Flash crashes in algorithmic trading markets, exemplified by the event on 6th May, 2010, reveal vulnerabilities that extend beyond isolated errors or single-point failures. This paper introduces the Swiss Cheese Model — originally conceptualised in aviation and safety engineering domains — as a systems-level framework for comprehensively understanding the simultaneous failures of multiple defence layers within financial markets. The model is adapted to the intricate structure of contemporary trading ecosystems through a detailed conceptual analysis and case study approach, identifying critical defence layers, including circuit breakers, algorithmic controls, liquidity monitoring, regulatory oversight and human intervention. The paper advances theoretical understanding by integrating insights from Normal Accident Theory and the principles of complex adaptive systems, providing practical guidance for risk governance. The Swiss Cheese Model functions as a structured vocabulary and diagnostic framework for analysing emergent failures in high-speed markets, emphasising the necessity for multilayered, diversified and resilient defences in designing algorithmic trading systems and regulatory frameworks. This paper offers a novel approach to understanding and mitigating the risks associated with flash crashes by shifting the focus from single-point causality to systemic alignment. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Steven R. Haynes (Sun,) studied this question.