Curtailment as a mortgage performance indicator q Che-Chun Lin a,1 , Tyler T. Yang b, * a National Tsing Hua University, 101, Sec. 2, Kuang-Fu Road, Hsin-Chu 30013, Taiwan b Integrated Financial Engineering Inc., 51 Monroe Street, Plaza E-6, Rockville, MD 20850, USA Received 7 July 2005 Available online 12 September 2005 Abstract This paper studies the impact of mortgage curtailment behaviors on the subsequent default and prepayment performance. Although curtailment is not a popular event in the western coun- tries, it is the dominant form of prepayment in Asia and other high saving rate regions. Using a sample of loan-level mortgage performance records from Taiwan, the results of the multino- mial logit regressions indicate that curtailment is one of the most significant factors in predict- ing future default and prepayment probabilities of a seasoned mortgage pool. Mortgages with past curtailment are estimated to be 85% less likely to default and 23% more likely to prepay during the remaining life than a mortgage without any curtailment. Hence, ignorance of past curtailment records could lead to biased projection of default and prepayment and, hence, the pricing and hedging of a seasoned mortgage-backed security. By collecting and incorporat- ing curtailment information, investors could more accurately estimate the fair market value, disclose risk-based capital, and perform effective hedging of a particular mortgage portfolio. Ó 2005 Elsevier Inc. All rights reserved. JEL classification: G1; G21 Keywords: Mortgage; Prepayment; Curtailment; Default; Securitization; Multinomial logit; Competing risk; Portfolio management; Seasoned MBS; Loan level data; Taiwan 1051-1377/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jhe.2005.07.005 q This paper is partially sponsored by the Korean Development Institution and Taiwan National Science Council. * Corresponding author. Fax: +1 301 309 6562. E-mail addresses: chclin@mx.nthu.edu.tw (C.-C. Lin), tyler.yang@ifegroup.com (T.T. Yang). 1 Fax: 886 3571 5403. Journal of Housing Economics 14 (2005) 294–314 www.elsevier.com/locate/jhe JOURNAL OF HOUSING ECONOMICS