THE CORRELATION IN CORPORATE DEFAULTS: EVIDENCE FROM NON-FINANCIAL LISTED COMPANIES IN SRI LANKA

Abstract

Default correlation is the extent to which one company's default causes another's default. It measures the relationship between a firm’s individual default probability and the joint default probability among firms. This study makes theoretical contributions by examining the degree of default correlation among non-financial listed firms in terms of credit quality, time horizon, systematic risk, firm size, and liquidity. The researcher selected 550 firm-year observations (240 for default and 310 for non-default) consisting of 55 non-financial listed companies in Sri Lanka from 2012 to 2021. The default firms are selected using the three criteria during the selection period: suffering from losses, having negative net worth, or suffering from negative operating cash flows for more than three consecutive years. The non-default firms are selected by a couple with the default firms based on the same industry and the highest market capitalization. Altman’s (1968) Z-score model is used to determine the credit ratings as high, medium, and low. The study follows Lucas’s (1995) proposed standard binomial approach to measure the default correlation. The study revealed that time horizon, poor credit ratings, high beta, and small size increase the default correlation in the Sri Lankan context; however, liquidity is not a good measurement for default correlation analysis in Sri Lankan context. Credit risk management could underestimate the portfolio credit risk without considering the default correlation. The dynamic nature of default correlation shows that an increment in default risk of individual credit assets engages with a disproportionate increment in portfolio credit assets.