Characteristics of Mortgage Terminations: An Analysis on the Loan-Level Dataset

42 Pages Posted: 23 Nov 2016

See all articles by Hoon Cho

Hoon Cho

Korea Advanced Institute of Science and Technology (KAIST)

Hyeongjun Kim

College of Business, Korea Advanced Institute of Science and Technology (KAIST)

Doojin Ryu

Sungkyunkwan University

Date Written: November 1, 2016

Abstract

To valuate mortgage-backed securities, it is crucial to understand mortgage termination behavior. Analyzing the unique loan-level dataset, this study examines the characteristics of mortgage prepayment and default behavior in the Korean housing and housing finance markets. We find that loans to individuals with unhealthy financial characteristics — for example, a high loan-to-value ratio, a high debt-to-income ratio, a lower credit rate, or a high excess premium — have a higher prepayment risk in the Korean market, which contrasts with the findings on developed markets. The results of subsample tests across regions, loan purposes, and periods indicate that the prepayment risk in a high median price-to-income ratio region is sensitive to mortgage rates, whereas the prepayment risk in a low median price-to-income ratio region is sensitive to housing prices. Mortgages for purposes other than purchase or refinance are less sensitive to financial benefits in prepayment decisions. Prepayment and default behaviors have also changed over time.

Keywords: Default, Hazard model, Loan-level dataset, Mortgage; Prepayment

JEL Classification: G21, D12, R31

Suggested Citation

Cho, Hoon and Kim, Hyeongjun and Ryu, Doojin, Characteristics of Mortgage Terminations: An Analysis on the Loan-Level Dataset (November 1, 2016). KAIST College of Business Working Paper Series No. 2016-016, Available at SSRN: https://ssrn.com/abstract=2874603 or http://dx.doi.org/10.2139/ssrn.2874603

Hoon Cho

Korea Advanced Institute of Science and Technology (KAIST) ( email )

87 Hoegiro
Dongdaemun-Gu
Seoul 130-722
Korea
+82-2-958-3413 (Phone)

Hyeongjun Kim

College of Business, Korea Advanced Institute of Science and Technology (KAIST) ( email )

85 Hoegiro, Dongdaemoon-gu
Seoul 02455
Korea, Republic of (South Korea)

Doojin Ryu (Contact Author)

Sungkyunkwan University ( email )

53 Myeongnyun-dong 3-ga Jongno-ju
Guro-gu
Seoul, 110-745
Korea, Republic of (South Korea)

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
86
Abstract Views
702
Rank
527,705
PlumX Metrics