CONSTRUCTION AND EMPIRICAL STUDY OF A HIDDEN MARKOV MODEL-BASED STOCK PRICE PREDICTION PROCEDURE UNDER SHAREHOLDER DISENFRANCHISEMENT

Abstract

Shareholder disenfranchisement events due to shareholders' failure to pay their contributions in full and on time not only bring significant adjustments to the corporate governance structure, but also lead to unfavorable shocks to a company's stock price. By studying the impact of shareholder disenfranchisement on stock price volatility, this paper highlights the significant impact of shareholder disenfranchisement events on stock price volatility, which results in the difficulty of manually forecasting a company's stock price. To solve this problem, this paper adopts the Hidden Markov Algorithm to construct a stock price prediction model for shareholder disenfranchisement companies. By learning the historical stock price data of shareholder disenfranchisement companies, the model predicts the future trend of the company's stock price. The accuracy of the HMM model in predicting the stock price of three shareholder disenfranchisement companies, X1, X2 and X3, is higher, and the model's MAPE value in stock price prediction is significantly lower than that of other models such as LSTM. Using the model in this paper can fully grasp the movement of the company's stock price after shareholder disenfranchisement, and realize the accurate prediction of the stock price of shareholder disenfranchised companies.