.¹ú²ú¶ÌÊÓƵ

Info Systems Professor Donald Berndt Lands Externship with the Treasury Department

By Keith Morelli

TAMPA (Jan. 25, 2017) -- Just as Muma College of Business marketing professor Mike Mondello returns to class from his Tampa Bay Rays externship – in the front office, not in centerfield – a second professor is getting ready to take his turn in the real world.

Don Berndt, associate professor with the Information Systems Decision Sciences Department, is getting ready to head to Washington to share his expertise with the U.S. Treasury Department's Office of Financial Research.

His travels are part of a new externship program in which professors opt out of classes for a while to work in the real corporate world. There, they address issues using their research to solve practical problems.

"The dean and other leaders have come up with a truly innovative program," Berndt said. "I am the second faculty member to participate, but I am pursuing a somewhat different externship driven by ongoing research."

Under the externship program, which began last year, the private sector, or in this case, government, gets a seasoned Muma College of Business professor, who can use solid research and academic wisdom to address various challenges within the entity to which they are assigned. The faculty member gets real world experience that can be brought back for timely and relevant discussions in the classroom.

There is no cost to the business since ¹ú²ú¶ÌÊÓƵcontinues to pay the faculty members' salaries.

This ground-breaking plan also strengthens the interaction between the university and its surrounding community, be it local or across the state, the nation or the globe.

Berndt's research, which he will share with his Washington counterparts, involves figuring out systemic risks in financial institutions.

"A couple of years ago, I was fortunate to receive a $300,000 multi-year grant from the Office of Financial Research administered through the National Science Foundation," he said. "The focus of the grant is to come up with computing-oriented approaches for better understanding financial systemic risk, an area of tremendous interest following the recent financial crisis."

While most systemic-risk models are bound by available data and computing platforms, Berndt said, many researchers now say that the data needed to better understand that risk lies in the details and are working on models that include unrelated actions of individuals and groups that typically are discounted.

"The overall objective of the Washington project is to accelerate our understanding of systemic-risk monitoring approaches that leverage granular data (transactional, positional and other information) by establishing a model for underlying distributed computing approaches," he said.

Such a model, he said, could go a long way in understanding the ebb and flow of financial institutions, the stock market and other investment vehicles.

"We certainly live in an increasingly interconnected world," he said, "and our financial systems are a good example of the complex systems that surround us."

He and fellow ¹ú²ú¶ÌÊÓƵresearchers James McCart and Saurav Chakraborty, along with David Boogers of Finametrics.com, have implemented an agent-based model of the U.S. corporate bond market, with classes of agents that include mutual funds, insurance companies and hedge funds, he said. An agent-based model is a computational model for simulating actions and interactions of individuals or groups and the effect those actions have on systems.

"This gives us a laboratory to assess the impacts of regulatory rules, financial policies and structural changes on market dynamics under conditions of stress," Berndt said. "These simulations are a natural way to capture the many feedback loops in such complex systems.

"For me, this new stream of research is very special," he said. "As a student, I was captivated by research in artificial life and evolutionary computation, reading the working papers coming out of places like the Santa Fe Institute."

Berndt said he has never before had an opportunity to use agent-based methods in assessing risk and this externship will give him that chance.

"I am excited to have this chance now, especially in the context of an externship," he said. "Several talented researchers at the Office of Financial Research also have embraced agent-based modeling and simulation as a tool for understanding complex financial systems."

Berndt, whose teaches information systems, database management and data mining, said he plans to spend about a week each month during the spring semester in Washington sharing his expertise with the government financial researchers.

"My goals," he said, "are to gain insights into our research, uncover opportunities to collaborate on future projects and develop materials to bring agent-based modeling into the classroom."