hit the East Coast in October. Employment across East Coast states in the three sectors that we find are most sensitive to weather– retail, construction, and leisure and hospitality – declined by a total of 17k in October, compared to an average monthly increase of 15k over the prior six months. The November rebound in these state-level industries was fairly lackluster, with +10k total gains across that state-industry grouping. Accordingly, we see a possibility of above-trend December growth (or an upward revision to November) to restore employment levels toward their pre-hurricane trend.

Arguing for a weaker report:

  • Jobless claims. Initial claims for unemployment insurance benefits moved higher, averaging 261k during the 5 weeks between the November and December payroll survey periods, compared to 253k for the November report. Continuing claims also showed its largest survey-week-to-survey-week increase (+53k) in nearly a year (January 2016). Jobless claims can be difficult to seasonally adjust around this time of the year, and we partially discount the data accordingly. Some of the rise in initial claims may also reflect a return toward mid-year levels after especially low results in late-Q3/early-Q4.
  • ADP. The payroll processing firm ADP reported a 153k gain in private payroll employment in December, down from +215k in November and somewhat below expectations of +175k. In the past we have found minimal incremental predictive power for nonfarm payrolls in the ADP report, with the notable exception of large surprises/deviations. The new methodology ADP introduced in October creates some additional uncertainty around the translation of the somewhat softer ADP data into the outlook for tomorrow’s nonfarm payroll report. Notably, ADP’s measure of healthcare employment has slowed in recent months – even more so than official BLS estimates – and in December rose just 26k, matching the two-year low growth rate achieved last month.
  • Temperatures. The December payroll period featured few notable storms, but saw a considerable drop in temperatures following much warmer-than-usual weather in early and mid-November. Accordingly, December’s relatively cold payroll survey week could constrain job growth in weather-sensitive industries. For example, we anticipate some payback in the construction industry, which has averaged +20k payroll growth in the last three months, compared to trend monthly growth in the 10-15k range.

Neutral Factors:

  • End of the election. We do not expect the end of the election to be a major factor this month. If election-related uncertainty reduced or delayed hiring plans earlier in 2016, the end of the election could potentially catalyze a reacceleration in hiring, as we’ve documented in research studying past elections. However, the reacceleration historically tends to occur over several quarters, and perhaps more importantly, we would not characterize the post-election landscape as one of reduced uncertainty. Payroll growth did slow during the course of 2016, consistent with the possibility that election-related uncertainty reduced or delayed hiring plans, though dwindling labor market slack represents a second viable explanation. More granular data from JOLTS show a decline in the pace of hiring since the beginning of the year, with a further meaningful drop in hiring in September and October ahead of the election (Exhibit 2). We would note that considerable policy uncertainty remains. For example, the anticipation of policy changes related to health insurance and destination-based taxation could slow hiring in the healthcare sector and in import-reliant industries such as retail, wholesale, and parts of manufacturing. Much like the policy outlook itself, the evolution of business-sector expectations and their implications for aggregate-level job growth remain open questions. In terms of short-term hiring in election-related job categories, note that the BLS makes a special adjustment to estimate and remove these effects in election years, and November payrolls showed no pronounced spike in marketing research categories or government.

Exhibit 2: Were 2016’s Slower Hiring Rates Election-Related?

Source: Department of Labor, Goldman Sachs Global Investment Research

  • Online job ads. The Conference Board’s Help Wanted Online (HWOL) report reversed most of last month’s decline, and stands 9% lower than levels last year (vs. -15% yoy in November). However, we put limited weight on this indicator at the moment in light of research by Fed economists that argued that the HWOL ad count has been depressed by higher prices for online job ads.
  • Job cuts. Announced layoffs reported by Challenger, Gray & Christmas after our seasonal adjustment increased by 7k to 34k in December, but the level of announced layoffs remains range-bound and not far from cycle-lows.
  • Seasonals. Since 2010, December payroll growth has surprised positively relative to consensus two thirds of the time (4/6), however the average surprise is slightly negative at -3k.

We believe the unemployment rate most likely rebounded one-tenth to 4.7% after a 0.3pp drop last month. The unemployment rate fell 0.3pp to an unrounded 4.64% last month, but the decline was mainly driven by the participation rate, as opposed to an acceleration in household employment (which nonetheless rose a respectable 160k in the month). Sharp changes in the participation rate often reverse, and the fact that the second decimal nearly rounded up, we believe a modest rebound to 4.7% is more likely than another 4.6% reading.

Average hourly earnings likely rose 0.3% after a 0.1% decline in November, as the December survey period ended on the 17th, a relatively late calendar that is historically associated with average or favorable earnings growth. The year-over-year rate is likely to accelerate from 2.5% to 2.8%, which would match the cycle high. Our wage tracker, which captures the broader trend in wage growth across four major indicators, stands at 2.8% year-over-year as of Q3.

Tomorrow’s employment report will be accompanied by the annual revision to the household survey, with potential revisions to the last 5 years of seasonally adjusted household data. Annual revisions to the establishment survey – as well as the incorporation of updated population estimates from the Census into the household survey – are scheduled for February 3rd. The revisions to the household survey accompanying the December report are usually minor, and last year’s revisions did not result in a change to any of the monthly unemployment rates for 2015.

*  *  *

And Ransquawk details the potential market reaction:

As ever with the NFP release, the headline is likely to garner much of the initial focus with algorithms and fast money moves jumping on any large discrepancies, participants must be aware that moves could be exacerbated and choppy trade likely due to the beginning of January's historic thin trading conditions.

 

If there is an overwhelmingly strong report, the USD will strengthen across the board however, as the dollar index continues to ramp many analysts consider the dollar upside to be limited and the possible move to lookout for is a poor report, negatively affecting the USD.

 

As focus detatches from the Fed and interest rates, equity markets are likely to be a much cleaner trade than recent months. Furthermore, with whispers of a reversal in equity markets and many participants still long the market an overwhelming poor report may prove to be a slight catalyst in a bearish push. In terms of technical levels in the S&P 500, November's high at 2214.10 can prove to be some support on the downside and there is an internal downtrend line which originates on 23/08/16 to the next lower high on the 07/09/16 and support likely at 2158.20.

 

Gold and treasury markets are likely to act in tandem with classic price action in regards to the NFP likely to be evident in flight to safety asset classes, with a strong report likely to cause some risk on sentiment resulting in weakness in both gold and treasury markets and vice versa for a weak report. Participants are likely to keep an eye on fixed income markets, with tight trading ranges evident across the curve throughout December and any discrepancy in either direction could result in some traction.

And if that fails Buy The Fucking Payrolls Dip no matter what...