Fragmentation and Algorithmic Trading: Joint Impact on Market Quality

Abstract

We investigate the effects of algorithmic trading and dark venues on U.S. security market quality. Market quality refers to the nearly universal mandate of securities regulators to ensure that all market design changes should not detract from efficiency or fairness which we define as less market manipulation. A simultaneous equations model is used to estimate how fragmentation of the lit and dark order flow interacts with algorithmic trading to determine round-trip transaction costs and end-of-day manipulation. We find that lit market fragmentation lowers effective spreads and reduces manipulation but that fragmentation into the dark raises effective spreads and increases manipulation.