摘要
为了应对全球气候变化带来的挑战，旨在限制发达国家温室气体排放量以抑制全球变暖的《京都议定书》，把市场机制作为解决以二氧化碳为代表的温室气体减排问题的新路径。以二氧化碳排放权为基础的全球碳交易市场由此形成并迅速发展。2005年成立的欧盟排放交易体系（EU ETS）所开创的欧盟配额（EUA）现货及其衍生产品交易市场在全球碳市场中占据主导地位，是最具流动性和影响力的碳市场。但是随着欧盟配额市场的发展，欧盟配额价格却波动剧烈，欧盟配额现货和期货市场之间的相依性为价格波动带来的风险在现货和期货市场之间传播提供了渠道，并会使风险作用加强，从而可能引发系统性风险。这种市场之间的波动传导机制被称为风险溢出效应。碳现货和期货市场之间的相依性和风险溢出效应，成为投资者规避碳市场风险和政策制定者探索碳市场稳定机制必须首先解决的核心问题。中国作为发展中大国，目前虽然没有强制减排义务，但却是世界上第一排放大国。中国因此面临着巨大的减排压力。与此同时，中国也是国际碳市场最大的供应方，却必须遵循最大买方欧盟的价格机制。为了改变这种状况，中国也希望通过建立国内碳市场达到减排效果，并改变在全球碳市场所处的不利地位。中国因此已在积极探索如何建立国内碳市场。由于欧盟排放交易体系是目前全球最成功的碳市场，为了顺利建设国内碳市场，中国需要学习欧盟排放交易体系的经验和教训。基于此，本文主要以欧盟碳市场作为研究样本，研究碳市场相依结构和风险溢出效应，在此基础上，进一步研究其对碳排放权价格波动影响，并希望本文的研究结论能为中国建立国内碳市场提供帮助。
欧盟排放交易体系分三个阶段进行，由于第三阶段刚刚开始，因此，本文的研究主要集中在第一阶段和第二阶段。本文首先介绍欧盟排放交易体系的创建和运行情况，分析总结其经验教训，在此基础上，本文运用GARCH模型、DCC-GARCH模型、EVT-Copula模型、CoVaR指标，结合欧盟碳市场从2005年至2011年共6年的欧盟配额现货和欧盟配额期货日收盘价数据，实证研究欧盟配额现货和期货市场的相依结构和风险溢出效应。在以上研究的基础上，结合中国实际情况，对中国建立碳排放权交易市场进行深入讨论。
本文的主要研究结论如下：
首先，本文从总结欧盟排放交易体系经验为起点。欧盟排放权交易体系开创了一个全新的碳排放权交易市场，这一市场的建立和发展不仅仅是为碳排放权确立了一个价格，更重要的是在企业和个人中形成了未来碳排放权资产将有价值上涨空间的预期。不过，欧盟配额价格的大起大落，市场价格稳定问题已经引起关注。欧盟经验表明市场的规则对于碳市场的发展有显著的影响。在学习欧盟排放交易体系的成功经验和失误教训的同时，必须认真研究碳排放权交易市场有效运行的边界条件、前提条件，及其对市场经济发展阶段的要求等因素，并结合各自国家的国情和经济发展的特点深入开展研究。因此，本文希望通过研究碳市场的相依结构和风险溢出效应，能为中国建立碳排放交易体系做点边际贡献。
其次，本文通过DCC-GARCH模型来描述欧盟配额期货和现货收益率的动态条件相关性，实证研究结果表明从第一阶段到第二阶段，欧盟配额现货市场与期货市场之间的动态相关性是逐步加强的。在第一阶段欧盟配额现货市场和期货市场之间的互动关系不强，可能与第一阶段碳市场制度的不完善，市场之间的信息传递存在障碍有关。而从第二阶段开始，由于配额分配机制逐渐成熟，市场发展渐趋规范，欧盟配额现货市场与期货市场之间的动态相关性变得越来越强，但是两个阶段的动态相关系均存在大幅度波动，且大多数时间内都为正。这说明碳现货和期货这两个市场基本上保持同向变化，但关系不稳定，且受经济因素影响很大，进一步证明欧盟碳市场存在极端风险。
为了深入研究碳现货与期货市场的相依关系，本文通过采用极值分布为边缘分布构造Tawn Copula函数，Tawn Copula函数能极大限度地捕捉到第二阶段的欧盟配额现货市场与期货市场尾部的相依关系。经研究表明，在第二阶段，欧盟配额现货和期货市场在活跃时期的相关性显著高于其低迷时期的相关性。当碳价疯狂上涨时，欧盟配额现货和期货市场之间存在着同时相互影响、相互加强的双边风险关联关系。不过，它们的尾部相关结构也可以用BB7Copula函数进行刻画，这一结果意味着当碳市场出现处于持续下跌、剧烈波动等极端风险事件时，碳市场的风险关联性也相互加强。因此，针对欧盟配额现货和期货市场，本文建议碳市场监管者可以考虑设定一个恰当的稳定机制，并尽可能不干扰市场正常波动和交易情况。这对于应对碳市场突发风险，是极其必要的。
为了能进一步了解碳现货与期货市场间风险溢出的程度，本文采用条件在险值CoVaR的方法评估当碳现货市场出现风险事件后，碳期货市场的风险水平。研究表明，方法可以很好估计碳市场风险溢出效应强度。研究结果也显示，如果采用传统的VaR评估欧盟碳市场风险，会严重低估欧盟配额期货市场风险水平。与传统的VaR方法相比，CoVaR评估方法最大的作用是能够用来捕捉单个碳市场发生风险事件时，对整个碳市场体系的溢出效应，即系统风险的变化。对于碳市场监管部门来说，在清楚单个碳市场的风险溢出效应的强度前提下，市场监管不再拘泥于单个碳市场的风险监管，而是可以着眼于整个碳市场体系的潜在风险，从而可以防范、控制和化解系统性风险，维持稳定的市场环境，以促进整个碳市场健康、安全、高效地运行。
在此基础上，本文通过采用GARCH模型检验碳价的波动性，实证研究结果发现在第一阶段的欧盟配额现货和期货市场的波动具有显著的不对称性；而且第一阶段波动的衰减速度比第二阶段的快。通过分析碳市场波动的异方差图，发现欧盟配额市场存在着极端风险。
对于中国而言，建立国内的排放权交易体系有着更为重要的意义。结合碳市场相依结构和风险溢出效应的研究结论，本文提出了中国建立碳排放权市场应以核证减排量期货为起点，逐步建立中国碳金融衍生品市场；并提出在建立碳市场过程中，要注重市场制度建设，特别要注重市场稳定机制的建立，加强监管；要关注各个碳市场之间的相依关系，建立风险防范机制，以防范碳市场各种潜在风险通过市场间相依性进行传播并引发系统风险，从而实现整个碳市场体系的稳定与安全。另外，结合中国当前碳市场发展情况，本文还建议中国应加强金融机构特别是商业银行的支撑，加强立法及政策保障，加快相关领域人才的培养，增强国内科研实力，同时还需加强全球气候变化政策的演变等风险应对能力建设，以更好地规避在我国建立碳交易市场遇到的各类风险。
