Default analysis in mortgage risk with conventional and deep machine learning focusing on 2008–2009

Vikram Ojha and JeongHoe Lee ()
Additional contact information
Vikram Ojha: Standard & Poor’s (S&P Global Ratings)
JeongHoe Lee: Standard & Poor’s (S&P Global Ratings)

Digital Finance, 2021, vol. 3, issue 3, No 3, 249-271

Abstract: Abstract The rapidly growing mortgage market corresponds with the growth of mortgage backed securities. Since the economic crisis in 2008–2009, financial institutions that deal with mortgages have been working to develop more accurate numerical models for Residential Mortgage Backed Securities (RMBS) to minimize credit risk. Within this context, there is an increasing use of big data and artificial intelligence techniques accordingly. This research focuses on the U.S. RMBS analysis using machine learning to predict the Probability of Default (PD). Primary analysis involves the loan origination and performance characteristics and economic characteristics like home performance index (HPI) to investigate default probability in terms of credit risk. In this research, various machine learning models such as Logistic Regression, Random Forest, Linear Discriminant Analysis, K-Nearest Neighbors (KNN), Multi-layer Neural Network (MNN), Convolution Neural Network (CNN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM) are used. Ultimately, this research provides comprehensive understanding and comparison in applying various machine learning algorithms to the financial discipline of RMBS to develop predictive models for calculating mortgage credit risk using the Fannie Mae loan data that include around 1.5 million of mortgage loans originating from 2005 to 2009 in the United States.

Keywords: Machine learning; Deep learning; Ensemble machine learning (Voting); Residential mortgage backed securities (RMBS); Probability of default (PD); Default coverage ratio and credit risk (search for similar items in EconPapers)
JEL-codes: C13 C63 C80 G17 (search for similar items in EconPapers)
Date: 2021
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s42521-021-00036-4 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:digfin:v:3:y:2021:i:3:d:10.1007_s42521-021-00036-4

Ordering information: This journal article can be ordered from
https://www.springer.com/finance/journal/42521

DOI: 10.1007/s42521-021-00036-4

Access Statistics for this article

Digital Finance is currently edited by Wolfgang Karl Härdle, Steven Kou and Min Dai

More articles in Digital Finance from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().