Don’t forget: tomorrow is payrolls Friday.
There’s a lot going on geopolitically speaking what with the Trump administration essentially telegraphing military action in Syria and the President’s first meeting with Xi Jinping being dominated not by trade talk or currency manipulation but by the threat emanating from Pyongyang.
Meanwhile, fresh doubts were raised about the underlying health of the US economy earlier this week when auto sales numbers disappointed “bigly.” “Car trouble” aside, we did get a blockbuster ADP print on Wednesday which, all else equal, would seem to bode well for Friday’s NFP print.
To be sure, tomorrow’s jobs data comes at an interesting time for markets. 10Y yields are all anyone wants to talk about as it seems Treasurys are one headline away from rallying hard and simultaneously driving a stake through the heart of the reflation narrative. The absence of a sell-off in USTs following the ADP print and the subsequent move lower in yields following the release of the Fed Minutes just served to underscore how precarious the situation truly is.
So that’s the setup. Below, for those interested, find Goldman’s March payrolls preview.
We estimate nonfarm payrolls increased 170k in March, compared to consensus of +180k and February’s 235k increase. Our forecast reflects above-trend underlying job growth that is more than offset by a significant drag from unseasonable winter weather, which we believe may have boosted February payroll growth by 30-50k and could weigh on March growth by as much as 30-60k. We are not assuming a significant impact from the federal hiring freeze implemented in late January, which did not appear to materially affect the February report.
We estimate that the unemployment remained stable at 4.7%, as upside risk from a potential pause in household employment growth is offset by some scope for a pullback in the participation rate, following its sharp year-to-date increase. We also estimate average hourly earnings increased 0.2% month over month and 2.7% year over year, reflecting tightening labor markets offset by slightly negative calendar effects.
Labor market fundamentals remain encouraging, as March exhibited further improvement in manufacturing employment surveys, a new cycle high for the Conference Board labor market differential, and continued low readings of initial jobless claims (despite an uptick in the final week). However, we believe the sharp drop in temperatures and the early-month winter storms will depress payroll growth in weather-sensitive categories. Winter Storm Stella impacted the Midwest and East Coast early in the payroll survey week – and much of the snow that accumulated during the storm did not melt until Thursday or Friday. We believe the weather impact could be particularly large in comparison to February, which was marked by unseasonably warm weather and limited snowfall.
ADP. The payroll processing firm ADP reported a 263k rise in private payroll employment in March, a solid pace that substantially exceeded expectations of +185k. In past research, we have found that large surprises in the ADP report tend to be predictive of a subsequent nonfarm payroll surprise. Offsetting this consideration, ADP’s February estimates were revised down significantly, and we believe some of the March strength reflects the lagged impact of accelerating nonfarm payroll growth (as well as the other financial and economic inputs to the ADP model). These factors, as well as differences in data collection methodologies, may be partially masking the impact of Winter Storm Stella, which we expect to manifest more visibly in tomorrow’s employment report. Nonetheless, the ADP data clearly imply a solid pace of underlying job growth.
The ADP report also showed another above-trend increase in construction (+49k) but a surprising decline of 32k in education employment. In the revamped ADP model, growth in these subindustries show strong correlations with their nonfarm payroll counterparts (0.74 and 0.75 since 2011, respectively). While some of the accuracy may reflect the impact of methodological revisions to the ADP regression model, it’s worth noting that the first vintage ADP reports also successfully anticipated the weak education payrolls growth in October 2016 as well as the strong construction job gains last month.
We estimate the unemployment rate remained stable at 4.7% following the one tenth drop last month (to 4.703% unrounded). The pace of household employment growth picked up sharply to start the year, with a 905k cumulative increase over the past two months (population-adjusted). Offsetting this strength, the participation rate also moved higher, rising three tenths over the same period to 63.0%. Given the possibility of mean reversion in either household employment or the participation rate following the sharp year-to-date moves, we are left without a strong view about tomorrow’s unemployment rate, though our medium-term expectation envisions a further decline.
Finally, we expect average hourly earnings to increase 0.2% month over month and 2.7% year over year, reflecting the interaction of firming wage growth with slightly negative calendar effects. The March payroll period ended on the 18th – the latest day possible and same as last month. Based on our model, this calendar is associated with average or slightly below average wage growth.