Training for Wall Street used to mean an MBA degree in finance and a smattering of computer courses. Not anymore.

At Cornell University, finance and engineering students are putting financial models and applications to the test on IBM's largest supercomputer, the 512-node Scalable RS/6000 POWERparallel (SP) Systems at the Cornell Theory Center (CTC).

Cornell students are experiencing the rigors of applying financial models to real-world problems in a new course called Applied Financial Engineering. "Students aren't just hearing about hedging and trading strategies, they're testing them," explained Robert A. Jarrow, the Ronald P. and Susan E. Lynch Professor of Investment Management at Cornell's Johnson Graduate School of Management. "That means formulating the model, programming the computer, collecting the data and analyzing the results."

The experience is part of Cornell's new Financial Engineering program, developed jointly by the Johnson School and the College of Engineering. More than 35 MBA and master's of engineering students are participating. "We expect this course to expose Cornell students not only to one of the most robust computing systems in the world, but to the latest advancements in software," said David C. Heath, the Merrill Lynch Professor of Financial Engineering at Cornell's School of Operations Research and Industrial Engineering.

Working in teams, Cornell's applied financial engineering students tackle multiple projects. One team, for example, is testing PDESolve, a new software tool from a Cornell University spin-off company called BEAM Technologies Inc. "PDESolve could be used by finance industry professionals to solve derivative valuations problems with considerably less code development time," said Gal Berkooz, president of BEAM Technologies.

PDESolve is a high-level language that solves partial differential equations and defines geometry, boundary conditions, and discretization methods. "We expect that the use of PDESolve by Cornell's financial engineering students will demonstrate its value as a modeling tool for pricing and hedging exotic securities," Berkooz said.

Besides helping to test PDESolve, the Theory Center is testing MultiMATLAB, a new toolbox designed to facilitate parallel computation with MATLAB. MATLAB, a product of The MathWorks Inc., is used by quantitative analysts and portfolio managers for financial modeling.

"CTC is interested in helping to better prepare Cornell students for the business world while testing new software and assessing the value of parallelizing financial codes for speed-up," explained Peter M. Siegel, director of the Theory Center's Corporate Partnership Program. "We're eager to keep the financial industry of New York state at the forefront of algorithmic development as well as numerically intensive and data-intensive computing."

CTC is one of four high-performance computing and communications centers supported by the National Science Foundation. Its activities also are funded by New York state, the Advanced Research Projects Agency, the National Center for Research Resources at the National Institutes of Health, IBM, and other members of CTC's Corporate Partnership Program.