RiskMed: Collaborative Intelligence for Liquidity Risk Mediation
- J Leon ZHAO (Principal Investigator / Project Coordinator)Department of Information Systems
- Akhil KUMAR (Co-Investigator)
- Chak Sham Michael WONG (Co-Investigator)Department of Economics and Finance
- Daniel Dajun ZENG (Co-Investigator)
DescriptionWe propose a collaborative intelligence approach to liquidity risk mediation in banks by adopting, extending, and integrating a number of human and machine intelligence techniques. While liquidity management systems have been used in banks for many years, those systems are inadequate with effectively managing liquidity risks as financial events can be very complex, volatile, and difficult to track. As such, banks in Hong Kong and around the world need special solutions to liquidity risk mediation. This project strives to fill this gap between business needs and financial technologies by developing a collaborative process of human and machine intelligence (or simply collaborative intelligence) consisting of techniques from complex event processing, predictive analytics, and collaborative decision making.Liquidity risk mediation, also known as liquidity surveillance and mitigation, is a complex process because cash flow of banks is very uncertain and can change rapidly in real time. The financial solvency of a bank depends on the behavior of the bank customers, the financial positions of various units of the bank, and external market environments. As a result, many financial indicators need to be monitored continuously to estimate the liquidity position of the bank. When certain financial events occur, the bank must react to risky situations effectively by mobilizing appropriate personnel usually distributed around the globe, particularly for large multinational banks with hundreds of branches. In the aftermath of the 2008 financial tsunami, banks are terribly stressed to monitor liquidity indicators and maintain financial health of the bank. Therefore, effective monitoring techniques, decision support methods, and collaboration tools specially designed for liquidity risk mediation are in great demand. There is no accident that the Research Grants Council of Hong Kong decided in June 2009 to make financial subjects as one of the four areas of theme-based research.The RiskMed project will develop collaborative intelligence for risk mediation in three stages:(Analysis) Analyze the feasibility of adopting, in liquidity risk analysis, existing business intelligence techniques found in complex event processing, predictive analytics, and collaborative decision making. Create a portfolio of risk analysis techniques suitable for liquidity risk mediation in banks.(Design) Create a process-driven framework and testbed consisting of collaborative and intelligent risk mediation techniques that are either extended from those studied in Stage 1 or completely new. Assess the effectiveness of the risk mediation techniques under various business conditions.(Validation) Apply and evaluate, via case studies, surveys, and experiments, the RiskMed approach to liquidity risk mediation in a variety of banks ranging from local banks to multinational banks.The research outcome of this project will include novel computational algorithms and collaborative intelligence solutions for liquidity risk mediation. Our research will be based on business requirements from various banks in Hong Kong and the greater China region. One of the investigators, Dr. Michael Wong, has consulted over 20 banks on financial risk management and has been interacting with Hong Kong Monitory Authority on issues of financial engineering and risk analysis. The RiskMed team possesses the interdisciplinary expertise in finance, business intelligence, decision support, and process management needed to tackle challenges in managing systematic financial risks. We hope that our research efforts will add to the intellectual knowledge on controlling the severity of financial risks in the information economy, thus helping avoid certain financial crises in the future.
|Effective start/end date||1/10/10 → 18/03/15|