Bankruptcy prediction using a discrete‐time duration model incorporating temporal and macroeconomic dependencies

Abstract The purpose of this paper is to build an alternative method of bankruptcy prediction that accounts for some deficiencies in previous approaches that resulted in poor out‐of‐sample performances. Most of the traditional approaches suffer from restrictive presumptions and structural limitations and fail to reflect the panel properties of financial statements and/or the common macroeconomic influence. Extending the work of Shumway (2001), we present a duration model with time‐varying covariates and a baseline hazard function incorporating macroeconomic dependencies. Using the proposed model, we investigate how the hazard rates of listed companies in the Korea Stock Exchange (KSE) are affected by changes in the macroeconomic environment and by time‐varying covariate vectors that show unique financial characteristics of each company. We also investigate out‐of‐sample forecasting performances of the suggested model and demonstrate improvements produced by allowing temporal and macroeconomic dependencies. Copyright © 2008 John Wiley & Sons, Ltd.