Mortgage rates were crushed by today's jobs report.  Nonfarm payrolls (the main component of the report), came in  significantly higher than expected (353k vs 180k forecast) and were revised up significantly for December (333k versus 216k previously). January's jobs data (the stuff released today) is often plagued by major departures from expectations because it's the one month of the year where the Bureau of Labor Statistics (BLS) implements new benchmarks based on a comprehensive count of jobs conducted in March of the previous year.   To understand this better, consider the changing composition of jobs over time.  BLS adjusts job counts based on how big a percentage a certain industry accounts for. Let's imagine social media were invented overnight and thousands of people quit regular jobs to become social media influencers.  Because there wouldn't be any track record of "social media influencer" in the BLS benchmarks in the first year, it would look like a lot of people lost jobs and didn't find new ones.  Then when BLS does the full count in March, they'd find that influencers were prevalent and the numbers would be revised.  If you'd like to see the actual changes in each industry category, BLS publishes the data here: https://www.bls.gov/web/empsit/cesprelbmk.htm  Benchmark revisions, alone, don't explain the wild results today, but they help explain why results have been a bit more wild than their normal range amid the changing landscape of the post-pandemic economy.  The more rapidly the composition of the economy changes, the more we'll see volatility like this in the numbers.