From latent risk to market collapse: Explaining flash crashes through the Swiss Cheese Model.

This article focuses on explaining flash crashes in algorithmic trading markets through the application of the Swiss Cheese Model of accident causation, originally developed in safety engineering. It adapts this model to financial markets by identifying five critical layers of defense—circuit breakers, algorithmic risk controls, liquidity monitoring, regulatory oversight, and human intervention—each with inherent vulnerabilities that can align to cause systemic market failures. Using the 6th May 2010 Flash Crash as a case study, the paper demonstrates how simultaneous failures across these layers led to a rapid and severe market disruption. The study argues that flash crashes are not isolated anomalies but emergent phenomena in complex, tightly coupled trading ecosystems, emphasizing the need for multilayered, diverse, and independent safeguards in risk governance. It also discusses practical implications for trading firms, regulators, and infrastructure providers, advocating for integrated surveillance, stress testing of interdependencies, and enhanced human oversight within automated markets.