The Kyoto Protocol is designed to limit greenhouse gas emissions in developed countriesin order to respond to the challenges posed by global climate change. Cap-and-Trade of theKyoto Protocol becomes a new market tool to solve greenhouse gas emissions. The globalcarbon trading market based on carbon dioxide emissions was quickly set up and rapidlygrowing. Since the EU Emissions Trading Scheme (EU ETS) was established in2005, EUemission Allowance (EUA) market becomes one of the most liquid and influence emissionstrading markets in the world. However, the volatility of carbon price is severe. As a result, riskassociated with fluctuations in carbon price could spread through carbon markets. As carbonprice risk was quickly accumulated, it would give rise to systemic risk. This kind of marketvolatility transmission mechanism is known as risk spillover. Therefore, a research ondependency structure between EU carbon markets and their risk spillover effects becomesimportant for investors to avoid risk and for policy makers to explore the core of carbonmarket stability mechanism. China is the most large developing country but withoutmandatory emission reduction obligations. However, as the world's largest emitter, Chinafaces enormous pressure to reduce emissions. While as the largest supplier of global carbonmarkets, China has to follow the biggest buyer(EU)’ pricing. Fortunately, China has beenactively exploring how to create a domestic carbon market to achieve emission reductioneffects and change an unfavorable position in global carbon markets. As we know, the EUEmissions Trading Scheme is the world's most successful carbon market. So it need learnadvantages and disadvantages of EU ETS to establish domestic carbon market. In this paper,it mainly research EUA spot market and its derivatives markets dependency structure andtheir risk spillover effects, and hoping that the conclusion of this paper can help theestablishment of a domestic carbon market
In this paper, firstly analysis and summarize the operating rules and the lessons of the EUemissions trading system. Based on using the GARCH model and the DCC-GARCH model,using the Copula Function and the extreme value theory to establish the Copula-EVT Model,building CoVaR indicators, and using six years of daily closing price data of carbon spots andcarbon futures within EU carbon market from2005to2011, empirical research on dependency structure and risk spillover effects of EU carbon spots and carbon futures markets.And then discuss how to establish a domestic carbon emissions trading market in China.
The main conclusions are as follows:
The EU emissions trading system is a new emissions trading market, it not onlyestablishes a price for emission rights, more importantly, but also gives a valuable upsideexpectations in the formation of future emissions assets for businesses and individuals.However, the carbon market price fluctuations, and rule-making for the development ofemissions trading market have a significant impact. To learn the success of the EU ETSexperience, it need study the boundary conditions and preconditions of effective operation indepth, and requirements for stage of development of the market economy and other factors,combining with their respective national circumstances and economic development features.
