Equity Trading Exchange—Trader Malfeasance Prevention

Summary:  Booz Allen Hamilton developed an automated and statistically driven market surveillance algorithm that enabled a stock exchange to reduce the risk of illegal trading based on pre-defined parameters.

Client Challenge:  The client was transitioning to a new, automated trading environment with a hybrid model that needed a new set of surveillance mechanisms. In this trading model, the execution of buy and sell orders would be controlled by a third party, which created the potential for illegal trading activities. The client required a statistically driven methodology for identifying instances of potentially illegal trading.

How Booz Allen Helped:  Booz Allen analytics experts developed a program that utilized historical order flow information and the trading rules of the new platform to simulate trading data in the new market environment. We explored a variety of statistical methods to detect the presence of rogue trading schemes. We then inserted rogue trading algorithms into the simulated transaction order flow, and developed measures to determine the detection power of selected statistical surveillance methods.

Results Achieved:  Our analysts developed a number of measures of trader malfeasance, including both the frequency and magnitude of rogue trading opportunities taken. The statistical tests we selected were able to detect all significantly profitable rogue trading schemes. The program Booz Allen developed gave the client powerful insight into trading activity in its forthcoming automated environment.