The paper uses multivariate autoregressive conditional heteroscedasticity models to investigate the effect of dollar/sterling exchange rate fluctuations on coffee and cocoa futures prices on the London LIFFE and the New York CSCE. For both commodities and in both markets, the exchange rate...

This paper compares different models for volatility forecasts with respect to the value at risk performance (VaR) for daily stock index returns. The VaR measures the potential loss of a portfolio for the next period at a given significance level. We will focus on the question if the choice of...

We illustrate in this paper the use of multivariate time series forecasts for portfolio construction and address the following research questions: First, how can forecasts of time series models be used for portfolio weight selection? Second, what kind of time series information improves...