To drive greater efficiency, higher utilization and more flexibility in providing the tools and infrastructure the business needs to remain competitive, Merrill Lynch's business, development and infrastructure teams are collaborating on an enterprisewide grid computing project. In 2005, the firm launched two grid initiatives: a robust pricing application for vanilla and exotic derivative portfolios, and advanced analytics and simulations for risk and margin analysis. The result has been a 200 percent increase in the speed by which exotics are priced, according to Chintan Upadhyay, VP, equity derivatives, Merrill Lynch.
"This is taking off big time at Merrill," says John Cislo, the firm's director, client infrastructure solutions. "It's the largest trading and risk management application that Merrill has."
Merrill Lynch brought in DataSynapse to design and implement the grid software, particularly for its expertise in migrating applications, according to Cislo. "They have been very cooperative in helping us meet our deadline," he notes.
The business case for implementing grid computing at Merrill Lynch was easy to justify, Cislo suggests. Most of the firm's desktop capacity is idle through the night, he explains. "Being able to put compute jobs onto desktops in the still of the night gives us extra computing where we can do our QA [quality assurance] there," Cislo says. "It's like finding a $10 bill on the street -- it's found money."
This year, Cislo notes, Merrill Lynch established a Grid Computing Center of Excellence to provide a governance structure and define the long-term vision of an enterprise grid computing utility. Key objectives include boarding additional lines of business onto the grid, establishing the business model for borrowing/lending CPUs between applications in the grid, and research into future grid improvements such as data caching and automated build tools.
Now, Merrill Lynch is looking into applications of grid technology that can transform its business, Cislo relates. "By understanding our counterparty risk better, or pricing more efficiently, we'll look at how we can bring that back into our business and provide better services to our clients," he says. --