從2018年到現在最引人注目的全球議題就是中美貿易戰，全世界各國都持續在關注此議題。當然，中美貿易戰開戰以來衝擊無數個產業，全球的股票指數也受到一定的影響。美國與中國分別身為全世界第一及第二大經濟體，台灣的經濟主要也仰賴美國與中國，而這場貿易戰對台灣來說也會受到一定的影響。
    本研究主要探討台灣加權股價指數、中國上證A股指數、中國深證A股指數、美國紐約道瓊工業平均指數、美國紐約S&P500股價指數及美國那斯達克100指數在極端事件下的風險值，分別以GARCH、E-GARCH與GJR-GARCH模型與加入極端值理論的混和模型來進行比較，並根據上述模型計算出的結果進行回測，檢視何種模型具有較佳的結果。
    實證結果顯示，使用GARCH風險值模型與使用GARCH結合極端值理論模型計算出的風險值進行中美貿易戰前與後的比較，我們可以發現所有的模型在中美貿易戰後的風險值都較中美貿易戰前還來的高，這代表著中美貿易戰的開始也影響著這六個標的。最後我們透過概似比檢定比較不同模型間的評估績效，從結果我們可以發現使用加入極端值理論的混和模型計算的風險值有比GARCH、E-GARCH與GJR-GARCH模型計算的風險值來的更準確。

The most striking global issue from 2018 to the present is the U.S.-China trade war, and countries around the world continue to pay attention to this issue. Of course, since the start of the U.S.-China trade war, it has impacted countless industries, and global stock indexes have also been affected to a certain extent. The United States and China are respectively the first and second largest economies in the world, Taiwan’s economy mainly depends on the United States and China, and this trade war will also be affected to a certain extent for Taiwan.
	This research mainly discusses the Value at Risk of Taiwan Capitalization Weighted Stock Index, China Shanghai A-share index, China Shenzhen A-share index, Dow Jones Industrial Average, Standard & Poor’s 500 and NASDAQ 100 Index under extreme events. The GARCH, E-GARCH and GJR-GARCH models are compared with the hybrid model with Extreme Value Theory, and the backtesting is performed based on the results calculated by the above model to check which model has th better results.
	The empirical results show that using GARCH model and using GARCH combined with the Extreme Value Theory model to calculated Value at Risk to compared before and after the U.S.-China trade war. We can find that the Value at Risk of all models after the U.S.-China trade war is higher than before the U.S.-China trade war, which means that the beginning of the U.S.-China trade war also affects these six targets. Finally, we through backtesting to compare the evaluation performance between different model. From the results, we can find that calculated the Value at Risk by the hybrid model with Extreme Value Theory is more accurate than calculated the Value at Risk by the GARCH, E-GARCH and GJR-GARCH models.

目 錄
表 目 錄	VI
圖 目 錄	VII
第一章 緒論	1
第一節 研究背景與動機	1
第二節 研究目的	4
第三節 研究架構與流程	5
第二章 理論與相關文獻	7
第一節 風險值的相關文獻	7
第二節 GARCH模型的相關文獻	8
第三節 極值理論的相關文獻	11
第四節 動態極值理論的相關文獻	13
第三章 研究方法	15
第一節 研究資料定義與來源	15
第二節 GARCH、E-GARCH與GJR-GARCH模型定義與概念	16
第三節 GARCH-EVT、E-GARCH-EVT與GJR-GARCH-EVT模型定義與概念	20
第四節 績效評估模型	25
第四章 實證結果與分析	28
第一節 敘述統計量分析	28
第二節 GARCH、E-GARCH與GJR-GARCH模型風險值估計	36
第三節 GARCH-EVT、E-GARCH-EVT與GJR-GARCH-EVT模型風險值估計	39
第四節 績效評估結果	43
第五章 結論與建議	61
第一節 結論	61
第二節 建議	62
參考文獻	64
 
