Refinance Incidence and Information Loss in Predicting Prepayment Risk

29 Pages Posted: 16 Jan 2020

Date Written: December 29, 2019


This paper uses loan-level data to investigate heterogeneity in loan prepayment incidence, and argues that refinancing is affected by a mortgage pricing convention that underestimates co-borrowers' actual creditworthiness. Specifically, we find a substantial difference in prepayment incidence between sole borrowers and co-borrowers. We also find this difference to vary across sub-sets of co-borrowers in a predictable manner. To address endogeneity concerns, we exploit the variation across time in mortgage rates to confirm that the difference in prepayment incidence exists only during a period of declining mortgage rates. At an aggregate level, we find that geographic areas with higher concentration of co-borrowers are having a higher prepayment rate. These results are directly relevant to the valuation of mortgage-backed securities by offering an additional explanation for the observed variation in refinancing decisions that is related to institutional aspects of the loan process.

Keywords: Mortgages, prepayment, MBS pricing, risk-based pricing

JEL Classification: G21, G23.

Suggested Citation

Tzioumis, Konstantinos, Refinance Incidence and Information Loss in Predicting Prepayment Risk (December 29, 2019). Available at SSRN: or

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

Paper statistics

Abstract Views
PlumX Metrics