Forecasting The Withdrawal Risk Of The Chinese Listed Firms Base On The Latent Variable Model

This paper researches withdrawal risk problem of listed firms. Currently, the problem has become to be one of the social focus issues. However, researches on estimation of withdrawal risk are still in a beginning stage in our country, therefore, there are profound theoretic and realistic meanings in furthering researches.This paper uses latent variable threshold model, which evolves from credit risk structural model, to analyze and estimate the withdrawal risk problem of listed firms. In this article, some variables are introduced into the model, such as financial variable and administration structure variable in listed firms, macro economy variable, and an unpredictable variable which is used to depicting unquantifiable factors. The unpredictable variable can be comprehended as those factors who have stronger explanation strength to dependent variables and can't be quantifiable, such as supervising strength of Securities Commission's, improvement of relative laws, and promulgation of macro policies. In this paper ,we compared credit risk structural model with latent variable threshold model , proposed how to model withdrawal risk using latent variable threshold model, proved the validity of the method.This paper forecast withdrawal risk and withdrawal correlation of listed firms, using ML of integrated latent variable, base on latent variable threshold model. And the latent variable obeys a probit distribution.We also compared the annual withdrawal probability of listed firms and annual the average withdrawal probability and the real withdrawal ratio. The empirical result also shows that the bringing in of the latent variable gives better analysis and expression of the withdrawal risk.