表 目 錄
表 1 研究對象的基本資料	15
表 2 中美貿易戰前股票指數報酬率之基本敘述統計量	28
表 3 中美貿易戰後股票指數報酬率之基本敘述統計量	32
表 4 六個標的在中美貿易戰前後的GARCH風險值	37
表 5 六個標的在中美貿易戰前後的E-GARCH風險值	38
表 6 六個標的在中美貿易戰前後的GJR-GARCH風險值	38
表 7 六個標的在中美貿易戰前後的門檻值設定	40
表 8 六個標的在中美貿易戰前後的GARCH-EVT風險值	41
表 9 六個標的在中美貿易戰前後的E-GARCH-EVT風險值	42
表 10 六個標的在中美貿易戰前後的GJR-GARCH-EVT風險值	43
表 11 中美貿易戰前後道瓊之穿透率與穿透次數	44
表 12 中美貿易戰前後S&P500之穿透率與穿透次數	45
表 13 中美貿易戰前後那斯達克100之穿透率與穿透次數	46
表 14 中美貿易戰前後台灣加權之穿透率與穿透次數	47
表 15 中美貿易戰前後上證A股之穿透率與穿透次數	48
表 16 中美貿易戰前後深證A股之穿透率與穿透次數	49
表 17 中美貿易戰前後道瓊之LR 和 LRcc test	51
表 18 中美貿易戰前後S&P500之LR和 LRcc test	52
表 19 中美貿易戰前後那斯達克100之LR和 LRcc test	53
表 20 中美貿易戰前後台灣加權之LR和 LRcc test	54
表 21 中美貿易戰前後上證A股之LR和 LRcc test	55
表 22 中美貿易戰前後深證A股之LR和 LRcc test	56
表 23 穿透率及概似比回溯測試結果之綜合比較	58
表 24 中美貿易戰後各標的使用混和模型結果之綜合比較	60

圖 目 錄
圖 1 研究流程圖	6
圖 2 中美貿易戰前之道瓊、S&P500、那斯達克100每日報酬率時間序列圖	29
圖 3 中美貿易戰前之台灣加權每日報酬率時間序列圖	29
圖 4 中美貿易戰前之上證A股、深證A股每日報酬率時間序列圖	29
圖 5 中美貿易戰前之六個研究對象報酬率之常態檢定各圖形	31
圖 6 中美貿易戰後之道瓊、S&P500、那斯達克100每日報酬率時間序列圖	33
圖 7 中美貿易戰後之台灣加權每日報酬率時間序列圖	33
圖 8 中美貿易戰後之上證A股、深證A股每日報酬率時間序列圖	33
圖 9 中美貿易戰後之六個研究對象報酬率之常態檢定各圖形	35

