Extending the Merton Model: A hybrid approach to assessing credit quality

Alexandros Benos, George Papanastasopoulos

    Research output: Contribution to journalArticlepeer-review

    Abstract

    In this paper we have combined fundamental analysis and contingent claim analysis into a hybrid model of credit risk measurement. We have extended the standard Merton approach to estimate a new risk neutral distance to default metric, assuming a more complex capital structure, adjusting for dividend payments, introducing randomness to the default point and allowing a fractional recovery when default occurs. Then, using financial ratios, other accounting based measures and the risk neutral distance to default metric from our structural model as explanatory variables we estimate the hybrid model with an ordered probit regression method. Using the same econometric method, we estimate a model using financial ratios and accounting variables as explanatory variables and a model using our risk neutral distance to default metric as unique explanatory variable. We have found that by enriching the risk neutral distance to default metric with financial ratios and accounting variables into the hybrid model, we can improve both in-sample fit of credit ratings and out-of-sample predictability of defaults. Our main conclusion is that financial ratios and accounting variables contain significant and incremental information; thus the risk neutral distance to default metric does not reflect all available information regarding the credit quality of a firm.

    Original languageEnglish
    Pages (from-to)47-68
    Number of pages22
    JournalMathematical and Computer Modelling
    Volume46
    Issue number1-2
    DOIs
    Publication statusPublished - Jul 2007

    Keywords

    • Accounting variables
    • Credit quality
    • Distance to default
    • Financial ratios

    Fingerprint

    Dive into the research topics of 'Extending the Merton Model: A hybrid approach to assessing credit quality'. Together they form a unique fingerprint.

    Cite this