18 June 2002
HSBC, ClusterTech and CUHK
to Develop Artificial Intelligence
Technology for FX Flow Analysis
The Hongkong and Shanghai Banking Corporation Ltd is joining forces with
Hong-Kong-based Cluster Technology Ltd (ClusterTech) and The Chinese University
of Hong Kong (CUHK) to apply artificial intelligence and optimisation methods
to forecast market movements and capital flows in HSBC's growing Internet-based
foreign exchange business. HSBC operates Asia's leading online FX
trading platform.
ClusterTech will develop a multi-faceted system for HSBC that will, among
other things, provide value-added analysis and forecasting of capital flows
within the HSBC franchise. The system will also be used for modeling
of the foreign exchange market, and to provide exchange rate forecasts.
"We have been planning this project for some time and are now confident
that the right mix of expertise, data and market knowledge has been brought
together," said Michael Powell, HSBC's Head of Treasury and Capital Markets,
Asia-Pacific.
The new software will be based in part on software component libraries
co-developed by ClusterTech and CUHK under a project sponsored by the Hong
Kong Government's Innovation and Technology Fund (ITF).
The software libraries being developed in the ITF project are specifically
targeted at financial applications, and are written for parallel execution
on clusters of workstations to ensure that the most comprehensive analysis
can be produced in the shortest possible time. The technology behind
parallel execution of software on multiple CPUs is widely used in scientific
research, ranging from the study of black holes to the sequencing of DNA.
One area of focus of ClusterTech is to bring this capability to commercial
applications.
"This project is a great example of partnership between university technical
know-how, industry innovation and government sponsorship," said Professor
Kenneth Young, Pro-Vice-Chancellor of The Chinese University of Hong Kong.
"I am delighted that we have HSBC and ClusterTech as partners in exploring
the application of cutting edge technologies to an area where Hong Kong
is a regional leader."
ClusterTech will deliver software that provides a framework for the development,
evaluation and optimisation of time-series forecasting models based on
artificial neural networks. A proprietary backtesting engine - used
for testing trading strategies based on historical data - developed by
HSBC will be incorporated to provide realistic market conditions for model
evaluation. ClusterTech will also provide a supercomputing platform
on which the optimisation can be performed.
The system will implement a model conceived by HSBC's Barry Flower, Ph.D.,
Manager E-Commerce Strategic Development, who has more than 15 years experience
in modeling and forecasting systems. "Combining HSBC's market knowledge
and detailed flow data with the artificial intelligence software and parallel
computing power of Cluster Technology will result in FX models of unprecedented
detail and accuracy," Dr Flower said.
"Having HSBC as an industrial partner in the early stage of the ITF project
will be a tremendous asset," said Dr Suen Wai-mo, CEO of Cluster Technology.
"It will help in the honing of the project deliverable and ensure that
it is designed to serve the needs of market practitioners."
"It is very exciting to be able to run analysis on the relationship between
capital flows and exchange rate movements, particularly in markets where
HSBC has a dominant market share," said Dr Kenneth Chow, COO of ClusterTech
and himself a 13-year veteran of financial modeling. "I believe this
is the first time where serious computing power is applied to this kind
of analysis."
Disclaimer:
"Any opinions, findings, conclusions or recommendations expressed in
this material/event (or by members of the ITF project team) do not reflect
the views of the Government of the Hong Kong Special Administrative Region
or the Innovation and Technology Commission.¡¨
|