Author(s)

Bingjie Xu, Jingjing Ding

Corresponding Author

Bingjie Xu

ABSTRACT

In order to analyse the factors that affect corporate financial risk prediction more comprehensively, this paper analysed related indicators of listed companies in China, then introduced non-financial variables such as market, macroeconomics, and managerial education into the financial risk prediction model. Finally used the classification and regression tree (CART) method in machine learning to establish a corporate financial risk prediction model. Compared with the financial risk prediction method that only considers financial variables, the prediction model proposed in this paper has higher accuracy. In addition, this paper verified that the financial risk prediction model has an industry effects, that means the model needs to consider the impact of different industries when it is applied.

KEYWORDS

Financial risk prediction, Non-financial variables, Cart, Industry effects