臺灣博碩士論文加值系統

本研究主要目的為：1.比較GARCH、EGARCH模式以及透過Bi-EGARCH純化後帶入類神經網路以及狀態空間模型估算風險值的估算績效。2.探討現貨市場與期貨市場的連動情形，分析不對稱現象以及波動外溢效果(Spillover)。3.運用單變量GARCH系列模型，觀察單一市場波動效果與不對稱效果。研究以台灣發行量加權股價指數與台灣證券交易所股價指數期貨結算價格為研究標的，取樣期間為2003年1月1日至2005年12月31日，資料頻率為日，去除非交易日以及轉化為日報酬率後之有效樣本，共計745筆。研究結果如下：
1. 以GARCH模型進行實證分析，發現波動度受到本身誤差的正向影響，以及本身前期變異數的正向影響；而以EGARCH模型進行實證分析，發現兩個市場中存在著不對稱效果的現象。
2. 透過Bi-EGARCH的分析，無論是台股指數報酬率或是期貨指數報酬率，均受到本身前期報酬率的負向影響，外溢效果則均為正向影響；在波動度方面，僅有標準化衝擊的影響是顯著的，而僅有期貨市場波動度具有不對稱效果。
3. 風險值的估算模型經由Lopez漏損函數(loss function)以及Kupiec概似比檢定後發現，無論在現貨市場或是期貨市場中，類神經網路與狀態空間模型，即採用經由純化後再估算風險值的模式績效最佳，傳統估計方式，在台股指數方面以EGARCH較佳，在台指期貨方面以台指期貨為佳。

研究受到台灣股市受到法令上漲跌幅的限制，可能會導致市場無法充分的反應消息，對本研究結果造成影響；研究資料皆為事後資料，因此所得獲得的預測能力將可能有所限制。對後續研究者建議：1.可以使用狀態空間模型估算出平均數方程式，建立出變異數方程式之後，再進行波動度的估計，可能可以獲得更佳的效果。2.可以將倒傳遞的最陡坡降法的演算模式，轉化為基因演算法，應獲得更好的估算效果。3.可以納入各國的證券市場、外匯市場、或選擇權市場進行研究，可能會有不同的結果產生。

In this research, 1: Compare GARCH, EGARCH model and though purifying Bi-EGARCH will lead to artificial neural network and state space model’s Value at Risk (VaR) calculation. 2: Investigate spot market and future market’s interrelated incident and analyze non-symmetric effect and also spill over effect. 3. Usage of one-way GARCH series model to observe market fluctuate and non-symmetric effect.
The study will be based on Taiwan Stock and Future Market as a research objective and research period from Jan. 1, 2003 to Dec. 31, 2005, and data are base daily. Elimination of days that stock exchange that did not operate, there are 745 data being use in the analysis, and the results are as following:
1. Using the analysis of GARCH model, there’s a significant fluctuation positively. Moreover, using EGARCH model analysis, it shows that in between two market there exist non-systematic effect.
2. Through Bi-EGARCH analysis, no matter it its Taiwan Stock and Future market, will all be effect both positively and negatively with the previous and future return rate.
3. VaR model is resulted from Lopez’s loss function and Kupiec’s likelihood test, and no matter it it’s spot market and future market, or artificial neural network and then calculate the VaR model to obtain the best performance. The traditional calculation, using EGARCH provide Taiwan Stock index a better result. Similarly, Future market will be a better fit when using GARCH model.
This research is limited by Taiwan stock regulations to the restriction of limit move. That has caused the possibility of insufficient market information,
This might be a restriction in forecasting of this research. In recommending future researchers:
1. Using State Space Model to analyze means equation, and build variance equation, and then implement fluctuation analysis. This will obtain a surplus effect in the analysis.
2. Can back-propation neural network’s the gradient steepest descent method, and convert to genetic arithmetic method to obtain a better calculation.
3. This can also be used in all different countries, exchange market, foreign exchange market, and option market in different analysis, and this create different effect.

