新冠肺炎期間新台幣匯率對ETF股票報酬之EGARCH不對稱分析

本研究利用E-GARCH-M模型，探討新冠肺炎期間新台幣匯率對ETF股票報酬之不對稱分析，研究期間自2018年01月01日至2021年04月09日，共794筆日資料，資料來源為台灣經濟新報。實證指出利用ADF單根檢定法顯示研究變數報酬皆為穩定時間序列，可以進行後續利用GARCH模型進行報酬及風險波動之研究，符合研究假說1。Granger因果關係的檢定方法，台幣貶值對0050 元大台灣50 (RETF01)、0053 元大電子(RETF04) 0057 富邦摩台(RETF08)有雙向顯著的影響，符合研究假說2。E-GARCH-M模型，均數方程式估計結果看出新台幣前一天報酬為負向顯著影響效果，表示前一天新台幣貶值影響隔天ETF顯著為正報酬；變異數方程式估計結果可看出E-GARCH-M不對稱效果存在。因此，符合研究假說3。

關鍵字

E-GARCH-M模型；新冠肺炎；台幣貶值

並列摘要

This thesis uses the E-GARCH-M model to explore the asymmetric analysis of new Taiwan dollar exchange rate on ETF stock return uring the COVID-19 period. From 01 January 2018 to 09 April 2021, there were 794 daily data from the Taiwan Economic Journal. Empirically, this study uses ADF unit root test and it shows that all the return variables are stable time series, hence we can use GARCH model to explore the volatility of return and risk, conforms to the hypothesis 1. Granger causality results show that the depreciation of the new Taiwan dollar and 0050 Yuanta Taiwan50 (RETF01)、0053 Yuanta Electronics (RETF04)、0057 Fubang Motai(RETF08) influence each other significantly, conforms to the hypothesis 2. E-GARCH-M model, the mean equation estimation results show that the previous day's return of the new Taiwan dollar has a significant negative effect, it means that the depreciation of the new Taiwan dollar the day before affected the next day ETF return, which is significantly positive, the variation equation results show that exists E-GARCH-M asymmetry effect, in line with the hypothesis 3.