An Agent-Based Model of the Flash Crash of May 6, 2010, with Policy Implications
39 Pages Posted: 7 Oct 2013 Last revised: 27 Feb 2014
Date Written: February 26, 2014
Abstract
We describe an agent-based framework that successfully simulates the key aspects of the most famous flash crash in history: the Flash Crash of May 6, 2010. In our model, market making high-frequency traders collectively create a feedback loop system triggered by a large institutional sell, consistent with the widely cited "hot-potato effect." With the help of simulations, we discover functional relationships between the number of HFT market makers, their inventory sizes or speed, and the probability of another similar flash crash. The model can be used for stress-testing algorithms before their production stage and to give sounder policy advice.
Keywords: agent-based model, feedback loop, flash crash, high-frequency trading, hot-potato effect, market regulations, tick size
JEL Classification: C15, C63, G01
Suggested Citation: Suggested Citation