The next big data point will be the January jobs report (released Wednesday), which will include the annual benchmark revisions – where the Bureau for Labor Statistics realigns their sample-based employment estimates with unemployment insurance population counts.
The consensus is for a 71k increase in January non-farm payrolls with the unemployment rate to hold steady at 4.4%. We are forecasting a modestly stronger outcome of 80,000, but as is always the case with the jobs data, there is very little conviction in our forecast. Jobless claims remained low during the period and while the Challenger report noted the worst January for lay-off announcements since 2009, they will be spread over a number of months. At the same time, the ISM services employment reported a second consecutive print above 50, indicating expansion, after six months of contraction. The ISM manufacturing reported the slowest pace of job losses for 12 months.
Assuming we are in the right ballpark, the bigger market reaction may come from the annual benchmark revisions. The preliminary revisions were released 9 September, and they showed an overestimate of employment of 911,000 for March 2025, equivalent to 0.6% of total employment. Over the past ten years, the average error has been 0.2%.
911,000 divided by twelve months equates to a 76,000 per month overestimate, but Federal Reserve Chair Jerome Powell suggested in December that central bank officials believe the overestimate is around 60,000 per month. That would mean an annual downward revision of 720,000, so we should probably go with something closer to that rather than thinking the 911,000 will be confirmed.