Sumario: | Market risk valuation is very important not only for financial theoreticians but for the applied job. When using daily time series on exchange rates, the use of RiskMetricsTM strategy, in conjunction with GARCH models, is common place. Many authors, in several contexts, have pointed out the inconvenience in using GARCH models. In this paper the use of Stochastic Volatility (SV) models in risk valuation is considered. We describe the main advantages and disadvantages of SV models against GARCH models. We present results of market risk valuation using SV for a portfolio that contains daily returns of Euro/Dollar, Yen/Dollar, UK$/Dollar and exchange rates. Our results suggest that the SV model produce more reliable estimations.
|