Through use GARCH models to test the volatility of carbon prices, and find that there issignificant asymmetry in the first phase of the carbon market volatility, and its fluctuations inthe decay rate is faster than the second stage. Through analyzing its heteroscedasticitydiagram, find that extreme risk exist in carbon market.
Based on the above analysis, using the DCC-GARCH model to describe the dynamicconditions correlation of EUA futures and spot yields from the first phase to the second stage,the dynamic correlation between EUA spots market and futures market is graduallystrengthening. It is not strong interactive relationship between EUA spots and futures marketsin the first phase, which maybe relate to the imperfect carbon market system in the first phase,there are barriers between the transmission of information between carbon markets. From thebeginning of the second stage, due to emission Allowance allocation mechanism maturing, thedynamic correlation between EUA spot market and futures market is becoming increasinglystrong, investors can take advantage of the correlation between volatility of spot and futuresfor effective hedging to avoid systemic risk.
Through using extreme value distribution for the marginal distribution and constructingTawn Copula function, it greatly captures tail dependency between EUA spot market andfutures market in the second phase. It empirically shows that, in the second stage, taildependency was significantly higher during a boom than a downturn. When the carbonmarket is skyrocketing, yield of EUA spot and futures interact each other, mutually reinforce bilateral risk relationship. While it also can be used by BB7Copula function to portray, thisresult means that when extreme risk events happening such as carbon market is in decline,carbon market risk reinforced each other, it is a true for carbon market. Therefore, inparticular, it is extremely necessary to set a appropriate stabilization mechanism for thepre-sudden crisis of carbon markets, as possible not interfere normal market fluctuations andtransactions.
Based on researching the carbon market dependency structure using Copula theory andextreme value theory, using conditions value at risk(CoVaR) to assess risk level of carbonfutures market when carbon spot market exists risk events. It shows that CoVaR can estimaterisk spillover effect. If using VaR to assess market risk, it would seriously underestimate risklevel of carbon futures market. For market regulatory authorities, it need only know that asingle carbon market risk spillover strength. That is, no longer rigidly adhering to a singlemarket risk regulation, they just have to focus on the potential risks of the entire carbonmarket system to guard against, control and mitigate systemic risk, and maintain a stablemarket environment, to promote carbon market healthy, safe and efficient operation.
