While we may not have a great explanation or even the capacity to believe it, nonfarm payrolls came in at 517,000 this morning versus a median forecast of 185,000. The bond market has reacted about like you'd expect: with a large, immediate sell-off. While that number does indeed seem very high (and it is--especially relative to the forecast), there are two caveats.
1. It doesn't look ridiculously high in the context of the past 2 years.
2. Today's data was also affected by the annual re-working of seasonal adjustment factors. This isn't some conspiracy or deception. The BLS (the entity that collects and publishes this data) has an established procedure for updating seasonal adjustments once per year. The changing landscape of the post-pandemic labor market has thrown a few curve balls to revision processes such as this. And while we can't comment on whether this particular revision caused overly optimistic distortion of today's number, it's fair to say it may have helped the number be a bit higher than it otherwise would have been.
But before you cry foul and blame BLS math for overstating the level of employment, consider that the NON-ADJUSTED numbers suggest an average monthly job gain of more than 400k over the past year (same as the adjusted numbers).
Adding insult to injury is a significantly stronger number reported in ISM's non-manufacturing index 90 minutes after the jobs report. The index came in at 55.2 vs a median forecast of 50.4. That is definitely a big "beat," and an even bigger improvement from the 49.2 reported last month, but the caveat is much more obvious in this case. Last month really sucked and economists incorrectly guessed that today's number would follow suit. Here too, changes in seasonal patterns are likely contributing to forecasting errors.
Given all of the above, the 11bp sell off in 10yr yields and the half point loss in MBS seem reasonable. Both remain in stronger territory vs last week.