Research & Innovation

Crash Test: Stevens Builds a New Tool to Probe Financial Markets for Potential Dangers

The days of open trading pits and frenzied brokers waving chits of paper have disappeared over the last decade or so, swiftly replaced by electronic markets and algorithmic trading. Ever since, regulators and practitioners have raced to keep pace with ever-more rapid changes in trading technology.

Now CME Group Foundation  — the philanthropic arm of the largest exchange in the world — has awarded Stevens Institute of Technology a contract to perform a series of financial research projects that may reshape the way federal regulators prepare for electronic trading events. The research will not only help spot illegal trades; it will also help both researchers and agencies stay abreast of the tremendous quantities of routine automated trading activity happening daily at light speed.

"Things have changed, and very quickly," says George Calhoun, director of Stevens' Financial Systems Center and the university's pioneering undergraduate program in Quantitative Finance. "Finance is becoming a hard science, as technical as chemistry or biology. Systems are vulnerable, and markets and regulators need to get out in front of this as quickly as they are able."

That's where Stevens comes in.

The CME Group Foundation-sponsored suite of four projects includes an investigation of applications of quantum computing to complex financial problems; the creation of the world's first high-frequency finance journal, which will be based at Stevens; and support for the university's annual October high-frequency finance conference, the largest such conference in the world.

One of the most exciting components of the Stevens-CME collaboration is sHiFT: an ambitious effort to build a new simulation platform, from scratch, that will run real-time market data and introduce actual high-speed trading scenarios into the market flow to test global markets and exchanges for weaknesses and vulnerabilities.

"There's simply no tool like this currently available for regulators and researchers," says Calhoun. "Its scope will be broad and the platform will run live market data from all markets available."

Student ingenuity, computational firepower

Drawing on the computational power of the Hanlon Financial Systems Lab, a small graduate-student team is writing software that will enable multiple complex strategies and high-frequency algorithms to be simultaneously entered and tested to study interactions of different models and strategies with one another and the dynamics of the resulting asset prices. Market orders will be routed and matched much as they are in actual high-frequency electronic markets. The platform will also possess the capability to record and store both live market data and historical data, enabling repeated testing of alternate scenarios.

In addition to the painstaking coding of the software, student research will be vital to the creation of the algorithms used.

"As part of each graduate financial engineering student's curriculum at Stevens, we have a capstone course during which students are required to complete a practical project related to finance and financial markets," explains Ionut Florescu, director of the Hanlon Lab and the lead researcher behind the sHiFT project. "Many students chose to work on the sHiFT project and devise trading strategies, thus gaining valuable hands-on market experience. These strategies reflect real market choices and will be implemented into the actual software."

In addition to stress-testing financial markets, sHiFT will also be useful in testing the impact of new or proposed electronic-trading regulations — in any nation or jurisdiction — simply by implementing these rules in the sHiFT system and observing the resulting impact, Florescu adds.

The first commercial version of sHiFT is expected to become available by mid-2016 — future iterations will extend beyond equities modeling to energy trading, futures, options and treasuries — and the university is already exploring potential partnerships with academic and industry partners to market and distribute the platform.

The Stevens quantitative finance and financial engineering group is also conducting research with additional partners, as well, including the Montreal Exchange, notes Calhoun.

"Trading and finance are no longer about the open trading pits," he concludes. "The era of person-to-person execution has long since passed. Today, with trades overwhelmingly electronic, they are about technology, about quantitative thinking, about computer science — things we teach in the Stevens curriculum from the fall of freshman year."

"Now Stevens is in a unique position, with its proximity to Wall Street and the power of the Hanlon Lab, to become one of the nation's research leaders in this rapidly growing field of high-frequency finance. I know of no other financial research lab like this in the Northeast. In fact, there are very few in the world."