A Mathematical Method for Deriving the Relative Effect of Serviceability on Default Risk

Graham Andersen,David Chisholm
DOI: https://doi.org/10.48550/arXiv.1111.5397
2011-11-23
Risk Management
Abstract:The writers propose a mathematical Method for deriving risk weights which describe how a borrower's income, relative to their debt service obligations (serviceability) affects the probability of default of the loan. The Method considers the borrower's income not simply as a known quantity at the time the loan is made, but as an uncertain quantity following a statistical distribution at some later point in the life of the loan. This allows a probability to be associated with an income level leading to default, so that the relative risk associated with different serviceability levels can be quantified. In a sense, the Method can be thought of as an extension of the Merton Model to quantities that fail to satisfy Merton's 'critical' assumptions relating to the efficient markets hypothesis. A set of numerical examples of risk weights derived using the Method suggest that serviceability may be under-represented as a risk factor in many mortgage credit risk models.
What problem does this paper attempt to address?