謝 詞 I
中文論文提要 II
ABSTRACT III
目 錄 IV
表 次 V
圖 次 VI
第一章 緒論 1
第一節 研究背景與動機 1
第二節 研究目的 3
第三節 章節結構 4
第四節 研究流程 5
第二章 文獻探討 6
第一節 風險值的相關概念與用途 6
第二節 風險值的估計方式 9
第三節 國外相關研究 15
第四節 國內相關研究 20
第三章 研究方法 25
第一節 研究設計 25
第二節 資料處理與基本檢定 29
第三節 波動度估計模型 35
第四節 風險值估計模型績效之驗證 47
第四章 實證分析 48
第一節 資料處理與說明 48
第二節 資料基本分析與檢定 50
第三節 Bi-EGARCH的實證研究 52
第四節 標準差純化模式之分析 57
第五節 傳統模式之分析 66
第六節 風險值估算結果 71
第七節 模型績效驗證之實證結果 75
第五章 結論與建議 77
第一節 實證研究結果 77
第二節 管理意涵與研究貢獻 79
第三節 研究限制與後續研究建議 81
參考文獻 82

一、中文部份
1.王倩茵，金控公司市場風險值之研究，台灣大學商學研究所碩士論文，民國92年。
2.沈大白、柯瓊鳳、鄒武哲，風險值衡量模式之探討—以台灣上市公司權並證券為例，東吳經濟學報，第二十二期，民國87年。
3.林保霖，增進蒙地卡羅模擬法評估風險值之績效研究，台北大學企業管理學系碩士論文，民國91年。
4.翁勝彬，認購權證發行人市場風險衡量與評估，東吳大學經濟系碩士論文，民國87年。
5.陳宜玫，風險值估測模型之研究：以台灣股票市場為例，義守大學管理科學研究所碩士論文，民國89年。
6.陳嘉明，應用遺傳演化模糊類神經網路於風險管理之研究，東吳大學經濟系碩士論文，民國93年。
7.張簡彰程，增進模擬法估計風險值之績效研究—以台股票市場為例，義守大管理科學研究所碩士論文，民國90年。
8.萬文隆，兩岸三地連動之研究-狀態空間模型之應用，證券櫃檯月刊，民國91年4月，頁48-65。
9.劉興唐，國際股市連動效應之實證研究，國立中興大學企業管理研究所碩士論文，民國87年。







