PROBABILITY AND STATISTICS APPLIED TO THE PRACTICE OF FINANCIAL RISK MANAGEMENT - THE CASE OF MORGAN,J.P. RISKMETRICS(TM)

This work describes applications of probability and statistics in Risk Metrics(TM), J.P. Morgan's methodology for quantifying market risk. Th e methodology implements an analytical approach to financial risk in t rading, arbitrage, and investment based on the statistics of market mo ves in equities, bonds, currencies and commodities. The public unveili ng of RiskMetrics(TM) in October of 1994 attracted widespread interest among regulators, competing financial institutions, investment manage rs, and corporate treasurers, while the available technical documentat ion offers us a unique opportunity for informed statistical research o n the theory and practice of financial risk management. For the purpos e of identifying problems for further research, this discussion focuse s on five applications of statistics in RiskMetrics(TM), which range f rom data analysis of daily returns and locally Gaussian processes to s tochastic volatility models and Ito processes for the term structure o f interest rates. Another important theme of this discussion, however, is devoted to attracting statisticians to the study of financial risk management and developing the foundations for collaborative work with financial economists and practicing risk managers. For this reason, t his is also an expository document that touches several areas of activ e statistical research with applications to problems of risk managemen t.