Impact Factors on Prepayment Risk in MBS Using Vector Auto-Regressive Model

Abstract

This research aims to identify and analyze factors impacting on the prepayment risk involved in MBS(Mortgage-Backed Security). To do so, Some independent variables likely to affect prepayment in MBS and 143-month time series MBS(issued by the Korea Housing Finance Corporation) data from January 2006 to November 2017 are selected and applied to a VAR(Vector Auto-Regression) model using Eviews program. Finally, the results drawn from this study is that new loan interest rate, apartment price, and unemployment rate have significant statistical impact on prepayment risk.