Weekly applications to refinance single-family mortgages were out front of weakening activity. Applications for loans to finance a home purchase have strengthened from a year ago.
Loan originators completed nearly 3 percent fewer applications in the week ended Oct. 27 than they did the prior week based on the seasonally adjusted Market Composite Index.
The index, which quantifies the volume of retail residential loan applications, still declined by 3 percent from the week ended Oct. 20 when seasonal factors are not considered.
Producing the index is the Mortgage Bankers Association based on its Weekly Mortgage Applications Survey. More than three-quarters of all applications are reportedly covered by the survey.
A 5 percent week-over-week decline was recorded for refinances, which made up 48.7 percent of all activity. Refinance share was more narrow than 49.5 percent a week earlier and
62.7 percent a year earlier.
A less severe
1 percent seasonally adjusted dip from the preceding seven-day period was recorded for purchase-money loan applications. Without seasonal adjustments, applications for purchase financing fell 2 percent but have accelerated by one-tenth from the week ended Oct. 28, 2016.
MBA said that 10.4 percent of applications were for mortgages insured by the Federal Housing Administration. That was wider than 9.8 percent seven days prior but more narrow than 11.1 percent one year prior.
Another 9.9 percent of applications were for loans guaranteed by the Department of Veterans Affairs, thinning from 10.1 percent. VA share was broader last year at 12.4 percent.
At 6.8 percent, the share of applications that were for adjustable-rate mortgages was more thick than
6.4 percent in the last report and 4.4 percent in the same-seven days during 2016.
With average jumbo interest rates rising to their highest level since July, jumbo rates were just 6 basis points less than conforming rates versus 7 BPS lower in the last survey.
But the spread was wider than this week a year ago, when jumbo rates were just a single basis point less.
