Fitch Ratings is finalizing its new model framework for determining loan losses on pools of U.S. prime residential mortgage loans, which includes several important updates. Chief among them is the application of a proprietary sustainable home price model (SHP Model) that measures a property's current price against its sustainable value. The SHP model allows Fitch to take a forward looking, countercyclical view on the potential for negative equity, which has shown to strongly influence borrower default behavior.

Fitch's SHP model uses macroeconomic and mortgage market factors to determine fundamental or sustainable home prices at the state level. A core principle is that if actual home prices are above levels supported by underlying economic and housing fundamentals (indicating the risk of an asset bubble), they will revert to sustainable levels over time.

The model's development originated from the broader goal of identifying the risk of further declines in various regional housing markets. Data from the recent downturn has shown that a borrower's propensity to default is highly correlated with total equity remaining in the property. This relationship is particularly evident in the high levels of defaults seen for properties with updated loan-to-value ratios of 150% or more. Given the strength of this relationship, the ability to predict housing market overvaluation at the time of loan origination would provide significant uplift to any mortgage loss model. The SHP model's home price forecasts are a key input in Fitch's new loss model in estimating both default and loss severity expectations.

While Fitch's criteria recognizes the importance of initial down payment and borrower equity as a driver of borrower default behavior, the new approach measures the amount of real equity versus that created merely as a result of a price bubble.

In building its SHP model, Fitch found it essential to differentiate between a true housing boom and an overheated market. Price growth alone does not signify a bubble, as evidenced by some states and large cities that have seen sustained price growth over time as populations swelled and real incomes rose. In calculating market dynamics, the SHP model includes metrics on unemployment, income, population, housing units, mortgage rates, and real home prices, measured quarterly over an observed 35-year history. To capture the long-term dynamics of each market, expected home prices in any period are calculated as a function of four-year averages of each of the underlying fundamental drivers, including a 12-quarter historic look-back, and a four-quarter forward projection. By regressing movements in these factors on a restricted time series against the observed movements in home prices, a 'sustainable price' is generated. This price level serves as a baseline for comparison to observed prices in any given period.

Graph 1 demonstrates the sustainable price concept for California from 1979 to present.

What Happened in California?

California has long been a focus in the securitization industry due to its size (representing nearly 40% of the prime non-agency space) and its historic price volatility. The run-up in prices in the late 1980s and the 29% real decline that followed represented one of the biggest boom and bust cycles in American history. At the time, it was difficult to distinguish between supportable home price levels and an overheated market, as unemployment in 1989 had hit more than a decade's low of 5.1% and other fundamentals appeared strong. However, home prices had risen by more than 50% in the five-year period leading up to 1989, significantly outpacing the 20% growth predicted by the SHP model.

When defense industry spending cuts ravaged local economies and drove state unemployment to 10%, prices lost their buoy and sank back to sustainable levels. It would be 10 years before real prices recovered to the same level. Using the econometric and price information available at the time, the SHP model would have predicted an average home price decline of 21%. However, negative price momentum and a struggling economy would send prices temporarily below sustainable levels during the downturn.

By 2006, California's economy had benefited from nearly a decade of economic growth, experiencing more than a 20% increase in real incomes and population growth that had outpaced housing supply. Keeping pace with demand, housing construction boomed, accompanied by an expansion in mortgage credit availability. Led by easy money and market enthusiasm, prices doubled in barely five years. At the peak, the SHP model would have predicted more than a 50% decline in real values in California based on underlying fundamentals. Prices eventually reverted to sustainable levels in 2009. It should be noted, however, that the bounce in prices that followed the passage of the homebuyer tax credit and other government programs has left California with a projected correction of approximately 12%.

The degree by which home prices exceed their fundamental or sustainable value has proven to be an important predictor of mortgage defaults, and is a key variable in Fitch's new prime loan loss model. As calculated at origination, loans predicted to face declines of 40% and 50% or more have default rates of more than six and 17 times higher, respectively, than that of loans calculated to be at sustainable values. Loans expected to see a significant increase in values have defaulted at roughly 1/4 the rate of sustainable loans (see Graph 2).

The SHP concept is currently implemented at the state level. However, a model expansion which will provide regional results for more than 100 metropolitan statistical areas is in the late stages of development, and is expected to be released soon.

The preceding was derived from Fitch Ratings' upcoming new rating criteria for new U.S. RMBS, which Fitch expects to release this month.

Subscribe Now

Access to a full range of industry content, analysis and expert commentary.

30-Day Free Trial

No credit card required. Access coverage of the securitization marketplace, including breaking news updated throughout the day.