Paper Infomation
Bayesian Estimation of Value-at-risk Based on Gray Peaks over Threshold
Full Text(PDF, 487KB)
Author: Ruiqing Wang
Abstract: A two-stage model for estimating value-at-risk based on grey system and extreme value theory is proposed. Firstly, in order to capture the dependencies, seasonalities and volatility-clustering, an GM(1,2) model is used to filter electricity price series. In this way, an approximately independently and identically distributed residual series with better statistical properties is acquired. Then peaks over threshold is adopted to explicitly model the tails of the residuals of GM(1,2) model, and accurate estimates of electricity market value-at-risk can be produced. For conquering the difficulty lacking for sample data over threshold, Bayesian estimation based on Markov Chain Monte Carlo simulation is used to estimate the parameters of peaks over threshold model. The empirical analysis shows that the proposed model can be rapidly reflect the most recent and relevant changes of electricity prices and can produce accurate forecasts of value-at-risk at all confidence levels, and the computational cost is far less than the existing two-stage value-at-risk estimating models, further improving the ability of risk management for electricity market participants.
Keywords: Value-at-risk; Grey System Theory; Extreme Value Theory; GM(1,2); Peaks Over Thresholds; Bayesian Estimation
References:
[1] LIU Baohua, WANG Dongrong, SHU Anjie. Reconsideration on the energy crisis in California [J]. Automation of Electric Power System, 2007, 31(7): 1-5
[2] ZHANG Fuqiang, ZHOU Hao. Financial risk analysis in electricity market by analytical approach [J]. Proceedings of the CSU-EPSA, 2004, 16(3): 23-26
[3] LI Ming, ZHANG Qiang, SI Yung. Evaluating short-term financial risk based on Delta model for hydropower plants [J]. Power System Protection and Control, 2010, 38(14): 12-15, 27
[4] ZHONG Bo, LI Huamin. Financial risk analysis of electricity market by a copula based approach [J]. Journal of Shanxi Normal University (Natural Science Edition), 2007, 21(2): 15-19
[5] LIAO Jing, JIANG Hui, PENG Jian-chun, et al. Risk assessment on bidding strategy of power generation companies based on VaR and CVaR method [J]. Relay, 2007, 35(11): 30-34
[6] HUANG Renhui, ZHANG Ji, ZHANG Lizi, LI Zhaoli. Price risk forewarning of electricity market based on GARCH and VaR theory [J]. Proceedings of the CSEE, 2009, 29(19): 85-91
[7] Zhang X B. A Long-term price risk early-warning model of electricity company based on EGARCH and VAR [C]. 2010 International Conference on Advances in Energy Engineering, 19-20 Jun. 2010, Beijing, China: 295-298
[8] Hartz C, Mittnik S, Paolella M. Accurate value-at-risk forecasting based on the normal-GARCH model [J]. Computational Statistics & Data Analysis, 2006, 51(4): 2295-2312
[9] WANG Mianbin, TAN Zhongfu, ZHAN Rong. Purchase power portfolio model and an empirical analysis based on risk measure with fractal value-at-risk [J]. Proceedings of the CSU-EPSA, 2009, 21(6): 11-16
[10] WANG Ruiqing, WANG Hongfu, WANG Xian, LI Yuzeng．Calculating value-at-risk of electricity market considering the time-varying features of distribution’s parameters [J]. Power System Protection and Control, 2012, 40(24): 46-52
[11] Wang Ruiqing, Wang Fuxiong, GUO Xiaojiao. Research on price risk of electricity market based on GARCH model [J]. Journal of Hainan Normal University (Natural Science), 2012, 25(1): 36-40
[12] De Rozario R. Estimating value at risk for the electricity market using a technique from extreme value theory, University of New South Wales, 15 Nov. 2011. [http://www.sal.tkk.fi/publications/pdf-files/eleh07.pdf]
[13] Bystrom H N E. Extreme value theory and extremely large electricity price changes [J]. International Review of Economics and Finance, 2005, 14 (1): 41-55
[14] Chan K F, Gray P. Using extreme value theory to measure value-at-risk for daily electricity spot prices [J]. International Journal of Forecasting, 2006, 22(2): 283-300
[15] Gong X S, Luo X, Wu J J. Electricity auction market risk analysis based on EGARCH-EVT-CVaR model [C]. Proceedings of IEEE International Conference on Industrial Technology, Gippsland, Victoria, Australia, 10-13 Feb. 2009: 1-5
[16] WANG Ruiqing, WANG Xian, LI Yuzeng. Risk measure of electricity market based on ARMAX-GARCH model with conditional skewed-t distribution and extreme value theory [J]. East China Electric Power, 2013, 41(6): 1335-1440
[17] WANG Ruiqing, WANG Fuxiong, XU Miaocun. ARMAX-GARCHSK-EVT Model Based Risk Measure of Electricity Market [C]. Proceedings of the 32nd Chinese Control Conference, Xi’an, China, July 26-28, 2013: 8284-8288
[18] DONG Lili, HUANG Dao. Fault diagnosis and prediction based on new information GM(1,1) [J]. Control Engineering of China, 2006, 13(3): 252-255
[19] Balkema A, De Haan L. Residual life time at great age [J]. The Annals of Probability, 1974, 2(5): 792-804
[20] Pickands, J. Statistical inference using extreme order statistics [J]. The Annals of Statistics, 1975, 3(1): 119-131
[21] Gilli M, Kellezi E. An application of extreme value theory for measuring financial risk [J]. Computational Economics, 2006, 27(2): 207-228
[22] Longin FM. From value at risk to stress testing: The extreme value approach [J]. Journal of Banking & Finance, 2000, 24: 1097-1130
[23] Ibrahim J G, Chen M H, Sinha D. Bayesian survival analysis [M]. New York: Berlin Heidelberg, 2001
[24] MAO Shisong, WANG Jinglong, PU Xiaolong. Advanced Mathematical Statistics [M]. Beijing: Higher Education Press, 2004
[25] McNeil A J, Frey R. Estimation of tail-related risk measures for heteroscedastic financial time series: an extreme value approach [J]. Journal of Empirical Finance, 2000, 7(3-4): 271-300
[26] Kupiec P. Techniques for verifying the accuracy of risk measurement models [J]. Journal of Derivatives, 1995, 39(2):73-84
[27] Coles S. An introduction to statistical modeling of extreme values [M]. London: Springer-Verlag, 2001