Investor Structure and Stock Price Crash Risk in a Continuous Double Auction Market: An Agent-Based Perspective

Abstract

Multiple studies presume the institutional investors to be informed investors. However, some reports argue that this view is still under debate. In this paper, to avoid the informed investors proxy bias caused by the institutional investors, we construct an agent-based continuous double auction stock market model with both informed and uninformed investors and examine whether stock price crash risk can be affected by the change of investor structure. In particular, we employ four types of investor structures by gradually increasing percentage adjustments of informed investors from 20%, 40%, 60% to 80% within the market. We find that stock clear price and return show significant improved stability coming with the rising weight of informed investors. Beyond that, we recognize the situation that stock clear price falls below 95% confidence interval as crash event and count the number of the stock price crash events within each simulation of each different investor structure. We find that consistent with growing stability of stock clear price and return, stock price crash event number drops dramatically following the higher proportion of informed investors. These findings confirm our hypothesis that the involvement of informed investors contributes to the market stability.

Tick Size and Market Quality Using an Agent-Based Multiple-Order-Book Model

“…To avoid the quantity deficiencies of tick-size changes, this study used an agent-based multipleorder-book model, adopting six gradually increasing tick sizes to investigate the dynamics between tick size and market quality. Many studies have investigated the feasibility of agent-based modeling as an effective vehicle for stock-market simulation [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34]. However, most of those studies were oriented using one stock, which limited the credibility of the models.…”

“…First, our agent-based stock-market model originates from a multiple-order-book perspective. Using the model, our study investigated the dynamics between tick size and market quality with small-, medium-, and large-cap stocks, which improves upon the limitations of prior one-stock models [27,28]. Second, compared to previous tick-size studies [9,18], our study covers six tick sizes, providing more detailed insights into the dynamics of tick size.…”