Indonesian Political, Business & Finance News

IBM makes stock trading faster

| Source: JP

IBM makes stock trading faster

JAKARTA (JP): IBM announced yesterday it has developed a new
technology that represents a significant leap forward in the
speed and accuracy of one of the major analytical techniques used
to support securities trading.

A spokesman for PT Usaha Sistem Informasi Jaya, the Indonesian
distributor of IBM computers, said here yesterday that the new
technology, called IBM Deterministic Simulation Blaster, is the
result of more than a decade of research by IBM mathematicians
and computer scientists.

The new technology is estimated to be up to thousands of times
faster than the "Monte Carlo" simulation, which is the current
technique used to calculate theoretical prices of mortgage-backed
securities and complex derivatives.

"This is one of the most significant breakthroughs in trading
analytics in many years," the spokesman quoted Robert Howe,
general manager of IBM's Banking, Finance & Securities, as
saying.

Traders currently use software to price their mortgage-backed
securities and complex derivatives. At the heart of that software
lies the Monte Carlo simulation. But because traders often have
many securities to price, the process can take many hours and is
rarely precise.

The IBM Deterministic Simulation Blaster allows securities
firms to refine their own mathematical projections and determine
a price that may put the traders at less risk.

The IBM Deterministic Simulation Blaster also allows traders
to get important pricing information within seconds rather than
hours. Traders will now be able to accurately price their entire
inventory of investments as often as they like within the trading
day, Howe said.

Traders can even have near-real time access to the 3-D
information by using a high-performance parallel processing
computer, such as the IBM RS/6000 Scalable POWER parallel System
(RS/6000 SP).

The Simulation Blaster is designed to work with almost every
computer system available today from personal computers to
supercomputers. (31)

View JSON | Print