Tracking Retail and Institutional Investors Activity in China

Abstract

Using recent comprehensive account-level proprietary data, we validate and improve the order size algorithm that track retail and institutional purchases and sales in the Chinese stock market. We find the order size algorithm that classify small orders as small retail investors, medium orders as large retail investors, and large orders as institutions, gradually loses its power in recent years, especially after 2018. Observing patterns that small retail investors are more likely to submit orders with shares at a higher level of roundness than institutional investors, we propose tentative improvements for the order size algorithm by moving orders that are not multiples of 500 shares from retail orders to institutional orders. We find that our modified algorithm can better measure the order imbalances from retail and institutional investors, and our measurement of institutional investors’ order imbalances have stronger return predictive power.

Keywords: Order size, roundness, retail investors, institutional investors, Chinese stock market

JEL Classification: G12, G14, G15