Analysis of the Influence Short-Selling has on Crashes by using Agent-Based Simulation

Abstract

Financial markets have suffered large scale crashes, like the 1929 US stock market crash, and the global economy has been influenced significantly as a result. In this study, we focus on the relationship between financial crashes and short-selling, the latter of which is an important trading method in the modern financial market. To this end, we have developed an agent-based model incorporating the short-selling on the basis of a previous study by Friedman and Abraham. Portfolio-managers are modeled explicitly in the model, whose investment strategies, namely the leverage levels, are adjusted according to the payoff gradient. Like the previous study, comparison of the numerical result with the analytical solution is used to validate the model. We analyze short-selling's influence on the statistics of crashes through a series of simulations, in which both macro- and micro-mechanisms have been carefully investigated. As a result, we find that the volatility of markets with short-selling becomes about twice as large as that of markets without short-selling. In the meantime, the crash frequency increases, which means that short-selling could make financial markets more unstable. We also find that crashes are caused by agents' accidental loss which drives other agents to lower their leverage. On the other hand, traders in a market with short-selling become more active right after the crash than those in a market without short-selling.