Probability of default estimation in credit risk using a nonparametric approach Articles uri icon

published in

publication date

  • June 2021

start page

  • 383

end page

  • 405

issue

  • 2

volume

  • 30

International Standard Serial Number (ISSN)

  • 1133-0686

Electronic International Standard Serial Number (EISSN)

  • 1863-8260

abstract

  • In this paper, four nonparametric estimators of the probability of default in credit risk are proposed and compared. They are derived from estimators of the conditional survival function for censored data. Asymptotic expressions for the bias and the variance of these probability of default estimators are derived from similar properties for the conditional survival function estimators. A simulation study shows the performance of these four estimators. Finally, an empirical study based on modified real data illustrates their practical behaviour.

subjects

  • Mathematics
  • Statistics

keywords

  • kernel method; local lineal fit; probability of default; risk analysis; survival analysis