Predicting Early Warning Signals of Financial Distress: A Stochastic Approach*