Mortgage application activity dropped sharply last week as higher rates cut into both refinance and purchase demand. According to MBA’s Weekly Applications Survey for the week ending September 26, total volume fell 12.7% on a seasonally adjusted basis and 13% unadjusted. The Refinance Index decreased 21% from the previous week but remains 16% higher than the same week one year ago. The pullback was broad-based, with double-digit declines across conventional and VA refinancing after rates climbed to three-week highs. Apart from the previous 2 weeks, the index was at the highest levels in more than 3 years. “Mortgage rates increased to their highest level in three weeks as Treasury yields pushed higher on recent, stronger-than-expected economic data. After the burst in refinancing activity over the past month, this reversal in mortgage rates led to a sizeable drop in refinance applications, consistent with our view that refinance opportunities this year will be short-lived,” said Joel Kan, MBA’s Vice President and Deputy Chief Economist. Purchase applications edged lower, with the seasonally adjusted index down 1% and the unadjusted index down 2%, though both measures remain 16% stronger than a year ago. The refinance share of mortgage activity decreased to 55.0% of total applications. The adjustable-rate mortgage (ARM) share fell to 8.4%. The FHA share increased to 16.8%, while the VA share declined to 16.2%. Mortgage Rate Summary:

Both the FHFA and the S&P CoreLogic Case-Shiller indices published updated home-price data this week. The message hasn’t changed: prices are still higher than a year ago, but the pace of growth continues to slow. Case-Shiller is now at its weakest year-over-year level in more than 2 years, while FHFA remains stuck near the lowest growth since 2012. The eternal caveat with home price data is that the “lowest in x years” classification doesn’t mean home prices are falling if the percent change is still positive–something that’s still easily the case in annual terms.  Another way to visualize this is with the simple price indices themselves (NOT the percent change). Note: the following chart’s y axis is normalized such that 100 = 100 for both indices (which simply makes it easier to see correlation). The takeaway from this second chart is quite different. Prices remain near all-time highs and have only ebbed slightly in recent months. None of the moderation in prices over the past few years even belongs in the same conversation as the massive correction seen during the great financial crisis. FHFA House Price Index (seasonally adjusted, MoM)
July: −0.2%; June was unrevised at −0.2%
YoY: +2.8% from July 2024 to July 2025
All nine census divisions remained positive YoY, with gains ranging from +0.6% in the Mountain division to +6.5% in the Middle Atlantic. Case-Shiller National Index (unadjusted)

The National Association of Realtors’ Pending Home Sales Index (PHSI)—which tracks contract signings on existing homes—ticked higher in August, but remains locked in the same flat, depressed range that has defined the past two years. Pending home sales rose 4.0% in August, lifting the index to its highest level since March, and 3.8% above the same month last year. That all sounds pretty good, but the chart tells a more sobering story. The overall trend hasn’t changed: contract activity continues to bounce around within a narrow band, showing only modest sensitivity to month-to-month rate shifts (which could also simply be coincidental). Regional Breakdown (Month-Over-Month)
Northeast: −1.1%
Midwest: +8.7%
South: +3.1%
West: +5.0%
Regional YoY Change
Northeast: +2.6%
Midwest: +6.7%
South: +4.2%
West: +0.2%
Three of the four regions posted solid monthly gains, led by the Midwest and West. On a yearly basis, all four regions were slightly positive, with the Midwest again the strongest performer.

“It’s likely going to be in the thousands,” White House press secretary Karoline Leavitt told reporters on Thursday, saying that the “entire team at the White House” was working to identify possible cuts.

A group of homeowners filed a class action lawsuit against the largest homebuilding company in the United States, seeking all money lost.

The city-owned venue will promote the broker search website by team owner Mat Ishbia’s mortgage company for 10 years in a $115-million deal.

The partnership centers on a master purchase agreement for a portfolio of mortgage servicing rights that the real estate investment trust is buying.

While new jobs data contributed to falling Treasury yields, mortgage rates showed a mixed picture with some trackers moving in opposite directions.

Ultra Quiet Session, But Ultimately Stronger

With the day’s only relevant econ data on hold due to the shutdown, bonds didn’t have any objective inspiration. Add the Yom Kippur holiday as well as the absence of tomorrow’s jobs report, and there suddenly seems to be no compelling reason for the bond market to be open for the rest of the week. This was reflected in volume and AM volatility–both were muted. There was modest movement in the afternoon with trading levels moving into slightly stronger territory. This could be viewed as an incidental  byproduct of the shift in buyers/sellers after the European close, or an intentional front-running of what bond bulls hope is a weaker ISM report tomorrow. It doesn’t really matter either way. Bigger, more durable moves are on hold until the jobs report comes out.

Market Movement Recap

09:23 AM Flat overnight with modest selling early.  MBS down 2 ticks (.06) and 10yr up 1.4bps at 4.113

12:08 PM Sideway to stronger in MBS, now unchanged on the day.  10yr near best levels, up less than half a bp at 4.102

01:57 PM best levels of the day.  MBS up 3 ticks (.09) and 10yr down 1.2bps at 4.087

As a reminder, economic data published by government entities is on hold during the shutdown. Today, that means no Jobless Claims (also no Factory Orders, but we don’t care about that report anyway), and thus a random walk for trading momentum, absent any unexpected headline shock. Friday still has ISM Services data, and thus some old fashioned “cause and effect” potential.