Research Projects Supported by HKU's High Performance Computing Facilities
¡@
Researcher:
Dr Philip L H Yu, Department of Statistics and Actuarial Science
Project Title:

Statistical Methods in Finance

¡@
Project Description:

Statistical methods have been widely used in various problems in finance.  In this project, we focus on three topics, namely market risk measurement, liquidation risk management, and technical trading study.  The first topic involves the estimation of value at risk, one of the most important measurements of market risk.  The second one discusses how to estimate the cost due to liquidating a position on a portfolio of assets.  The last topic involves the construction of optimal technical chart patterns and optimal technical trading rules.  Various Statistical models and data mining techniques will be used in this project.

¡@
Project Duration:
Since 2002

- Back to top- / Contents

¡@
Project Significance:

Some results of the first two topics are used in a contract research project for the Hong Kong Exchanges and Clearing Limited.  We hope that the proposed methodologies can provide better risk management policies for Hong Kong financial markets.  The third topic is related to the concept of market efficiency.  It has a long debate whether the market is efficient.  Recent research did not support market efficiency and claimed that significant profit can be earned by constructing trading rules based on historical prices.  Thus, technical chart patterns are abound but they have no formal definitions.  In this project, we would like to determine the optimal technical chart patterns and provide a formal definition for a chart pattern.  This is of course useful to technical analysts, portfolio managers and market traders.   

¡@
Results Achieved:

Yu, P.L.H., Lam, K. and Ng, S.H. (2003). Automating technical analysis.  Advances in Data Mining and Modeling: Hong Kong 27 - 28 June 2002. (Edited by Wai-Ki Ching & Michael Kwok-Po Ng), World Scientific. 85-96.

 Yu, P.L.H. and So, M.K.P.: Empirical analysis of GARCH models in Value at Risk estimation.  Submitted for publication.

 Working paper in preparation:
 Estimation of liquidation cost 
 Using genetic algorithms to find optimal head-and-shoulders patterns

¡@
Remarks on the Use of High Performance Computing Cluster:

Carried extensive simulation studies, parallel programming and data analysis.

¡@
Email Address:
plhyu@hku.hk
¡@

- Back to top- / Contents