ST Enterprises Survival Analysis Base On KMV Model

ST enterprises have occupied lots of social resources. The conditions of their future have relationship with vital interests of the stakeholders such as investors, creditors, enterprise managers, regulators. Credit conditions of ST enterprises play an important role in the survival condition of the enterprises. Most of previous studies used traditional credit measurement methods, such as solvency indicators, to describe the credit condition of ST enterprises. The reality requires putting forward ST enterprises credit risk measurement technology. KMV model that combines the thought of Merton model and Black-Scholes option pricing model brings a new perspective for credit risk measurement. At the meanwhile, in the process of researching ST enterprises researchers often have to deal with censored data. Research method that can handle censored data is needed urgently. Survival analysis method, one of the hottest fields of statistics, can handle censored data. Survival analysis began to be widely followed since the mid-70s and put forward as a representative branch by the committee of the national academy in 1986.This article selects all the ST enterprises as the research samples from 1998 to 2011. Firstly, use the KMV model to calculate default distance and default probability of ST enterprises, which combine other financial indicators to build the risk factors index system of ST enterprises.Secondly, use the method of statistical analysis, life table analysis and Kaplan-Meier analysis to process survival data of ST enterprises, looking for the laws of ST enterprises survival data. Again, make use of the established index system to build COX model, which can identifies the key factors of ST enterprises survival time and calculate the survival probability of ST enterprises. Finally, ROC curve is used to test the prediction ability of model and find the ideal threshold of survival probability.This article concluded that the ST enterprises cost 2.953425 years to pick hat on average, KMV model’s default probability and equity structure indicator have significant effects on the future conditions of ST enterprises, the model is good at judging which ST enterprises have the ability to pick hat and the ideal threshold of survival probability is 0.74 which has a sensitivity of 0.802 and a 1-specific of 0.156.