Investigating the impact of company announcements on stock prices: an application of machine learning on Australian lithium market

Abstract

This paper investigates the effects of various types of announcements made by lithium producers on stock prices. We collected data from 40 lithium-producing companies listed on the world's largest stock exchanges, spanning from May 2020 to September 2022. To analyze the impact of announcements such as quoted and unquoted securities, market announcements, company reports, public meetings and presentations, financial announcements, and technical announcements on stock prices, we employed an extreme gradient boosting (XGBoost) model. Our results indicate that stock exchange market announcements and announcements about public meetings and presentations significantly influenced the stock prices of all eight large-cap companies studied. Announcements about public meetings and presentations were crucial predictors of stock prices for 73% of all companies analyzed. Additionally, positive financial announcements were key predictors for 70% of the companies. These findings suggest that investors should consider these predictors when making investment decisions in the lithium-related stock market. This study contributes to the existing literature by providing empirical evidence on the impact of different types of announcements made by lithium producers on stock prices. Furthermore, the XGBoost model used in this study can be applied to other industries and markets to analyze the impact of various types of announcements on stock prices.