Too Fast, Too Furious? Algorithmic Trading and Financial Instability

To what extent can algorithmic trading-based strategies explain the propagation of flash crashes on financial markets? This question has to be discussed at the intersection of two disciplinary fields: management of information systems and finance. Built on realistic assumptions on traders' strategies, on their use of algorithmic information systems and considering the role of transactions systems at the market level, an agent-based approach is presented. Final results show that speed-oriented trading strategies and the increasing use of new trading technologies can arm markets' stability and resiliency, facing intraday operational shocks. The article also shows the central role played by transactions systems in the propagation of flash crashes, when a new regulation based on the principle of decimalization is introduced.