Stevens students develop new computer tools to stop stock trading fraud
The stock market is booming, and while corporations and investors are reaping profits, financial criminals are also looking for easy cash. But a pair of Stevens students and their professor are designing new tools to help catch schemers before they cheat.
The students, doctoral candidates Stavros Tsarpalis and Sheung Yin Kevin Mo, work with Dr. Steve Yang, assistant professor with the School of Systems & Enterprises. Together they have developed computer simulations to automatically check for trading irregularities.
“The goal of this project is to create trading analytical tools and an automated market surveillance system that financial market regulators can use to investigate trading fraud,” Yang said. “The idea is for them to be able to automatically identify unknown malpractices on the market.”
Tsarpalis called trading fraud a serious problem. “The effects of fraud in financial markets ripple through the whole system with broad, unpredictable and at times catastrophic results,” he said. “It’s a systemic risk that needs to be addressed and prevented at the proper levels.”
Yang and the students run various simulations of stock markets to detect fraud. For example, they may run a simulation showing X, Y and Z to be a normal trading sequence, which would tip regulators to potential irregularities in real life if they ever see X, Y and Q happen. Or the Stevens team might run a simulation with X, Y and Q so that regulators can understand more about what Q looks like.
The Stevens team is operating with a research grant from Northrop Grumman. The company has assigned two employees to work with the Stevens team, and it will continue to fund the project into the next year.
During a recent visit to campus, Jim Sowder, the chief technical officer of Northrop Grumman’s civil financial group, said he was very impressed with the students’ work on the project.
"I found their presentation to be very informative and impactful,” Sowder said. “Stevens is developing an impressive program with students well-prepared for entry into industry or academia.”
There are multiple regulatory agencies, both governmental and independent, that look for trading fraud, including the Securities and Exchange Commission, the Commodity Futures Trading Commission, and Financial Industry Regulatory Authority. However, Yang said that trading fraud has become more sophisticated because of technology, forcing regulators to develop new ways of keeping pace.
“Trading fraud has become harder to detect these days, and it’s exerting a bigger impact on the financial market,” Yang said. “Financial regulators need new advanced technology and tools to battle these forces.”
Given the size and scope of the American trading market, fraudsters have an incentive and opportunity to target millions and perhaps billions of dollars. Yang cited one recent case where the CFTC fined Red Bank, New Jersey based Panther Energy over $2.8 million dollars for using new trading algorithms to manipulate commodities markets. Panther had been projecting a false impression of wide industry demand, which would inflate the company’s sale price for its shares.
Yang said that financial regulators have already expressed interest in the Stevens project and are eager to implement the new technology. He estimated that with more development the system could be ready for use within two years.
He added that companies, investors, regulators and the general public all have an interest in stopping trading fraud. “We need a healthy market where trading is fair and where people can participate with a sense of trust,” Yang said.
Tsarpalis agrees. “We strongly hold that for a market to thrive there needs to be a level field for all players,” he said. “Our concentrated efforts in fraud detection contribute in creating a secure and resilient market and therefore an efficient and more profitable one.”