Loan origination decisions based on multiple scores with application to installment loan portfolio selection

2008 2008

Other formats: Order a copy

Abstract (summary)

The loan origination decision problem is that of deciding whether or not to accept a borrower's application for a loan. Lenders are increasingly adopting broader business objectives that extend beyond a single period of risk control. This trend points to the need for a more sophisticated loan origination decision framework designed to take advantage of a collection of scores, each of which predicts a different event of interest associated with the loan (e.g., default, prepayment, bankruptcy).

This dissertation presents an analytic framework for loan origination decision making using multiple scores. I establish both statistical and utilitarian metrics for measuring the performance of multiple-score systems. I provide an analytical exploration of the conditions under which an additional score provides new information that can be the basis for improved loan origination decision making. I also study the factors that affect the magnitude of the economic value brought in by the new score. I study optimal decision policies when multiple scores are used. I present a simple cutoff policy in the form of a derived score that can significantly simplify the loan origination decision with multiple scores and up to two objectives. As an application of my theory on multiple-score decisions, I address an installment loan origination decision problem in which default and prepayment predictions (scores) are available. The multiple-outcome framework has the characteristic that it provides the decision-maker with a clear vision of the entire future risk spectrum, rather than the risk at a particular epoch in time. Consequently, the multiple-score framework yields a decision that dominates that of the single-score system in business performance objectives of interest. Finally, I address ways in which variance of portfolio profit may be considered input to the origination decision. I present an analytic model to calculate the variance of portfolio profit by assuming that the performances of two individual accounts are conditionally independent given the same economic conditions. I show that considering variance in origination decisions yields a more conservative decision policy, assuming that the decision policy is based on a predetermined cutoff score.

Indexing (details)

Operations research;
Loan originations;
0508: Finance
0770: Banking
0796: Operations research
Identifier / keyword
Social sciences; Applied sciences; Credit scoring; Installment loan; Loan origination; Portfolio selection
Loan origination decisions based on multiple scores with application to installment loan portfolio selection
Gao, Lu
Number of pages
Publication year
Degree date
School code
DAI-B 69/07, Dissertation Abstracts International
Place of publication
Ann Arbor
Country of publication
United States
Beling, Peter A.
University of Virginia
University location
United States -- Virginia
Source type
Dissertations & Theses
Document type
Dissertation/thesis number
ProQuest document ID
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Document URL
Access the complete full text

You can get the full text of this document if it is part of your institution's ProQuest subscription.

Try one of the following:

  • Connect to ProQuest through your library network and search for the document from there.
  • Request the document from your library.
  • Go to the ProQuest login page and enter a ProQuest or My Research username / password.