投資組合報酬率風險值與條件風險值估計方法之實證分析

The Empirical Analysis on the Value-at-Risk and Conditional Value-at-Risk Estimation of the Portfolio Return

楊博丞(Bo-Cheng Yang)

指導教授 ： 林志娟

淡江大學/商管學院/統計學系碩士班/碩士(2016年)

https://doi.org/10.6846/TKU.2016.00720

摘要

條件風險值是由風險值衍生而來的風險衡量指標，其意於偵測極端事件的發生，在本研究中，使用歷史模擬法，變異數-共變異數法以及超越門檻值法估計風險值與條件風險值，其中變異數-共變異數法是以指數加權移動平均來估計資產間的變異數-共變異數矩陣;而超越門檻值法是以動差法作參數之估計，利用模擬研究以及實證分析透過回溯測試以及模型精確性分析比較各方法之優劣，在模擬研究方面，分為常態分配模擬以及具有厚尾特性的廣義柏拉圖分配模擬
最後，實證分析結果中，顯著水準 為0.05的情形下，以變異數-共變異數法估計風險值較為精確，而當顯著水準 為0.01時，則以超越門檻值法較其他方法好，在估計條件風險值方面，無論顯著水準 為0.05或0.01，皆以超越門檻值法最為精確，其結果與常態分配模擬結果一致。

關鍵字

風險值 ； 條件風險值 ； 超越門檻值法 ； 回溯測試

並列摘要

Conditional Value-at-Risk is derived from Value-at-Risk and is also a risk measurement instrument used to detect the extreme events. In this research, historical simulation, variance-covariance method incorporating the exponential weighted moving average (EWMA), and peak over threshold method are adopted. Backtesting and modeling accuracy tests are used to compare estimation methods through the simulation and the empirical analysis. In the simulation, the data is assumed to follow the normal distribution and the generalized pareto distribution with the fat tail property.
Regarding to the Value-at-Risk estimation, the variance-covariance method is better than others when equals to 0.05 in empirical or simulation result. On the other hand, the peak over threshold method is better than others when equals to 0.01. Regarding to the Conditional Value-at-Risk estimation, the peak over threshold method outperforms the rest estimation methods whenever equals to 0.05 or 0.01.

並列關鍵字

Value-at-Risk ； Conditional Value-at-Risk ； Peak over threshold method ； Backtesting

參考文獻

林志娟、張慶暉、羅惠瓊、吳旻珊 (2008)，“資產風險值估計方法之探討與比較”，永豐金融季刊，第42期，131~164。

Adrian, T and Shin, H. S. (2012). Procyclical Leverage and Value-at-Risk. Federal Reserve Bank of New York Staff Report 338.

Embrechts, P., Klüppelberg, C., and Mikosch, T. (1997). Modelling Extremal Events for Insurance and Finance. Berlin: Springer Verlag..

Jorion, P. (2006). Value at Risk: The New Benchmark for Managing Financial Risk, third edition, McGraw-Hill.

Mausser, H., and Rosen, D. (1998). Beyond VaR: From Measuring Risk to Managing Risk, Algo Research Quarterly 1/2, 5-20.