Warning: Tomorrow's Mortgage Rate News Will be Wrong
Mortgage rates have had a great week. Although rates are slightly higher today versus yesterday, they're still exceptionally close to the lowest levels since November 2016. That's actually been the case since late last week as more than half of the improvement in underlying bond markets (which dictate rates) was in the books by Friday afternoon. It took lenders a few days to get caught up with the market movement, but by yesterday afternoon and this morning, top tier scenarios were seeing rate quotes of 3.625% in many cases (25%+ down, perfect credit, etc).
Tomorrow will bring a raft of mortgage rate headlines from major news outlets due to the typical Thursday release of the widely-followed Freddie Mac Primary Mortgage Market Survey. Those stories will all be wrong. How can I be so sure? Freddie's methodology means it only collects survey responses from Monday through Wednesday. The results are heavily skewed in favor of Monday and Tuesday's rates. Thursday and Friday are not represented at all, and the same is arguably true for Wednesday afternoon based on past precedent.
Therein lies the problem. From noon on Wednesday last week through noon today, there was nothing but good news for bond markets and mortgage rates. On Monday and Tuesday of last week, rates were flat and holding at their highest levels in nearly 2 weeks. So all of the massive improvement we've had in the past week was yet to be reflected in Freddie's 3.75% figure from last week. Based on actual lender rate sheets, as of this morning, the average lender was pricing out 3.5% 30yr fixed rates at almost the exact same cost as last Wednesday's 3.75% rates. I don't know if Freddie's survey is going to quite reflect a full 0.25% drop, but anywhere from .15-.20 should not be a surprise in the slightest.
That means tomorrow's mortgage headlines will be shouting from the rooftops about "the biggest drop in years" and "30yr fixed rates near or under 3.6%." That's a problem for a few reasons. Unless the bond market has a great night, rates under 3.6% will be a lot harder to find tomorrow morning (they're already pretty exceptional). Beyond that, in purely technical terms, rates stand a very good chance of being fairly similar to last week's latest levels due to bond market weakness this afternoon (which, again, Freddie's methodology doesn't capture).
The bottom line is that the timing of the various big market moves over the past week could not have done a better job of evading Freddie's survey methodology. This will probably be the single biggest difference between Freddie's week-over-week survey observations and actual mortgage rate changes that we've seen in years. Don't expect rates to be any lower than they were on Friday, and expect them to be higher than they were earlier this week, no matter what any news article says.
Today's Most Prevalent Rates
- 30YR FIXED - 3.625% - 3.75%
- FHA/VA - 3.25%
- 15 YEAR FIXED - 3.375%
- 5 YEAR ARMS - 3.375-3.75% depending on the lender
Ongoing Lock/Float Considerations
- Early 2019 saw a rapid reevaluation of big-picture trends in rates and in markets in general
- The Federal Reserve has been a key player, and while they aren't the ones pulling the global economic strings, their response (and even their EXPECTED response) to the economy has helped rates fall more quickly than they otherwise might.
- Based on the Fed's laundry list of concerns, the bond market (which determines rates) will be watching economic data closely, both at home and abroad, as well as trade-related concerns. The stronger the data and trade relations, the more rates could rise, while weaker data and trade wars will lead to new long-term lows.
- Rates discussed refer to the most frequently-quoted, conforming, conventional 30yr fixed rate for top tier borrowers among average to well-priced lenders. The rates generally assume little-to-no origination or discount except as noted when applicable. Rates appearing on this page are "effective rates" that take day-to-day changes in upfront costs into consideration.