Deep Learning for Mortgage Risk

68 Pages Posted: 23 Jun 2016 Last revised: 9 Jul 2016

Justin Sirignano

Imperial College London - Department of Mathematics; University of Illinois at Urbana-Champaign

Apaar Sadhwani

Stanford University

Kay Giesecke

Stanford University - Management Science & Engineering

Date Written: July 7, 2016

Abstract

This paper analyzes multi-period mortgage risk at loan and pool levels using an unprecedented dataset of over 120 million prime and subprime mortgages originated across the United States between 1995 and 2014, which includes the individual characteristics of each loan, monthly updates on loan performance over the life of a loan, and a number of time-varying economic variables at the zip code level. We develop, estimate, and test dynamic machine learning models for mortgage prepayment, delinquency, and foreclosure which capture loan-to-loan correlation due to geographic proximity and exposure to common risk factors. The basic building block is a deep neural network which addresses the nonlinear relationship between the explanatory variables and loan performance. Our likelihood estimators, which are based on 3.5 billion borrower-month observations, indicate that mortgage risk is strongly influenced by local economic factors such as zip-code level foreclosure rates. The out-of-sample predictive performance of our deep learning model is a significant improvement over linear models such as logistic regression. Model parameters are estimated using GPU parallel computing due to the computational challenges associated with the large amount of data. The deep learning model's superior accuracy compared to linear models directly translates into improved performance for investors. Portfolios constructed with the deep learning model have lower prepayment and delinquency rates than portfolios chosen with a logistic regression.

Keywords: Deep Learning, Machine Learning, Mortgages, Loans, Credit Risk

Suggested Citation

Sirignano, Justin and Sadhwani, Apaar and Giesecke, Kay, Deep Learning for Mortgage Risk (July 7, 2016). Available at SSRN: https://ssrn.com/abstract=2799443

Justin Sirignano (Contact Author)

Imperial College London - Department of Mathematics ( email )

South Kensington Campus
Imperial College
LONDON, SW7 2AZ
United Kingdom

HOME PAGE: http://jasirign.github.io

University of Illinois at Urbana-Champaign ( email )

601 E John St
Champaign, IL 61820
United States

Apaar Sadhwani

Stanford University ( email )

Stanford, CA 94305
United States

Kay Giesecke

Stanford University - Management Science & Engineering ( email )

473 Via Ortega
Stanford, CA 94305-9025
United States
(650) 723 9265 (Phone)
(650) 723 1614 (Fax)

HOME PAGE: http://www.stanford.edu/~giesecke/

Paper statistics

Downloads
1,652
Rank
7,763
Abstract Views
4,497