With housing policy on the backburner during this election year, the Federal Housing Finance Agency has partly filled the vacuum by proposing several reforms to improve the mortgage market. Unfortunately, the FHFA's latest proposal, to start charging five states higher guarantee fees to cover default costs well above the national level, sets a bad precedent.
Mortgage losses are comprised of two parts; the percentage of defaults on a pool of mortgages (default incidence) over a period of time times the dollars of losses suffered once a default occurs (loss severity). Historically, Fannie Mae and Freddie Mac priced the cost of insuring a mortgage sold to them by a lender by estimating such losses expected to occur over the life of the portfolio plus assigning additional costs associated with capital required to cover losses for particularly adverse outcomes. Over time, with extensive loan level performance data, the government-sponsored enterprises were able to price g-fees based on a number of borrower, product and other attributes determined to be statistically significant determinants of mortgage loss. Macroeconomic effects such as house price, employment and economic growth trends at the state or local level would often be included in such analyses.
In estimating mortgage loss, it is well-established that state foreclosure timelines affect the severity of loss in two important ways. First, a great deal of variation exists between states in the number of days it takes to foreclose and take title back on a property, due in part to whether the process requires a court to intervene in the proceedings or not. For example, the range in the FHFA's state estimates of total time to obtain marketable title is 270 to 820 days.
The second factor driving losses is the actual cost associated with carrying a nonperforming loan. These costs include such items as property taxes, maintenance expenses, legal and marketing costs and GSE financing costs on delinquent loans.
The FHFA's proposal calls for a state-level assessment of these costs. Borrowers residing in states deemed to be significant "outliers" would be charged additional upfront g-fees. The affected states are Florida, Connecticut, New Jersey, New York, and Illinois.
In theory, the FHFA's application of risk-based pricing for state foreclosure cost differences is founded on solid economic principles of charging customers based on the inherent risks they present to the insurer. FHFA acknowledges that under the g-fee model used by the GSEs, borrowers in low-cost states subsidize those in higher-cost states since the agencies do not differentiate their pricing by state.
But introducing state-level g-fees for a subset of high foreclosure cost states would set up the FHFA for a major fight with the states without really serving the purpose of such pricing.
The FHFA directly asserts in its proposal that targeted high cost states electing to modify their laws, requirements and even property tax regimes, such that their foreclosure costs conform to the national average, could face lower or no state-specific g-fees. In effect, the pricing scheme envisioned by FHFA would attempt to "incent" state and local governments to implement changes to an admittedly byzantine foreclosure process that has stymied housing recovery while embedding huge inefficiencies in managing foreclosure activity.
Thus, the proposed remedy to high state foreclosure costs would introduce a precedent for allowing federal agencies to design regulations intended to shape state laws and practices. The radical implications of this idea, and the potential backlash from the states, overshadow any economic benefit that may accrue from risk-based pricing.