Dynamic Prediction of Financial Distress in Hedge Funds and Funds-of-Hedge Funds

Abstract

This study establishes survival/hazard models with time-varying covariates, as well as fixed covariates, under three specifications of the Cox Proportional Hazards (CPH) model for Hedge Fund (HFs) and Funds-of-Hedge Funds (FOHFs). With the development of the SAS Macro program for generating survival probabilities, the dynamic changes in survival probabilities are predicted over the lifetime of HFs and FOHFs after the estimation of baseline hazards using mixed CPH model. This model incorporates both fixed and time-varying covariates. The resulting dynamic survival probabilities show that the mixed model developed in this study is effective for real-time prediction of the cumulative survivals of HFs and FOHFs. The estimated models exhibit satisfactory predictive accuracy in forecasting the occurrence of failures in HFs and FOHFs.