This wasn’t supposed to happen. The bond market shouldn’t have cared about today’s jobs data. The Fed had already declared victory on jobs and moved on to focusing solely on inflation. But the eternal caveats of the jobs report are twofold: NFP day can always serve as the jumping-off point for bolder trading strategies and there’s always a number that’s crazy enough to cause a reaction. Today’s numbers were more than crazy, sort of…
If we only examine the headline payroll count, 467k vs 150k f’cast is substantial–especially in light of Wednesday’s suggestion of the possibility of a negative print from ADP. Looking at recent revisions only adds to the sense of surprise (last month revised up to 510k from 199k for instance).
But there’s a huge caveat that pertains to all of the above. January’s jobs report is the one that contains the annual benchmark revisions (more here). This happens every year. During non-covid times, it helps the Labor Department keep pace with the slowly changing landscape of the labor market and any shift in seasonal patterns. During covid times–especially coming out of the winter of 20/21 (revisions are based only on the month of March)–it makes for some fairly intense changes to recent numbers.
It’s not important to understand exactly how the benchmark revision process works. What’s important is that, in this case, most of the big recent upgrades came at the expense of downgrades earlier in 2021. The following chart shows the monthly payroll counts before and after today’s revision was applied.
What you should be seeing here is that the updated adjustment factors (red line) results in a more even-keeled labor market in the past year. If the revisions were never applied, today’s number would be closer to 200k as opposed to 467k.
Any way you slice it though, the jobs report was strong. The unemployment rate may have ticked up, but the labor force participation rate ticked up more. Wages grew at a 0.7% pace versus 0.5% previously. The 55+ crowd is returning to the workforce much faster than all the doom and gloom of the “early retirement” narrative would suggest. Incidentally, they never dropped out of the workforce nearly as quickly as the 25-54 crowd either. Chalk this up primarily to the service industry.
Regardless of the yeah-buts, the bond market reacted–and swiftly so! Again, a move of this magnitude was not expected, and is likely being exacerbated by algo/technical stop loss levels prompting snowball selling. In other words, the initial selling only needed to be slightly more brisk than expected to take out stop-loss levels that were previously considered unlikely to break. Once they broke, the ensuing selling took out higher stop loss levels that were even less likely to break. Meanwhile, if you’re a buyer, you’re probably sitting on your hands until you see what the market is doing with all this drama next week. The result is a very obvious breakout of a very symmetrical consolidation pattern.
Despite the brisk nature of the sell-off, and despite the clear correlation to today’s events, trading levels in Fed Funds Futures are still moving up in a linear fashion.
One question that’s come up in MBS Live today is whether the market reaction has something to do with the inflation implications of the jobs data. That’s a good question, and it’s technically possible, but I don’t think it accounts for much of the selling. Hot inflation is a known known. This jobs report is more about the revision process than stunning new information. Most importantly, most of the recent rate spike is due to factors other than new increases in inflation expectations. The following chart shows 10yr yields in blue (which include inflation and non-inflationary components, always) and then the non-inflation components in green. From there, simply take note of the much sharper spike in the green line in the past month. That’s speaks to extent to which rates are rising with zero regard for inflation.