Downloads: 21

Files in This Item:
File Description SizeFormat 
jor.2022.032.pdf7.6 MBAdobe PDFView/Open
Title: A new approach to detecting change in credit quality
Authors: Kevkhishvili, Rusudan  kyouindb  KAKEN_id  orcid (unconfirmed)
Keywords: credit quality
credit risk
mean reversion
long-term mean
Issue Date: Jun-2022
Publisher: Infopro Digital Services Limited
Journal title: Journal of Risk
Volume: 24
Issue: 5
Start page: 51
End page: 73
Abstract: This paper provides a new framework for the detection and quantification of change in the corporate credit quality of established companies. These companies are important players in the economy, and analysis of their creditworthiness is of great interest to investors. Our approach is based on the observation that the deterioration of credit quality in established companies is a long-term process, because these firms use their resources as well as strong customer and supplier relationships to respond to market and technological changes. Based solely on the realized path of the stochastic process that represents creditworthiness, we propose an economically plausible endogenous mechanism that governs the shift in the long-term mean of the process. Specifically, we construct a model that detects a change in the mean-reversion level of an Ornstein–Uhlenbeck process representing the company’s leverage. Our model captures this change based on a number of realized upcrossings of a certain level (estimated from data) by the process itself, without introducing an additional source of uncertainty. Our approach is computationally simple and provides an efficient tool for monitoring changes in quantities related to credit quality, such as default probability.
Rights: © 2022 lnfopro Digital Risk (IP) Limited
The full-text file will be made open to the public on 12 JULY 2023 in accordance with publisher's 'Terms and Conditions for Self-Archiving'.
DOI(Published Version): 10.21314/jor.2022.032
Appears in Collections:Journal Articles

Show full item record

Export to RefWorks

Export Format: 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.