二、西文部分
1.Akaike, H., Canonical Correlations Analysis of Time Series and the Use of an Information Criterion, Academic Press, New York, 1976.
2.Alexander, C.O., and Leigh, C. T., “On the Covariance Metrices Used in Value at Risk Models”, The Journal of Derivatives, Spring, 1997, pp. 50-62.
3.Backman, Daniel, Choi Jongmoo, Jay, Jeon Bang, Nam, and Kopecky Kenneth, J., “Common Factor in International Stock Prices: Evidence from a Cointegration Study”, International Review of Financial Analysis, Vol. 5, 1996, pp.39-53.
4.Beder Ts., S., “VaR: Seductive but Dangerous”, Financial Analysts Journal, Vol.51, Sep/Oct 1995, pp. 12-24.
5.Berndt, E. K., Hall, B.H., Hall, R.E., and Hausman, J.A., “Estimation and Inference In Nonlinear Structural Model”, Annals of Economic and Social Measurement, Vol.34, 1974, pp.653-665.
6.Bollerslev, T., “Generalized Autoregressive Conditional hetreoscedasticity”, Journal of Econometrics, Vol. 31, 1986, pp. 307-327.
7.Boyle, P., ”Options: A Monte Carlo Approach”, Journal of Financial Economics, Vol.4, 1977, pp.323-338.
8.Brinson, G..P., Hoodand, L.R., Beebower, G..L., “Determinants of Portfolio Performance”, Financial Analysts Journal, Vol.42, Iss. 4, July/August 1986, pp. 33-44.
9.Brooks, C., and Persand, G.., “Volatility forecasting for risk management”, Journal of Forecasting, Vol.22, 2003, pp. 1-22.
10.Cassuto, A.E., “Non-normal Error Patterns: How to Handle Them”, Journal of Business Forecasting Method & System, Vol. 14, 1995, pp. 15-16.
11.Chen, Z.P., and Zhao, C.E., “Is the MV Efficient Portfolio Really that Sensitive to Estimation Errors?”, Asia-Pacific Journal of Operational Research, Vol.19, Iss. 2, Nov 2002, pp. 149-168.
12.Parker, D.B., “Learning-logic, Center for Computational Res. in Economics and Management Sci”, Technical Report TR-47, MIT, April, 1985.
13.Rumelhart, D.E., Hinton, G.E., and Williams, R.J., “Learning Internal Representations by Error Propagation”, Nature, Vol.323, 1986, pp.533-536.
14.Dickey, D.A., and Fuller, W.A., “Distribution of the Estimates for Autoregressive Time Series with Unit Root”, Journal of the American statistical Association, Vol.74, No.366, 1979, pp.427-431.
15.Duffie, D. and Pan, J., “An Overview of Value at Risk”, Journal of Derivative, Vol.4, no.3, 1997, pp.7-49.
16.Engle, R.F., and Granger, C.W.J., “Co-Integration and Error Correction: Representation, Estimation, and Testing”, Econometrics, Vol.55, Iss.2, 1987, pp. 251-277.
17.Engle, R.F., and Granger, C.W.J., Long-run Economic Relationships, Oxford University Press, New York, 1991.
18.Engle, R.F., “Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation”, Econometrics, Vol.50, 1982, pp.987-1007.
19.Gloria, G.R., Lee, T.H., and Mishra, S., “Forecasting Volatility: A reality check based on option pricing, utility function, value-at-risk, and predictive likelihood”, International Journal of Forecasting, Vol.20, 2004, pp. 629-645.
20.Golub, B.W., and Tilman, L.M., “Measuring Yield CurveRisk Using Principal Components Analysis〞, Journal of Portfolio Management, Vol.23, 1997, pp.72-84.
21.Granger, C.W.J., and Newbold, P., “Spurious regression in econometrics”, Journal of Econometrics, Vol. 2, 1974, pp.111-120.
22.Hendricks, Darryll, “Evaluation of Value-at-Risk Models Using Historical Data”, Economic Policy Review,Vol.2, no1, 1996, pp.39-70.
23.Jackson, P., Maude, D.J., and Perraudin, W., “Bank Capital and Value at Risk”, Journal of Derivative, Vol.4, no.3, 1997, pp.73-90.
24.John, Hull, and Alan, White, “Values at Risk When Daily Changes in Market Variables are not Normally Distributed”, Journal of Derivative, Vol.5, no3, 1998, pp.9-19.
25.Jorion, Philippe, “Risk2 : Measuring the risk in value at risk”, Financial Analysts Journal, November/December 1996, pp. 47-56.
26.J.P. Morgan, Risk Metrics Technical Document, 4th,1996.
27.Kalman, R.E., “A New Approach to Linear Filtering and Prediction Problems”, ASME Journal of Basic Engineering, Vol.83, 1960, pp.34-57.
28.Koutmos, G., and Tucker, M., “Temporal Relationship and Dynamic Interactions between Spot and Futures Stock Markets”, Journal of Future Markets, Vol.16, no1, 1996, pp.55-69.
29.Kupiec, P.H., “Techniques for Verifying the Accuracy of Risk Measurement Models,” Journal of Derivatives, Vol.3, Winter, 1995, pp.73-84.
30.Linsmeier, T.J., and Pearson, N.D., “Risk Measurement： An Introduction to Value at Risk”, University of Illinois at Urbana-Champaign, 1996.
31.Ljung, G.M., and Box, G.E.P., “On a measure of Lack of Fit in Time Series Models”, Biometrica, 1978, pp.297-303.
32.Lopez, J.A., “Evaluating the Predictive Accuracy of Volatility Models.” Journal of Forecasting, Vol.20,no 2, 2001, pp.87-109.
33.Markowitz, H., “Portfolio Selection”, Journal of Finance, Vol.6, 1952, pp. 77-91.
34.Marshall, C., and Siegel, M., “Value at Risk：Implementing a Risk Measurement Standard”, Journal of Derivatives, Vol.4, no.3, 1997, pp.91-111.
35.Nelson, D., “Conditional Heteroskedasticity in Asset Returns : A New Approach”, Econometrica, Vol.59, 1991, pp.347-370.
36.Werbos, P., Beyond Regression: New Tool for Prediction and Analysis in the Behavioral Sciences, Ph.D. Thesis, Harvard University, 1974.
37.Said, S., and Dickey, D., “Testing for Unit Roots in Autoregressive-Moving Average Models with Unknown Order”, Biometrics, Vol.71, 1984, pp.599-607.
38.Venkatarman, S., ” Value at Risk for Mixture of Normal Distribution：The Use of Quasi-Bayesian Estimation Techniques”, Economic Perspectives, Vol.23, 1997, pp.2-13.
39.Cun, Y.Le., “Une procedure d’apprentissage pour reseau a seuil assymetrique”, Cognitiva, Vol.85, 1985, pp.599-604.