參考文獻
一、	中文文獻
1.	楊愛軍、張杰(2011)，「基於極值理論的滬市VaR和ES風險價值度量」，山東經濟，第27卷，第5期，頁100-105。
2.	翟永會、趙飛(2019)，「比特幣市場風險度量-基於GARCH-t模型與GEV模型的比較研究」，金融，第9卷，第1期，頁19-27。
3.	蔣晶晶、葉斌、馬曉明(2015) ，「基於GARCH-EVT-VaR模型的碳市場風險計量實證研究」，北京大學學報，第51卷，第3期，頁511-517。
二、	英文文獻
1.	Altun, E. and H. Tatlidil.(2016), “Forecasting Value-at-Risk with Novel Wavelet Based Garch-EVT Model,” Gazi University Journal of Science, Vol.29, No.3, pp.599-614.
2.	Aliyev, F., R. Ajayi and N. Gasim(2019), “Modelling Market Volatility with Univariate GARCH Models: Evidence from Nasdaq 100,” Retrieved from https://www.preprints.org/manuscript/201909.0280/v1 .
3.	Bollerslev,, T. (1986), “Generalized Autoregressive Conditional Heteroskedasticity,” Journal of Econometrics, Elsevier, Vol.31, No.3,  pp. 307-327.
4.	Balkema, A. and Haan L. D.(1994), “Residual Life Time at Great Age,” Annals of Probability, Vol.2, No.5, pp.792-804.
5.	Christoffersen, P. F. (1998), “Evaluating Interval Forecasts,” International Economic Review, Vol.39, No.4, pp.39-841.
6.	Cerovic, J., M. Lipovina-Bozovic, S . Vujosevic.(2015), ”A Comparative Analysis of Value at Risk Measurement on Emerging Stock Markets : Case of Montenegro,” Business Systems Research, Vol. 6, No.1, pp.36-55.
7.	Cheteni, P. (2016). “Stock Market Volatility Using GARCH Models: Evidence from South Africa and China Stock Markets,” MPRA Paper 77355, University Library of Munich, Germany.
8.	Danielsson, J. and C. G. de Vries (1997), “Beyond the Sample: Extreme Quantile and Probability Estimation,” Mimeo, Tinbergen Institute Rotterdam. 
9.	Dowd (1998), “Beyond Value at Risk”, John Wiley and Sons.Erasmus University Rotterdam.
10.	Diebold, F. X., T. Schuermann and J. D. Stroughair(1999), “Pitfalls and Opportunities in the Use of Extreme Value Theory in Risk Management,” Decision Technologies for Computational Finance, Vol.2, pp.3-12.
11.	Delis, P. and J. H. Plikas (2017), “Comparing Methods to Forecasting Var: Fhs Versus Evt Approaches”, Journal of Economics and Finance, Vol. 8, Issue 2, pp.63-68
12.	Engle, R. F. (1982), “Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation,” Econometrica, Vol. 50, No.4, pp.987-1008.
13.	Etuk, I. E., Y. Musa and S. U. Gulumbe.(2019), ”Estimating and Predicting Value at Risk in Selected Bank of Nigeria Stock Market,” International Journal of Statistics and Applications, Vol.9, No.4,  pp.117-121.
14.	Glosten, L. R., R. Jagannathan and D. E. Runkle (1993),” On the Relation between the Expected Value and the Volatility of the Nominal Excess Return on Stocks,” The Journal of Finance, Vol. 48, No. 5, pp. 1779-1801.
15.	Gencay, R. and F..Selcuk.(2004), “Extreme Value Theory and Value-at-Risk : Relative Performance in Emerging Market,” Inernational Journal of Forecasting, Vol.20, Issue 2, pp.287-303.
16.	Gutiérrez, R. D. and R. J.. Salgado (2019), ” Conditional Extreme Values Theory and Tail-related Risk Measures: Evidence from Latin American Stock Markets,” International Journal of Economics and Financial Issues, Vol. 9, No.3, pp.127-141.
17.	Jiang, W.(2012), “Modeling and Predicting of Different Stock Markets with GARCH Model”, Department of Statistics Uppsala University.
18.	Kupiec, P. H. (1995), “Techniques for Verifying the Accuracy of Risk Measurement Models,” The Journal of Derivatives Winter, Vol.3, No,2, pp.73-84. 
19.	Longin, F. M. (1996), ”The Asymptotic Distribution of Extreme Stock Market Returns,” The Journal of Business, Vol. 69, issue 3, pp.383-408.
20.	Li, L. (2017) "A Comparative Study of GARCH and EVT Model in Modeling Value-at-Risk," Journal of Applied Business and Economics, Vol. 19, Iss.7,  pp. 27-48.
21.	McNeil, A. J. (1999), “Extreme Value Theory for Risk Managers”, Internal Modeling and CAD II, RISK Books, pp.93-113.
22.	McNeil, A. J. and R. Frey.(2000),” Estimation of Tail-Related Risk Measures for Heteroscedastic Financial Time Series : an Extreme Value Approach,” Journal of Empirical Finance, Vol.7, pp.271-399.
23.	Nelson, D. B. (1991),”Conditional Heteroskedasticity in Asset Returns : A New Approach,” Econometrica, Vol. 59, issue 2, pp.347-70.
24.	Omari, C. O., P. N. Mwita and A. G. Waititu (2017), ”Using Conditional Extreme Value Theory to Estimate Value-at-Risk for Daily Currency Exchange Rates,” Journal of Mathematical Finance, Vol.7, No.4, pp.846-870.
25.	Pickands, J. (1995), “Statistical Inference Using Extreme Order Statistics,” The Annals of Statistics, No.3, pp.119-131.
26.	Paul, S. and P. Sharma (2017), ” Improved VaR Forecasts Using Extreme Value Theory with the Realized GARCH Model,” Studies in Economics and Finance, Vol. 34, issue 2, pp.238-259.
27.	Pratiwi, N., C. Iswahyudi and R. I. Safitri (2019)”Generalized Extreme Value Distribution for Value at Risk Analysis on Gold Price,” Journal of Physics : Conference Series, Vol. 1217, pp.1-8.
28.	Stoyanov, S., L. Loh and F. J. Fabozzi (2017), ”How Fat are the Tails of Equity Market Indices?” International Journal of Finance & Economics, Vol.22, Issue 3, pp.181-200.
29.	Velasco, AAF and DKP Lapuz (2018)” Extreme Value Modelling for Measuring Financial Risk with Application to Selected Philippine Stocks,” Journal of Applied and Computational Mathematics, Vol.7, Issue 3, pp.1-13. 
30.	Williams, R., J.D. van Heerden and W.J. Conradie (2018)” Value at Risk and Extreme Value Theory : Application to the Johannesburg Securities Exchange,” Econometrica, Vol.42, No.1, pp.87-114.