Tracking Retail Investor Activity, a paper co-authored by our school’s Assistant Prof. Zhang Xinran, Prof. Ekkehart Boehmer from Singapore Management University, Prof. Charles M. Jones from Columbia University, and Prof. Zhang Xiaoyan from Tsinghua University, was officially published in Journal of Finance, a world-class academic journal on finance.
Studying the behaviors of retail investors is of great importance to other market participants, researchers of behavioral finance, regulatory policymakers and retail investors themselves. In China’s stock market, retail investors contribute to over 80% of the total trading volume, but they are generally seen as lacking stock picking ability and lose money on average given that they usually buy when share prices rise and sell when share prices fall. However, some latest academic studies find that in recent years, trading activities of retail investors can correctly predict the future trends of stocks in the stock markets of developed countries. But existing research usually employs non-public and a small portion of retail investor transaction data, or identifies small trade orders in public transaction data as placed by retail investors.
Based on the trading mechanism and regulatory requirements of the US market in recent years, this paper creatively provides a simple algorithm by which to identify retail investors in the open trade and quote (TAQ) high-frequency database. TAQ contains open data, including all the intraday transaction data about stocks listed on all the exchanges in the US. In America, most marketable retail orders initiated by retail investors do not take place on exchanges, but are instead executed either by wholesalers or via internalization by brokers. Orders executed in this way are usually reported to the TAQ via a Trade Reporting Facility (TRF) set up by the US Financial Industry Regulatory Authority (FINRA).
Often, orders from retail investors are given a small amount of price improvement relative to the National Best Bid or Offer (NBBO). This price improvement is typically only a small fraction of a cent, such as 0.01 cents, 0.1 cent, and 0.2 cents. The incentive for securities wholesalers to provide price improvements is to guide brokers of retail investors to transfer retail orders to them for closing deals. Whereas brokers which are constrained by regulatory policies and internalize orders need to prove that the orders they execute for customers are optimal. Therefore, they also need to provide price improvements to retail orders. On such a basis, transactions that are just below a round penny are classified as retail purchases, while transactions at prices that are just above a round penny are classified as retail sales.
Based on this algorithm, this paper finds that retail investors in developed countries (such as the United States) show some stock picking ability, and can correctly predict the future trend of stocks. By analyzing this ability to predict returns, the authors find that part of this predictability can be explained by the continuity of order flows, while the remaining predictability shows that retail investors may possess corporate-level information about share prices.
