Abstract: Enterprise digital transformation came into being under the effective coupling of digital technology and enterprise development, which has profoundly impacted China's stock price. Based on the optimization of characteristic relationships, this paper constructs an ED-SPCBoost model to explore the impact of enterprise digital transformation on the risk of stock price collapse and the underlying mechanism. The findings are as follows. (1) Digital transformation can effectively reduce the risk of the stock price crash, and the risk of stock price crash shows a phased downward trend of "stable-sensitive-stable" as the degree of transformation increases. (2) Digital transformation plays a more significant role in reducing the risk of the share price collapse of state-owned enterprises. In the mature period of transformation, non-state-owned enterprises account for 78.03%. Non-state-owned enterprises are more willing to accept digital transformation than state-owned enterprises. (3) Digital transformation can reduce the risk of stock price collapse by enhancing enterprise information transparency, improving market evaluation expectations, and improving internal financial stability. The mechanisms underlying the three paths of impact are different. The ED-SPCBoost model proposed in this paper has high robustness and a low error rate. Empirical data verify that it aligns with the prediction law of digitization and stock price collapse. It can provide a valuable reference for digital technology to optimize the governance environment of listed companies and improve the monitoring of financial market risks.

Key words: enterprise digital transformation, risk of share price collapse, information asymmetry, ED-SPCBoost model