MBA: “Mortgage Delinquencies Increase in the Fourth Quarter of 2022”
Calculated Risk: MBA: “Mortgage Delinquencies Increase in the Fourth Quarter of 2022” calculatedriskblog.com, Bill McBride
by Calculated Risk on 2/16/2023 01:42:00 PM
From the MBA: Mortgage Delinquencies Increase in the Fourth Quarter of 2022
The delinquency rate for mortgage loans on one-to-four-unit residential properties increased to a seasonally adjusted rate of 3.96 percent of all loans outstanding at the end of the fourth quarter of 2022, according to the Mortgage Bankers Association’s (MBA) National Delinquency Survey.
The delinquency rate was up 51 basis points from the third quarter of 2022 but still down 69 basis points from one year ago. The percentage of loans on which foreclosure actions were started in the fourth quarter fell by 1 basis point to 0.14 percent.
“As expected, the overall national mortgage delinquency rate increased in the fourth quarter of 2022 from its previous quarterly survey low,” said Marina Walsh, CMB, MBA’s Vice President of Industry Analysis. “The weaker economy and ongoing inflationary pressures contributed to the uptick in delinquencies. The delinquency rate – while still low – increased from the previous quarter across all loan types and across all stages of delinquency.”
According to Walsh, for the past 15 years, mortgage delinquencies have tracked very closely with employment conditions. Despite recent indicators of resiliency in the job market, including the unemployment rate declining to 3.4 percent in January, MBA still forecasts for slower hiring and rising unemployment, with the rate rising to 5.2 percent by the end of the year. This will likely mean further increases in mortgage delinquencies.
Added Walsh, “Notwithstanding the fourth-quarter increase in mortgage delinquencies, the foreclosure starts rate of 0.14 percent was well below the historical quarterly average of 0.40 percent. Many distressed homeowners have loss mitigation options available to them and have accumulated home equity, which can ease financial hardship and avert foreclosure actions.”
This graph shows the percent of loans delinquent by days past due. Overall delinquencies increased in Q4 from a record low in Q3.
Click on image to enlarge.
Compared to last quarter, the seasonally adjusted mortgage delinquency rate increased for all loans outstanding. By stage, the 30-day delinquency rate increased 26 basis points to 1.92 percent, the 60-day delinquency rate increased 13 basis points to 0.66 percent, and the 90-day delinquency bucket increased 11 basis points to 1.38 percent.
The delinquency rate includes loans that are at least one payment past due but does not include loans in the process of foreclosure. The percentage of loans in the foreclosure process at the end of the fourth quarter was 0.57 percent, up 1 basis point from the third quarter of 2022 and 15 basis points higher than one year ago.
The sharp increase in 2020 in the 90-day bucket was due to loans in forbearance (included as delinquent, but not reported to the credit bureaus).
The percent of loans in the foreclosure process increased year-over-year in Q4 with the end of the foreclosure moratoriums but are still historically low.
Anyone know what seasonal adjustment would apply here? Seems an odd kind of metric to adjust to me. I could accept that there could be 12 month cyclic elements somehow, but I would not adjust the data unless that was some kind of measuring or reporting variance cycle. L
Seasonally Adjusted Annual Rate (SAAR): Calculations and Examples (investopedia.com)
Later, I am going back to sleep. Woke up too early.
Thanks. In the work I did we did not adjust much data and then report that as some version of the data. But we did incorporate seasonality and other cyclical effects when using the data in prediction models. The model carried the adjustments, when appropriate. So here maybe delinquencies are high, but the model of whether lenders are going to feel a lot of pressure might understand that 1Q (calendar) always is higher pre-tax refunds, so minimal pressure is the output of the prediction. Same data in June maybe looks like a big problem.