貨幣型ETF與經濟因素之關聯性：應用ARIMAX-GARCH模型分析

The Relationehip between Currency ETF and Economic Factors： The Application of ARIMAX-GARCH Model

黃國銘(KUO-MING HUANG)

指導教授 ： 陳若暉

中原大學/商學院/企業管理研究所/碩士(2016年)

https://doi.org/10.6840/cycu201600591

摘要

近年來衍生性金融商品吸引許多企業及民眾投資的焦點，其中指數型基金（ETF）為穩定成長並可獲利之金融商品。2005年首支貨幣型ETF(歐元貨幣基金)問世，投資報酬備受國際金融市場肯定，陸續推出主要國際流通貨幣型ETF。故本研究以主要貨幣流通國家為主軸，選取澳洲、英國、加拿大、瑞士、日本、美國、紐西蘭等國家所發行貨幣型ETF為研究對象，研究期間於2007年1月至2015年12月31日止之每日報酬率資料為探討樣本，運用ARIMAX-GARCH模型分析，驗證及預測貨幣型ETF之報酬率與總體經濟因素之跨期性影響效應。研究結果如下：
一、當恐慌性指標(VIX)上升時，當期報酬率均為顯著混合影響。表示投資人在面對前期報酬率(Return)、通貨膨脹指標(CRB)、股票指數(Index)、短期利率(Rate)等變數之係數波動而影響到投資方向。故當期報酬率與投資人之投資方向相互影響，最終影響下期或後幾期報酬率。
二、當期報酬率趨勢向上時，前期報酬率(Return)呈現負向相關，通貨膨脹指標(CRB)、利率(Rate)則呈現正向相關。
三、報酬率預測分析結果，發現加拿大貨幣型ETF的MAE、RMSE為最小值，表示預測報酬率最佳，適合投資標的。
上述證實結果，利率(Rate)、股價指數(Index)、通貨膨脹指標(CRB)、恐慌性指標(VIX)等經濟因素確實與貨幣型ETF具跨期性影響效應。各投資機構與投資若可參考關聯性指標，選擇與判斷符合投資偏好的貨幣型ETF投資，以有效掌握報酬率及降低投資風險。

關鍵字

貨幣型ETF、報酬率預測、ARIMAX-GARCH 模型

並列摘要

Recently, the number of financial derivatives have been created to attract the business and public to invest. Among them, the Exchanged-Traded Fund (ETF) is one of the most stable financial derivatives and profitable. In 2005, the first currency ETF (euro IMF) has been established, and it’s performance was affirmed by the international financial markets. The major currencies were used to issue a series of currency ETFs. This study takes the major currencies –such as Australia, United Kingdom, Canada, Switzerland, Japan, USA, New Zealand and other countries as study subjects. The sample data of daily return is from the beginning of January 2007 to the end of December 2015 as the objects. This paper uses the ARIMAX-GARCH model to analze, test and forecast the dynamic relationship of currency ETFs returns and macroeconomic factors. The results are as follows:
1.When volatility index (VIX) rises, the current returns have significantly mixed effect. This shows that the investors may change investment strategies which affected by the previous returns, CRB, stock index and short term interest rate. Therefore, the current returns interacted with investors’s investment strategies may ultimately affect the next-term or few-terms of return.
2.When the current retrrns have upward trend and the previous returns have a negative impact , while CRB, and short-term interest rate showed positice correlation.
3.The results of forecast for the rerurns, reveal that Canada currency ETF's MAE and RMSE have the minimum value, which means the best predicting performance for the returns and most suitable for investment.
The above results indicate the economics factors such as short-term interest rate, stock index, CRB, and VIX having dynamic effects with the currency ETFs. Institution and the investors can take indicators as references, to choose the favor currency ETF based on their investing preference in order to effectively control the returns and avoid the investment risk.

並列關鍵字

Currency ETF, Rate of Return Prediction, ARIMA-GARCH Model

參考文獻

常皓蓉(2013)，投資情緒對股價報酬持續性的影響—縱橫平滑轉換自我迴歸模型之應用，中原大學國際經營與貿易研究所論文。

楊仁吉 (2008)，亞洲單一貨幣化與短期資本流動之關聯性研究，中原大學企業管理研究所論文。

洪珮菁(2012)，兩種商品市場指標與股市、匯率以及經濟指標關聯性之研究—以美國與金磚四國為例，中原大學企業管理研究所論文。

蔡易成(2015)，債券型ETF之長期記憶及預測，中原大學企業管理研究所論文。

黃麒元(2004)，全球主要指數股票型基金之外溢效果與績效之研究，中原大學企業管理研究所論文。