EXCLUSIVE: How Much Do Price Trends Matter in Residential Mortgage Default Forecasting?

Carrie Rossenfeld Residential & Mixed Use

The following is excerpted and adapted from an article that was presented exclusively to SoCal Real Estate. To read the article in its entirety, complete with charts and graphs, click here.

By Norm Miller, PhD, Hahn Chair of Real Estate Finance at the University of San Diego School of Business and affiliated with the Burnham-Moores Center for Real Estate, and Michael Sklarz, PhD, chairman and CEO of Collateral Analytics

Headshots courtesy a representative of the Burnham-Moores Center for Real Estate at the University of San Diego School of Business

Default rates on mortgage loans are driven by a number of factors, including but not limited to: income to mortgage payment ratio, credit behavior and attitudes toward paying obligations as reflected in credit scores, loss of a job or unexpected medical expense, available liquid assets that may be tapped for mortgage payments, and perceived and/or actual equity in the home as reflected in current values.

While all of the above are important, the last factor seems to dominate all the others. Home equity is a function of current market value and the loan-to-value (LTV) ratio of the original mortgage, along with any other debt added during ownership, such as second mortgages and home equity lines of credit (HELOC). Here, we focus on this last factor, home equity.

When home values decline to the point where borrowers have no equity or negative equity, and they see no immediate reversal in local price trends, they often decide to default. This is sometimes called “rational default” or “strategic default,” and while it may affect credit ratings for some time, U.S. borrowers know they will seldom be chased for mortgage-balance deficiencies. This encourages the decision to default when the mortgage balance appears to exceed the value of the home. This creates a situation where mortgage lenders are on the short end of a “heads I win, tails you lose” situation. If prices go up, borrowers’ benefit, but if prices go down too far, the mortgage lender suffers.

The tendency towards default may be reinforced when borrowers know it will take several months to evict them from the home, and meanwhile they can live rent free, extracting more capital from mortgage investors. (State laws and the borrower’s rights after default vary considerably by state, which in turn affects borrower behavior. For example, see “Recourse and Residential Mortgage Default: Evidence from US States” by A. Ghent and M. Kudlyak in The Review of Financial Studies, Volume 24, Issue 9, 1 September 2011, Pages 3139-3186.) Some borrowers use the excuse that lenders have misled them or charged too high of a rate, allowing rationalization for not making their mortgage payments.

Michael Sklarz

The typical pattern of mortgage default in response to negative price trends and high LTV mortgages can be shown using Chula Vista, California — a market with higher-than-average LTV mortgage loans — as an example. While set in San Diego County, this example is not unique but typical of all housing markets where prices have changed significantly.

In Chula Vista, loans originated in the years 2000, 2001, 2002, and 2003 were the beneficiaries of strong price increases of 12 percent, 14 percent, and 25 percent in average home prices, so when prices reversed in 2006, there was still sufficient cushion to preserve real home equity. Independent of factors 1 through 4 above that influence default, factor 5 (home equity) was sufficient that few home owners defaulted. Note that during this period of time the average purchase mortgage LTV was climbing every year.

Mortgage loans originated in 2004 saw increasing default rates peaking out in 2007 as home prices started to decline in 2006. Loans originated in 2005, 2006, and 2007 at the peak of home prices in the cycle, and cumulatively half or more defaulted in the subsequent years. The high LTV loans declined as subprime loans disappeared; however, the FHA and loan modifications allowed many homeowners to continue with high LTVs through 2012.
In looking at the annual average percentage change in price by year in Chula Vista, we see that prices increased rapidly as mortgage credit became easier in the 2003-through-2005 period, then fell as easy mortgage credit reversed. Those who were late in following the herd of homebuying enthusiasts often paid prices that were unsustainable with current household income. Modest home-price appreciation, closer to inflation and wage growth, is sustainable, but we see the market overshoot and undershoot these more sustainable levels.

To expand our analysis, we selected a range of markets to study from the same county, where we saw significantly different levels of original mortgage leverage. We looked at six markets — Spring Valley, El Cajon, Carmel Valley, Coronado, La Jolla, and Rancho Santa Fe — and noted the average mortgage LTVs for the combined years of 2005 and 2006 as well as the percentage of transactions during 2005 and 2006 where no mortgage was used at all. This is significant as an indicator of contagion from the local market. We noted that areas with all-equity homes tend to be more stable with fewer foreclosures. The lower the rate of foreclosures, the fewer “distressed” sales that might be used by appraisers or buyers as indicators of value for any local property.

Next, we combined all of these markets, comparing default rates for loans originated during the peak of easy credit, 2005 and 2006. What we observed is that those markets with higher LTV mortgages and similar price trends as Chula Vista defaulted at much higher rates. It is very clear from these patterns that home-price trends dominate the decision of home borrowers to default. Price declines in 2008 were especially influential in that most borrowers expected these negative trends to continue. (In many surveys by Case and Schiller of homeowners, asking what next year’s home appreciation rate will be, the typical answer was whatever happened in the most recent year.)

Last, we looked at the many zip codes in San Diego and default rates for all mortgages originated from
2000 through 2007. We noted that the areas that had the lowest default rates were areas where mortgages are less relied upon and LTVs tend to be lower. The areas where mortgage-default rates are slightly higher are neighborhoods where home buyers have less access to wealth for down payments and use higher LTV mortgages, thus are more apt to go into financial stress should home prices decline.

In this case, the math is quite simple. If the average LTV was 92 percent in a local market and home prices slip just 8 percent, that home equity is wiped out. Add in transactions and legal costs, and we can see that mortgage investors are very vulnerable. In previous articles, we have discussed how mortgage investors and, in turn, lenders, could deal with these price-induced risks. They need to account for not only current values, but also price uncertainty and future price trends. This is certainly possible by adjusting LTVs to a statistically driven range of confidence.

We are not suggesting any sort of neighborhood-level, absolute screening for mortgage underwriting criteria. We are suggesting that price uncertainty and actual and expected equity are extremely important drivers of mortgage default. Thus, good valuation models with information about price uncertainty and even price forecasts are factors that should be considered by mortgage investors.