Survival Analysis of Corporate Delisting Risk in the Korean Stock Market

This paper applies survival analysis to model corporate delisting risk among Korean Exchange (KRX) listed companies over the period 2000-2023. Using a sample of 365 firms drawn from the DART disclosure system and FinanceDataReader (KOSPI: 127, KOSDAQ: 238; 158 delisting events, 43.3% event rate), we estimate Kaplan-Meier survival functions, Cox Proportional Hazards (Cox PH), and Accelerated Failure Time (AFT) models to examine how financial fundamentals and market characteristics determine time-to-delisting. Our principal findings are: (1) KOSDAQ-listed firms face significantly higher delisting hazards than KOSPI counterparts (log-rank ,); (2) 12-month price momentum (,) and annualized return volatility (,) are the dominant predictors of delisting risk-markets appear to price corporate distress in equity prices before it manifests in accounting statements; (3) accounting ratios (leverage, profitability) carry the expected signs but lack individual significance at this sample size; and (4) the Weibull AFT model with heteroscedastic scale achieves the best fit (AIC = 1620.6, over the homoscedastic specification), confirming that failure-time variance itself differs significantly across market segments. The overall model discrimination is C-index = 0.738, competitive with Shumway (2001)'s US hazard model benchmark (). This framework offers a principled, censoring-aware alternative to static scoring models (e.g., Altman Z-score) for Korean delisting risk assessment.