Today's much-anticipated jobs report ended up coming out much stronger than expected.  A stronger result was all but guaranteed to cause carnage (relative) in the mortgage market and that's definitely what we're seeing.  A caveat is that rates are still much lower than they were several months ago, but the average lender is now back in line with mid August levels.  Additionally, this is one of the largest single day jumps we've seen with the average 30yr fixed rate moving from 6.26 to  6.53. 

A move of more than 0.25% in a single day is tremendously uncommon, but it can happen due to the underlying structure of the mortgage bond market.  For those who would like to nerd out on those details, here you go:

Whether a mortgage lender is lending their own stockpiles of cash or temporary cash obtained from a credit line, the chunk of cash wired to escrow at closing carries a cost.  For a majority of mortgage lenders, the day to day changes in those costs are determined by the trading of mortgage-backed securities (MBS).

MBS are similar to bonds like Treasuries in that investors pay a lump sum of cash and earn interest over time.  They're different in several key ways.  The most important difference is that the “borrower” of US Treasuries (i.e. the US Government) cannot return principal to the investor and end the deal.  It must continue to pay for as long as it agreed.

Mortgage borrowers, on the other hand, can sell/refi/etc and end the mortgage that underlies the mortgage-backed security.  This introduces an element of uncertainty for investors that will be important in a moment.

Another key difference for MBS is that they are offered in increments of half a percent and only a certain range of mortgage rates can fall into each 0.5% bucket.  To make matters more confusing, there’s overlap between the buckets.  For instance a rates of 6.75 to 7.125 could go 2 different buckets.

Bucket choice matters.  It corresponds to the prevailing rates in that bucket and certain ranges of rates exhibit different behaviors depending on what’s happening with rate trends. The simplest example is that a higher rate bucket is more likely to refinance when rates are trending lower.

Is it good or bad for the investor when it owns mortgages that are refinancing?  There’s no single, correct answer, because it depends on the price the investor paid at the time of purchase, but it’s almost always the case that the investor does NOT want mortgages to refinance when rates had been high, but might be starting to fall.

Now let’s bring all this together with the example above of 6.75 to 7.125 rates having a choice between 2 MBS buckets.  One of those buckets will contain lower average rates.  That bucket will be LESS susceptible to the risk that borrowers refinance when and if rates begin to fall.  Because of that, investors are willing to pay more for that lower bucket and conversely, they may shun the higher bucket by comparison.  At this point, it’s important to note that mortgage lenders determine their pricing based on what investors will pay for loans.

The net effect is that the highest allowable rates in the lower of the two buckets can actually be MORE PROFITABLE for a mortgage lender than the lowest allowable rate in the higher bucket.

Here’s a real world example from an actual lender this week: a rate of 6.625% generated about 10% more profit for the lender, thus allowing them to decrease upfront costs by about 0.2% of the loan amount.  On a $400k loan, that’s an $800 difference.  In other words, you’d pay higher upfront costs if you opted for the 6.75% loan.

The reason we’re diving so deep behind the curtain is to explain why rates can suddenly lurch sharply higher or lower by more than it seems like they should.   The structure explained above means there are certain rates that are disproportionately more economically efficient.  When rates fall enough to get close to the next lowest 0.5% MBS bucket, the highest rate in that bucket makes more and more sense to pursue, even if it means paying a bit more upfront.  

Depending on the methodology underlying any given mortgage rate index, this can make for quicker movement across certain ranges of rates, and large drops on days where the next lowest bucket suddenly becomes a viable option (or large increases on days where that lower bucket is no longer as viable).