Share Distribution

In an AI-driven Taiwan stock market, mastering chip distribution (ownership structure) is the key to profitability. This study delves into the Share Distribution data from the TDCC, transforming 15 tiers of shareholding data into behavioral finance factors such as investor attention, opinion dispersion, and retail speculation. By utilizing Fama–MacBeth two-stage regression and the alphalens-tej quantitative tool, we precisely validate the predictive power of psychological biases on stock returns, providing investors with actionable Alpha strategies and robust risk management solutions.

Share Distribution data is a dataset published by the Taiwan Depository & Clearing Corporation that classifies investors into 15 groups based on the number of shares held. This information reveals the shareholding ratios and number of shareholders for each group to observe investor behavior. By analyzing these distributions, the data can be transformed into quantitative factors to predict the relationship between investor trading behavior and expected stock returns.

In the Taiwan stock market, retail investors account for the highest proportion and have the highest trading volume in the market. This article uses factor analysis to delve into the relationship between retail investor trading behavior and stock expected returns.