Modeling the Extreme Risk of Financial Consecutive Losses in Generalized Pareto Distributions

以一般柏拉圖分配函數建立財務連續損失之極端風險模型

陳尚武(Winfred Sun-Wu Chen)

《東亞論壇》 473期 (2011/09) Pp. 73-80

https://doi.org/10.29705/EAR.201109.0007

摘要

本研究探討如何有效建構極端風險值之數學模式。相較於傳統風險值分析（VaR），本研究導入極端值理論，運用一般化柏拉圖分配函數（GPT），推導出DaR是－可行的極端值實證及新的研究方向。

關鍵字

DaR ； 極端值理論 ； 一般化柏拉圖分配

並列摘要

This study intends to explore the modeling of drawdowns variables. Although there are no previous evidences that financial drawdowns are normal, thin-tailed, or thick-tailed distributions, the extreme value theory (EVT) provides flexibilities to model the drawdowns. Throughout our study, we apply limit laws for maxima and uniformity of the convergence to present a comprehensive justification of generalized Pareto distribution (GPD) modeling on drawdowns variables, based on the peak over threshold (POT) framework of EVT. Our justifications provide a theoretical foundation for future studies on the estimation of various promising empirical Drawdown-at-risk (DaR) values.

並列關鍵字

Drawdown ； Drawdown-at-risk ； Extreme value theory ； Generalized Pareto Distribution

參考文獻

Artzner, P., F. Delbaen, J.-M. Eber, D. Heath, and H. Ku. (2003), “Coherent Multi-period Risk Adjusted Values and Bellman's Principle”, Working Paper, ETH Zurich, Switzerland.

Fisher, R. A. and L. H. C. Tippett (1928), “Limiting Forms of the Frequency Distribution of the Largest or Smallest Member of a Sample”, Proc. Cambridge Philos. Soc., 24, 180-190.

(1999).Internal Modeling and CADII.Risk Books.

Riedel, F. (2003), “Dynamic Coherent Risk Measures”, Working Paper, Department of Economics, Stanford University Stanford University.

Adler, R.(ed.),Feldman, R.(ed.),Taqqu, M. S.(ed.)(1998).A Practical Guide to Heavy Tails: Statistical Techniques and Applications.Boston:Birkhauser.