臺灣博碩士論文加值系統

一般在金融市場期貨的主要功能之一，就是規避現貨價格變動的風險，而市場在2019年末因遭遇COVID-19疫情的重創，期貨商品是否還能有效地為現貨來規避風險是一個值得研究的議題。本研究旨在探討對於疫情的前後哪些期貨商品還能保有較佳的避險績效，哪項模型更能為避險提供最佳的預測效能。
本研究採用從2018年01月至2021年12月，共4年整的期間以期交所中的期貨商品中挑選出十項上市期間滿足樣本期間的商品，並以日資料做為本研究之分析。透過四種模型其中分別為屬於不對稱模型的GJR-GARCH，廣義正交的GO-GARCH，以及動態條件相關係數DCC-GARCH，上述三種模型皆屬多變數廣義自回歸條件異方差模型，最後以OLS簡單線性回歸模型來作為一個基準的比較點，並由上述主要的三項多變數GARCH模型的績效與之進行比較。
本研究之實證結果主要發現如下: (1)在疫情後期貨商品避險績效大多數具有下降的趨勢；(2)以疫情前為避險提供最佳預測效能的模型為GO-GARCH，而疫情後為避險提供最佳預測效能的模型為GJR-GARCH；(3)無論是疫情前或後單從整體上升下降的績效排名中可以得知美國標準普爾500(SPF),及美國道瓊期貨(UDF)的避險效果不論在任一多變數GARCH 模型下都是具有最佳的避險能力。

Generally, one of the main functions of futures in the financial market is to hedge the risk of changes in spot prices, and the market was hit hard by the epidemic at the end of 2019. Whether futures instruments can still effectively hedge risks for the spot market, so this study aims to explore which futures commodities can still maintain better risk hedgeing performance before and after the epidemic, and which model can provide the best predictive performance for risk hedgeing.
In this study, from January 2018 to December 2021, a total of 4 years is used to select ten futures commodities in the futures exchange and select the commodities whose listing period meets the sample period, and use the daily data as the analysis of this article. Through four models including GJR-GARCH which belongs to bivariate asymmetric model, GO-GARCH which belongs to general benefit orthogonality, and dynamic conditional correlation coefficient DCC-GARCH, the above three models are all multivariate generalized autoregressive conditional heteroscedasticity models. The OLS simple linear regression model was used as a benchmark point of comparison and was compared by the performance of the main three multivariate GARCH models described above.
The empirical results of this study show that: (1) After the epidemic, most futures commodity hedging performance has a downward trend; (2) The model that provides the best predictive performance for hedging before the epidemic is GO-GARCH, while the model that provides the best prediction performance for hedgeing is GJR-GARCH; (3) no matter before or after the epidemic, from the overall performance ranking of rising and falling, we can know that the US Standard & Poor's 500 (SPF) and US Dow Jones Futures (UDF) have the best hedging ability regardless of switching any multivariable GARCH model.

摘要 ... i
Abstract ... ii
誌謝 ... iv
目錄 ... v
表目錄 ... vii
圖目錄 ... viii
第一章、 緒論 ... 1
1.1 研究動機與目的 ... 1
1.2 研究架構 ... 2
第二章、 文獻回顧 ... 3
2.1 避險模型文獻 ... 3
2.2 績效衡量文獻 ... 4
第三章、 研究設計與方法 ... 7
3.1 資料來源與研究期間 ... 7
3.2 研究流程說明 ... 8
3.3 單根檢定 ... 9
3.4 Jarque-Bera Test檢定 ... 9
3.5 ARCH檢定 ... 10
3.6 GJR-GARCH模型 ... 11
3.7 DCC-GARCH模型 ... 12
3.8 GO-GARCH模型 ... 13
3.9 OLS迴歸模型 ... 14
3.10 避險績效衡量 ... 14
第四章、 實證結果與分析 ... 16
4.1 指數趨勢圖 ... 16
4.2 單根檢定 ... 21
4.3 敘述性統計與常態分配檢定結果 ... 22
4.4 ARCH檢定 ... 24
4.5多變量GARCH模型 ... 25
4.5.1GJR-GARCH ... 25
4.5.2DCC-GARCH ... 26
4.5.3GO-GARCH ... 28
4.6參數估計及避險結果 ... 30
4.7避險績效排名比較 ... 35
4.8疫情前後效能最佳GARCH模型 ... 36
第五章、 結論與建議 ... 37
5.1 結論 ... 37
5.2 研究建議 ... 38
5.2.1 對投資人建議 ... 38
5.2.2對期交所建議 ... 38
參考文獻 ... 39
Extended Abstract ... 41

