| Summary: | 碩士 === 國立虎尾科技大學 === 財務金融系碩士班 === 105 === Value at Risk (VaR) has become one of the most popular methodologies for financial risk management. Besides, since ETFs have several advantages over index funds, their development and popularity is grown enormously over the past years, especially in Europe. Therefore, in this paper, we use two main approaches, namely, Variance Covariance approach and Historical Simulation approach to estimate VaR of eight European country-specific ETF portfolios. In particular, three different allocative strategies, including equally weighted moving average (EW), EWMA and GARCH models for constructing ETFs portfolios based on these two approaches are respectively developed. Moreover, we also introduce the Sharpe ratio and Modified Sharpe ratio, which apply the semi-variance and variance in turn for referring them to determine the optimally allocative weights for ETFs portfolios. Finally, we conduct two backtesting techniques to examine the validation of estimation with regard to the prediction of future risks. Our empirical results exhibit that the EWMA (Variance-Covariance) model using modified Sharpe ratio as measuring downside risk followed by the GARCH model (Variance-Covariance) can indeed outperform others and beat a benchmark as well.
Key words: Variance-Covariance Approach; Historical Simulation; Semi-Variance; Exchange-Traded Funds; Backtesting
|
|---|
