| Summary: | 碩士 === 國立中興大學 === 統計學研究所 === 107 === Exponentially weighted moving average (EWMA) is one of the most used commonly models in the field of finance. One of the major advantages of EWMA is that it gives more weight to the recent returns while calculating the returns. More and more papers suggest that generally weighted moving average (GWMA) is more extensively applicable because of its sensitivity over micro measurements.
In this study I examine the effectiveness of cross hedging the USD/TWD exchange rate changes with USD/JPY. We take advantage of Historical model, EWMA, GWMA and Generalized Auto-Regressive Conditional Heteroscedasticity (GARCH model) to estimate optimal hedge ratio and compare these methods in different rolling data in this thesis. Moreover, we find out the optimal parameters in GWMA and EWMA by using numerical method.
The empirical results suggest that GWMA model is a powerful model for hedging and should be recognized as a useful hedging strategy.
|
|---|
