Recent past gives mixed signals for expanded mortgage lending
An improving economy and tight rental housing are leading more moderate-income renters to consider homeownership. In response, some community lenders are exploring whether to expand loan programs targeted to this market. With the housing collapse and high foreclosure rates still a fresh memory, new analysis by the Minneapolis Fed shows that the timing of home purchases can play a large role in foreclosure rates, especially when combined with credit scores.
Using data provided by Lender Processing Services (LPS) Applied Analytics, mortgages in Minnesota since 2000 were analyzed to better understand the likelihood of foreclosure. Borrowers with lower credit risk scores at loan origination had higher rates of foreclosure over this period. For example, among mortgages originated in 2003, 1.5 percent of borrowers with a 720+ credit score entered foreclosure within seven years, compared with 4.5 percent of borrowers with scores of 680 to 719 and 9.4 percent of borrowers with scores of 620 to 679 (see Chart 1).
Foreclosure rates were also strongly linked to timing. Mortgages originated in 2000 performed better than 2003 mortgages and much better 2005 mortgages with similar credit scores. For mortgages with low credit scores (620 to 679), the foreclosure rate after seven years was just 3.2 percent for mortgages originated in 2000, but reached 26.6 percent for mortgages originated in 2005.
These differences may partly reflect the boom and bust in Minnesota housing prices between 2000 and 2010. As home prices fell after 2007, homes with year 2000 mortgages were more likely to remain “above water”—that is, to have a value greater than the balance due on the mortgage—than homes with mortgages originated in 2003 or 2005. Borrowers with above-water mortgages have both stronger incentives and better options for avoiding foreclosure, such as simply selling. And many more mortgages originated in 2000 may have been refinanced early on, given that mortgage rates fell between 2000 and 2003.
High debt burdens were also linked to foreclosure (see Chart 2). For example, controlling for credit score, foreclosure rates for 2003 mortgages were about 50 percent higher for high-debt borrowers than lower-debt borrowers. For 2005 mortgages, foreclosures were considerably higher for both high- and low-debt borrowers compared with 2003 mortgages. The specific effect of debt levels was present but appears less pronounced, as foreclosures among high-debt mortgages were about 35 percent, while low-debt foreclosures reached almost 30 percent. (A comparison with 2000 mortgages was not possible due to insufficient data on debt-to-income ratios.)
If housing prices rise or at least remain stable over the coming years, the more encouraging performance of Minnesota mortgages originated in 2000 may provide useful guidelines to organizations contemplating new home lending programs. However, the weaker performance of loans from 2003 and 2005 shows the risks that could arise if housing prices or employment fall significantly again.
For related information, see the Minneapolis Federal Reserve Bank’s web page Housing Market and Mortgage Conditions in the Ninth District.