Regulators have historically relied upon exchanges to monitor their members' trading activities for insider trading and market manipulation. Over the last 30 years, however, structural changes – namely, fragmentation and demutualization – have made it increasingly difficult for exchanges to do so effectively. In response, many regulators have attempted to subsume exchanges' traditional oversight role by undertaking responsibility for market surveillance or transitioning that responsibility to a non-profit, quasi-governmental organization. Their success has been mixed. To explain this variation, this paper constructs a theory of strategic interaction between regulators, exchanges and broker-dealers. The theory is tested against two case studies on the U.S. Securities and Exchange Commission's attempts to create national trade surveillance systems in the 1980s (Market Oversight Surveillance System) and 2010s (Consolidated Audit Trail). Empirical evidence is obtained from archival materials, a Freedom of Information Act request, and 60 interviews conducted in Oxford, London, Toronto, New York and Washington, D.C.