Research on Financial Early Warning of Chinese Listed Companies Based on Z-value Model

This paper adopts z-value financial warning model, which has been widely used and recognized in recent years, as the research model. The selected z-value warning model includes Z-Score model, ZETA model and modified Z-Score model. Select the listed companies declared st in Shanghai and Shenzhen from 2018 to 2020 and non ST listed companies as the research samples. Study and evaluate the effectiveness and applicability of the three models in the financial early warning three years before being declared St. The results show that the modified Z-Score model has the best test effect on ST company among the three models, and the average prediction accuracy is 83.2% in the first three years of ST. However, the discrimination of non ST companies by this model is far inferior to that of ST companies. At present, if the modified Z-Score model is used to do financial early warning and forecast, auxiliary analysis means such as model index analysis and z-value volatility analysis are needed to improve the accuracy of identifying financial crisis companies.