Follow Insights Blog

CoreLogic

CoreLogic Econ

risk summit

LATEST CORELOGIC ECON TWEETS

Purchase-Credit Demand Weakens

What Caused the Drop?

Sam Khater    |    Housing Trends

For the first half of 2013, the MBA mortgage purchase application index increased steadily, rising 10 percent from the prior year. Then it began to weaken in response to the more than 100 basis-point rise in the 30-year fixed rate mortgage (FRM) that occurred between early May and July. By late July growth in the purchase application index decelerated and was up only 5 percent from the previous year. The 5-percentage-point deceleration was small given the sharp and unexpected rise in mortgage rates and the purchase market stabilized in August and September.

However, when the government shutdown occurred between October 1st and October 17th, the index weakened again. Between late September, when talk of a possible shutdown had reached a fever pitch, and mid-November, a full month after the shutdown, the growth rate of the index fell from 5 percent to -1 percent on a year-over-year basis – or a 6-percentage-point deceleration. The net short-term impact of the government shutdown on purchase-credit demand was slightly larger than the more than 100 basis-point increase in rates.

While purchase-credit demand stabilized after the shutdown, beginning in late November, the index declined sharply again and has dropped for six consecutive weeks by an average of 10 percent year over year and as of the second week of January stood at 15 percent below a year ago. This decline in the index was almost the same size as the combined impact of the mortgage rate increase and the government shutdown. Looking more closely at the data reveals that the drop in applications occurred in both conventional and government purchases, with a slightly larger decline in conventional purchase applications.

What caused the sudden decline? First, the 30-year FRM has increased from 4.10 percent in late October to 4.53 percent in the last two months – a 43-basis-point increase. While that is a noticeable increase, the rise was less than half the rise earlier this year that rate increase only led to about a 5-percentage-point decline in the year-over-year growth rate of the index. Moreover, there have been other instances of similar rate increases during the last 20 years that didn’t lead to a drop in applications. The largest two-month increases in the 30-Year FRM over the last 20 years were in the spring of 1994, summer of 2003, and spring of 2004. All three generated very minor declines in purchase application activity on a year-over-year basis. The difference between today’s market and those eras is that in the past borrowers would often switch to more affordable ARM products, many of which today are less available. The December drop in applications reveals that purchase market today is more sensitive to a rise in 30-year FRM rates than in prior years because borrowers are less able to switch to more affordable products. 

Another explanation could be weaker labor markets. Weekly initial unemployment claims have been very volatile lately, but after claims hit a cyclical bottom of 305,000 in late September, they increased to 349,000 as of early January – a 14-percent rise[1]. The weakness in unemployment claims was further confirmed by the very low 74,000 increase in employment. Weaker labor markets typically lead to a drop in consumer sentiment, which impacts borrowers’ confidence in applying for mortgage credit. Another possibility is that holiday and weather effects have made an impact. The Thanksgiving holiday fell later than normal in 2013 and it was the shortest gap between Thanksgiving and Christmas in 11 years. This constricted holiday season could have led to a shorter period for house hunters to submit mortgage applications. It has also been unusually cold during the last few weeks, likely impacting homebuyers’ desire to go out and visit houses.

© 2013 CoreLogic, Inc. All rights reserved.



[1] The initial unemployment claims data represents a 4 week moving average.