A Semiparametric Hazard Estimation of Multifamily Mortgage Prepayments

Posted: 17 Apr 1998

See all articles by James R. Follain

James R. Follain

Federal Home Loan Mortgage Corporation (FHLMC); National Bureau of Economic Research (NBER)

Jan Ondrich

Syracuse University - Center for Policy Research

Gyan P. Sinha

Bear, Stearns & Co., Inc.

Date Written: Undated

Abstract

Estimates of a prepayment function for multifamily mortgages are reported in this paper. These are among the first attempts to estimate such a function; most previous work along these lines focuses on single family mortgages. A further distinguishing aspect of the paper is its attempt to incorporate the impact of unobservable factors on the mortgage refinancing decision. A variant of the maximum likelihood procedure first developed by Meyer (1987) is employed. The results indicate an overall positive duration dependence for the conditional prepayment rate. The estimated response of prepayments to a change in the market rate of interest is significant with the expected sign; it is also larger once the effect of unobserved heterogeneity is taken into account. Nonetheless, the magnitude of the response is substantially less than that predicted by the ruthless option pricing model.

JEL Classification: C14, G21

Suggested Citation

Follain, James R. and Ondrich, Jan and Sinha, Gyan P., A Semiparametric Hazard Estimation of Multifamily Mortgage Prepayments (Undated). Available at SSRN: https://ssrn.com/abstract=9021

James R. Follain (Contact Author)

Federal Home Loan Mortgage Corporation (FHLMC) ( email )

8200 Jones Branch Road
McLean, VA 22101
United States

National Bureau of Economic Research (NBER)

1050 Massachusetts Avenue
Cambridge, MA 02138
United States

Jan Ondrich

Syracuse University - Center for Policy Research ( email )

Maxwell School of Citizenship 426 Eggers Hall
Syracuse, NY 13244
United States

Gyan P. Sinha

Bear, Stearns & Co., Inc.

245 Park Avenue
New York, NY 10167
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
1,167
PlumX Metrics