Professors search for key to economic crisis

Three professors at the University just may hold the key to understanding the national economic crisis.

Professors Anne Villamil, Stefan Krasa and Jamsheed Shorish in the department of economics are currently working on a project that could predict different outcomes for small businesses financially, and in turn, for the entire economy.

“In economics, entrepreneurs are very important,” Villamil said. “Between 50 and 80 percent of Gross Domestic Product is accounted for by entrepreneurs.”

Villamil said that the project is part of a research program that works to build models that evaluate alternative policies and show their effects on small businesses. The model could then show different outcomes related to “the size of firms, firms’ capital structure, firms’ access to credit, default rates, all those kinds of things that we see in the economy right now,” Villamil said.

The project itself is a branch-off from a similar project Krasa and Villamil launched a few years ago, studying the characteristics of what makes an entrepreneur. The current project started in fall 2008, when professors applied for a grant at the National Center for Supercomputing Applications for use of the supercomputer.

Krasa said that because this project deals with too many parameters and large numbers, 300,000 service units (basically 30 years of computing time) is needed in order to process data and make predictions on a larger scale.

Tricia Barker, senior public information specialist at the NCSA, said that people use the supercomputer because it can calculate quickly.

“If they were to run [a program] on their own computer or cluster of computers, it might take months or even years,” Barker said. “If they move it to a supercomputer, it could take days or hours.”

Although the grant proposal is still waiting for approval, professors hope to see results this spring.

Professors take research gathered from the Federal Reserve on businesses’ past financial outcomes and decisions, and put it into the program. The computer will take the data and make a prediction on the outcome of certain choices, and professors then compare that with the actual outcome.

“So far the predictions have been accurate,” Krasa said.

The importance of the project not only deals with financial decision-making for businesses, but also with how things can be changed in the future. One major problem over the years has been the fact that different policies have been used to fix the economy, but with little insight as to how they will actually turn out, Villamil said.

“We try something, it doesn’t work. Then we try something else, and it also doesn’t work,” she added.

The model would allow for less costly experimentation, and actually allow for insight as to which policies would work better in certain situations. Villamil said the model could also help in making policies to help entrepreneurs operate better in the future. Not only that, but various changes in the way businesses are operated and the consequences of such changes can also be analyzed.

Krasa noted that currently, the program is working with smaller-scale data, analyzing a few businesses first before applying larger sets of data. The professors hope to move on to bigger things after making the system as accurate as possible.

“Our goal would be to see what would happen on a national level,” Villamil said.