There's been widespread speculation as to the effect that falling rates will have on refinance demand in a world where 2/3rds of mortgages have rates under 4%. This week's data suggests it's worth discussing.

While the blue bars on the right side of the chart above may not be as big, they still represent 100s of millions of dollars of loan balances.  In fact, there are nearly 10 billion dollars outstanding at rates of 7% and higher.  As such, when last week's rates dropped briefly into the low 6% range, there was a mini glut of refi demand.  The following chart shows the latest refinance application index from the Mortgage Bankers Association (MBA).  

That may not look too "mini" at first glance, but context is important.

In other words, we're barely back to what had been historically low levels of refi activity seen in late 2018.  The counterpoint is that rates were only in the low 6's for an hour or two last Monday morning.  Bottom line: this is still proof of concept for more widespread refi demand in a scenario where rates move back into the 5s and hold those levels for more than a few hours.

So how likely is a move back into the 5s?  Based on the recent trajectory of economic data 2 weeks ago, it's only a matter of time.  Since then, the data has been more of a two way street.  The first few days of this week saw rates move lower in response to lower inflation.  Thursday brought a sharp bounce in response to higher retail sales numbers and lower jobless claims.

The weekly Jobless Claims report is becoming rapidly more important due to the confluence of the most recent jobs report (a different, significantly larger and more important report compared to the jobless claims data) and the Federal Reserve's increased level of focus on the health of the labor market.

In not so many words, the Fed just talked about watching employment more closely in the announcement 2 weeks ago.  2 days later, the big jobs report came out much weaker than expected.  Markets figured the Fed would have been even more rate friendly if they'd known about the jobs data.

From that point on, every little shred of employment data has been more important than normal.  Jobless Claims, in particular, are very useful because they're reported on a weekly basis, thus allowing the market to potentially get ahead of any shift in the trend.  

Rather than confirm the downbeat message from the big jobs report, Jobless Claims have instead been signaling business as usual, cutting a substantially similar path to the past 5 non-lockdown-distorted years.  

As seen in 10yr Treasury yields, rates have responded accordingly.

Thankfully for fans of low rates, these bumps have been contained by a broader consolidation near the lowest rate levels of the year.

Between now and the September Fed meeting, there are FIVE more jobless claims releases, another big jobs report, and another installment of the Consumer Price Index (CPI) in addition to plenty of strong supporting actors of the economic data world.  If there's a cohesive message, rates should move accordingly (i.e. weaker data = lower rates, stronger data = higher rates).

As always, mortgage rates and Treasury yields would move well before the Fed announces a rate cut (seen as 100% likely for September).  The market is constantly trading its expectations for the Fed Funds Rate, and those expectations tend to look a lot like the day to day movement in consumer rates.

Here's a longer term chart of a different Fed Funds Rate futures contract (June 2025) overlaid with a 5yr Treasury yield.  5yr Treasuries are actually more correlated with mortgage rate movement these days than 10yr Treasuries. Either way, the point is to show that the rate market will have already made its move by the time the Fed meeting rolls around.