[1] Glosten.L.R, and Jagannathan.R,1993,“On the Relation between the Expected Valae and the Vohfiliw of the Nominal Excess Return on Stocks”,The Journal of Finance,Vol 48,Iss 5 ,p.p.1779-1801.
[2] Engle, R.,2002,“Dynamic Conditional Correlation: A simple class of multivariate generalized autoregressive conditional heteroskedasticity models”,Journal of Business and Economic Statistic,Vol 20,p.p.339-350.
[3] Weide, R.V.D.,2002, “GO-GARCH: A Multivariate Generalized Orthogonal GARCH Model”,Journal of Applied Econometrics, Vol 17, p.p.549-564.
[4] 曾楹庭，2019，“應用GJR-GARCH模型研究美股報酬對台灣上市金控銀行報酬風險之不對稱傳遞”， 嶺東科技大學財務金融學系碩士班碩士論文。
[5] Bollerslev, T.,1990,“Modeling the Coherence in Short-run Nominal Exchange Rates Generalized ARCH.” Review of Economics and Statistics, Vol 70, p.p.498-505.
[6] Engle, R. F. and K. F. Kroner,1995,“Multivariate Simultaneous Generalized ARCH.” Econometric Theory, Vol 11, p.p. 122-150.
[7] Ripple, R.D., and A. A. Moosa.,2007,“Hedging Effectiveness and Futures Contract Maturity: The Case of NYMEX Crude Oil Futures.”,Applied Financial Economics,Vol 17,Iss 5,p.p.683-689.
[8] 凌愛凡，2021，“極端事件衝擊下商品期貨是好的避險工具嗎?”，經濟評論，10卷，3期，頁 105-125。
[9] 洪熾賢，2010，“台股msci指數與國際股價指數動態波動性研究”，國立中山大學財務管理學系碩士班碩士論文
[10] 張巧宜、賴靖宜、莊益源，2013，“期貨最適組合避險模型：新興市場為例”，管理與系統學報，20卷，2期，頁355-383。
[11] 呂冠誼，2017，“台指選擇權避險績效之比較”，國立東華大學財務管理學系碩士班碩士論文。
[12] Yu-Sheng Lai,2019,“Evaluating the hedging performance of multivariate GARCH models.”, Asia Pacific Management Review,Vol 24,p.p.86-95.
[13] Johnson, Leland L.,1960,“The Theory of Hedging and Speculation in Commodity Futures”,The Reviews of Economic Studies, Vol.27, Iss.3, p.p.139-151.
[14] Phillips, C.B. and P. Perron ,1988, “Testing for a Unit Root in Time Series Regression.” Biometrika, Vol 75,p.p.335-346.
[15] Engle, R. F.,1982, “Autoregressive Conditional heteroscedasticity with Estimates of the Variance of United Kingdom Inflation.” Econometrica, Vol. 50, Iss.4, p.p. 987-1007.
[16] Ederington, L.,1979,“The Hedging Performance of the New Hedging Markets＂,The Journal of Finance, Vol.34, p.p. 157-170.
[17] Rabemananjara, R. and J. M. Zakoian,1993, “Threshold ARCH Models and Asymmetries in Volatility”, Journal of Applied Econometrics, Vol. 8, p.p.31-49.