RENTAL PRICE AND MULTIFAMILY MORTGAGE PERFORMANCE
Larry Boyer
Freddie Mac
8200 Jones Branch Drive
McLean, Virginia 22102
(703) 714-2899
E-mail:
Corresponding Author:
Jan Ondrich
Center for Policy Research
426 Eggers Hall
Syracuse University
Syracuse, NY 13244
(315) 443-9052
E-mail:
RENTAL PRICE AND MULTIFAMILY MORTGAGE PERFORMANCE
Larry Boyer, and Jan Ondrich
This study will provide an econometric analysis of multifamily mortgage default and prepayment performance using data from 1980 to 1995, and will simulate this performance under various rental path scenarios. The data are for the market rate 221(d)(4) New Constructions or Substantial Rehabilitation of multifamily rental properties (OMI) program, one of the largest FHA multifamily mortgage insurance programs.
The purpose of the 221 (d) (4) program is to assist private industry in the construction or rehabilitation of rental and cooperative housing for low- to moderate-income and displaced families. The FHA insures loans originated by private, HUD-approved lenders for multifamily rental and cooperative housing. The mortgages are issued at market rates, although other subsidies may be associated with the properties or the tenants. For example, some properties associated with 221 (d) (4) program loans receive property-based Section 8 assistance.
Evaluating the price of a mortgage is complicated by the fact that the borrower has the option to default or prepay at any point in time. Furthermore, once a borrower decides to default on a mortgage, she loses the option to prepay, and vice versa. The most common approach in the literature is to rule out the possibility of defaulting when valuing the right to prepay and to rule out the possibility of prepaying when considering default. However, it can be shown that contracts with only one of the options imply different borrower behavior. This paper presents a model of the competing risks of mortgage termination by prepayment or default that estimates the two risks simultaneously.
Another important issue for the estimation stems from the differences in risk preferences and idiosyncrasies across borrowers. We allow for such individual heterogeneity by incorporating individual-specific random effects into the estimation. The model accounts for unobserved heterogeneity among borrowers and estimates the parameters of the heterogeneity distribution simultaneously with the parameters of the conditional prepayment and default probabilities.
The key covariate is the local (MSA-Level) rental price of apartments. Rent is an important determinant of debt-coverage ratio (DCR) volatility; DCR is widely used to underwrite and rate multifamily mortgages. Furthermore, rental price volatility plays an important role in the regulation of the government-sponsored enterprises. Follain, Kogut, and Marschoun (2000) have analyzed how rental prices evolve over time and their analysis will be an important component of our simulation. Once the model is estimated, mortgage performance will be simulated using probability-weighted rental price scenarios. The results will be compared to the average scenario, which is what Federal law mandates that FHA use in its decision-making.
References
Follain, James, David Kogut, and Michael Marschoun. 2000. “Analysis of the Time-Series Behavior of Rents of Mulitfamily Housing Units.” Paper presented at the 2000 ASSA Meetings in Boston.
Ondrich, Jan and Wenyi Huang. 2002. “Stay, Pay, or Walk Away: A hazard Rate Analysis of Federal Housing Administration-Insured Multifamily Mortgage Terminations.” Journal of Housing Research. Fannie Mae Foundation.13 (1):85-117.