Frank Ecker, Xitong Li, Yilan Li, Fan Wu
This empirical study analyzes how stock market participants use generative AI for investment information processing and its impact on the market using large-scale platform data.
There is a lack of systematic empirical evidence on how retail investors integrate generative AI into their investment decision-making. It is also necessary to determine what micro- and macro-level effects the widespread use of this new information processing tool has on financial market information efficiency and trading activity.
The study utilizes a dataset of 1.7 million stock-related queries collected from a major Chinese generative AI platform in the first half of 2024. It analyzes the topics and tasks of user queries and examines their correlation with corporate disclosure schedules, media coverage, and voluntary managerial disclosures. Furthermore, it analyzes the characteristics of answers that influence sustained user engagement, changes in users' query patterns, and the market-level impact (informed trading, liquidity, abnormal returns) using aggregated platform data.
User queries cover a wide range of topics and tasks, varying systematically with usage intensity and financial sophistication. Query activity increases around corporate disclosures but largely tracks contemporaneous media coverage. Answers that are concise and contain directionally accurate trading signals are more likely to follow continued platform engagement. At the market level, GenAI usage is associated with higher measures of informed trading and lower liquidity. This research offers insights into early-stage GenAI adoption by retail investors and informs discussions on how AI shapes information processing in financial markets.