銀行、證券商、投信投顧等金融相關機構的經營模式與市場風險有高度相關性、風險的不確定性將會影響公司的現金流，比如說銀行調整資本適足率、證券商在違約交割時提列的賠償準備金、投信投顧機構提列營業保證金等都會受到市場風險的影響，若準備金提列不足，一旦市場出現黑天鵝，整體投資市場系統性走跌，便會瓦解金融秩序並造成嚴重損失，提列過多會讓機構營運閒置資金過多無效率，換句話說若潛在的風險沒有被正確地估計，可能導致金融機構高估(低估)他們的市場風險，從而維持過高(低)的資本水平，其結果便會財政資源的低效率配置，最終可能直接或間接地導致經濟衰退。由此可知精準的預測風險並提列準備金是金融機構營運的一個重要面向，當市場風險高時相關準備金要求就會相應提高，因此許多金融機構將風險視為重點事項，著重於風險的辨識與預測上。
本研究使用了2019年Patton and Ziegel 提出的GARCH架構的VaR與ES風險模型，先使用歷史報酬資料訓練風險模型參數、再使用訓練完成的模型去預測未來一天的風險，風險指標選擇風險值Value at Risk（VaR）以及期望損失Expected Shortfall(ES)，如果能使用預測的VaR以及ES精準識別未來的風險，則可以降低對於提撥準備金的不確定性，增強金融機構在營運上的穩健度。

The business models of financial-related institutions such as banks, securities firms, investment credit investment advisors are highly correlated with market risks. The uncertainty of risks will affect the company’s cash flow. The compensation reserves listed at the time and the margins of investment and investment advisory institutions will be affected by market risks. If the reserves are insufficient, once a black swan appears in the market and the overall market crash systematically, the financial order will be disrupted. And then cause serious losses, entries will make the organization’s operating idle funds too much and inefficient. In other words, if the potential risks are not correctly estimated, it may cause financial institutions to overestimate (underestimate) their market risks and maintain them too high (Low) capital level, the result will be the inefficient allocation of financial resources, which may eventually directly or indirectly lead to economic recession. It can be seen that accurate prediction of risks and provision of reserves are an important aspect of the operation of financial institutions. When market risks are high, relevant reserve requirements will increase accordingly. Therefore, many financial institutions regard risk as a priority and focus on risk identification and prediction.
This research uses the VaR and ES risk model under the GARCH process proposed by Patton and Ziegel in 2019. First, the historical return data is used to train the risk model parameters, and then the trained model is used to predict the risk of the next day, and the risk indicator in this paper are Value at Risk (VaR) and Expected Shortfall (ES). If the predicted VaR and ES can be used to accurately assess future risks, the uncertainty about the provision of reserves can be reduced, and the operation of financial institutions can enhance the robustness.

論文審定書 i
Acknowledgements ii
摘要 iii
Abstract iv
目錄 v
圖目錄 vi
表目錄 vii
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究方法與目標 2
1.3 研究架構 4
第二章 文獻回顧 5
2.1 風險值(Value at Risk, VaR) 5
2.2 期望損失(Expected Shortfall, ES) 7
2.3 GARCH模型 9
2.4 FZ損失函數(Fissler and Ziegel loss function) 10
2.5 文獻總結 15
第三章 研究方法 16
3.1 理論模型 16
3.2 估計方法 17
第四章 實證研究 21
4.1 訓練資料來源與處理 21
4.2 模型參數估計結果與預測 23
4.3 評估預測結果 26
第五章 結論與建議 28
5.1 結論 28
5.2 後續研究建議 29
參考文獻 30

1. Abad, P., Benito, S., & López, C. (2014). A comprehensive review of Value at Risk methodologies. The Spanish Review of Financial Economics, 12(1), 15-32.
2. Brie, L., Genest, B., & Arsac, M. (2018). Back-Testing of Expected Shortfall: Main Challenges and Methodologies. Available at SSRN 3174711.
3. Brehmer, J. (2017). Elicitability and its application in risk management. arXiv preprint arXiv:1707.09604.
4. Creal, D., Koopman, S. J., & Lucas, A. (2013). Generalized autoregressive score models with applications. Journal of Applied Econometrics, 28(5), 777-795.
5. Catania, L., & Luati, A. (2021). Quasi Maximum Likelihood Estimation of Value at Risk and Expected Shortfall. Available at SSRN.
6. Dionne, G., Duchesne, P., & Pacurar, M. (2009). Intraday Value at Risk (IVaR) using tick-by-tick data with application to the Toronto Stock Exchange. Journal of Empirical Finance, 16(5), 777-792.
7. Dendoncker, V., & Lebègue, A. Comparative backtesting of the Expected Shortfall.
8. Engle, R. F., & Manganelli, S. (2004). CAViaR: Conditional autoregressive value at risk by regression quantiles. Journal of business & economic statistics, 22(4), 367-381.
9. Francq, C., & Zakoïan, J. M. (2015). Risk-parameter estimation in volatility models. Journal of Econometrics, 184(1), 158-173.
10. Fissler, T., & Ziegel, J. F. (2016). Higher order elicitability and Osband’s principle. The Annals of Statistics, 44(4), 1680-1707.
11. Fissler, T., Frongillo, R., Hlavinová, J., & Rudloff, B. (2021). Forecast evaluation of quantiles, prediction intervals, and other set-valued functionals. Electronic Journal of Statistics, 15(1), 1034-1084.
12. Gebizlioglu, O. L., Şenoğlu, B., & Kantar, Y. M. (2011). Comparison of certain value-at-risk estimation methods for the two-parameter Weibull loss distribution. Journal of Computational and Applied Mathematics, 235(11), 3304-3314.
13. Gneiting, T. (2011). Making and evaluating point forecasts. Journal of the American Statistical Association, 106(494), 746-762.
14. Komunjer, I. (2005). Quasi-maximum likelihood estimation for conditional quantiles. Journal of Econometrics, 128(1), 137-164.
15. Lazar, E., & Xue, X. (2020). Forecasting risk measures using intraday data in a generalized autoregressive score framework. International Journal of Forecasting, 36(3), 1057-1072.
16. Nolde, N., & Ziegel, J. F. (2017). Elicitability and backtesting: Perspectives for banking regulation. The annals of applied statistics, 11(4), 1833-1874.
17. Omari, C. O. (2017). A comparative performance of conventional methods for estimating market risk using value at risk.
18. Patton, A. J., Ziegel, J. F., & Chen, R. (2019). Dynamic semiparametric models for expected shortfall (and value-at-risk). Journal of econometrics, 211(2), 388-413.
19. Storti, G., & Wang, C. (2022). Nonparametric expected shortfall forecasting incorporating weighted quantiles. International Journal of Forecasting, 38(1), 224-239.
20. Taylor, J. W. (2019). Forecasting value at risk and expected shortfall using a semiparametric approach based on the asymmetric Laplace distribution. Journal of Business & Economic Statistics, 37(1), 121-133.
21. Ziegel, J. F., Krüger, F., Jordan, A., & Fasciati, F. (2017). Murphy diagrams: Forecast evaluation of expected shortfall. arXiv preprint arXiv:1705.04537.