It is more important for China to establish a domestic emissions trading system.Combined with the findings of dependency structure and risk spillover effect between EUcarbon markets, this paper presented for the first time that it should be based on CERs futuresas a starting point, and then to establish carbon financial derivatives step by step in China. Itneed not only to pay attention to rationalize the relationship between the carbon spot andcarbon futures market, but also to establish risk prevention mechanism to prevent potentialrisk of carbon price fluctuations in order to achieve stability and security of the entiredomestic carbon market system.
引文
[1]《气候变化——人类面临的挑战》编写组主编.《气候变化——人类面临的挑战》[M].气象出版社,2007
[2] World Bank. STATE AND TRENDS OF THE carbon market2011.[EB/OL].http://web.worldbank.org/,2011-6-20
[3]王遥，刘倩.碳金融市场：全球形势、发展前景及中国战略[J].国际金融研究.2010(9):64-74
[4]刘茂平.国际碳金融发展现状及我国的选择[J].海南金融.2010(11):18-21
[5]樊纲主编.走向低碳发展：中国与世界——中国经济学家的建议[M].北京：中国经济出版社,2010
[6]魏一鸣，王恺，凤振华，等.碳金融与碳市场——方法与实证[M].北京：科学出版社,2010:9
[7] Paolella M S, Taschini L. An econometric analysis of emission allowance prices[J]. Journal ofBanking&Finance.2008,32(10):2022-2032.
[8] Daskalakis G, Markellos R N. Are the European Carbon Markets Efficient?[J]. Review ofFutures Markets.2008,17(2):103-128
[9] Frunza, E.P.M.C., Guegan, D. Derivative Pricing and Hedging on Carbon Market[R]. CESWorking Paper,2010
[10]王恺，邹乐乐，魏一鸣.欧盟碳市场期货价格分布特征分析[J].数学的实践与认识.2010(12):59-65
[11] Seifert, Uhrig-Homburg, Wagner. Dynamic behavior of CO2spot prices[J]. Journal ofEnvironmental Economics and Management.2008,56(2):180-194.
[12] Chevallier. Carbon futures and macroeconomic risk factors: A view from the EU ETS[J].Energy Economics.2009,31(4):614-625
[13]陈晓红，王陟昀.欧洲碳排放权交易价格机制的实证研究[J].科技进步与对策.2010(19):142-147
[14] Benz, Trück. Modeling the price dynamics of CO2emission allowances[J]. EnergyEconomics.2008,31(1):4-15.
[15] Daskalakis, Markellos, Psychoyios, et al. Modeling CO2emission allowance prices andderivatives: Evidence from the European trading scheme[J]. Journal of Banking&Finance.2009,33(7):1230-1241
[16] Chevallier J. Nonparametric modeling of carbon prices[J]. Energy Economics.2011, In Press,Corrected Proof
[17] Conrad, Rittler, Rotfu. Modeling and explaining the dynamics of European Union Allowanceprices at high-frequency[J]. Energy Economics.2012, In Press, Corrected Proof
[18] Uhrig-Homburg M, Wagner M. Derivative Instruments in the EU Emissions Trading Scheme-An Early Market Perspective[J]. Energy&Environment.2008,19(5):635-655.
[19] Chevallier J, Le Pen Y, Se Vi B. Options introduction and volatility in the EU ETS[J].Resource and Energy Economics.2011,33(4):855-880
[20] Feng Z, Zou L, Wei Y. Carbon price volatility: Evidence from EU ETS[J]. Applied Energy.2011,88(3):590-598
[21] Arouri M E H, Jawadi F, Nguyen D K. Nonlinearities in carbon spot-futures pricerelationships during Phase II of the EU ETS[J]. Economic Modelling.2012(1):450-463
[22] Alberola, Chevallier. Price drivers and structural breaks in European carbon prices2005-2007[J]. Energy Policy.2008,36(2):787-797.
[23] Chevallier J. Carbon futures and macroeconomic risk factors: A view from the EU ETS[J].Energy Economics.2009,31(4):614-625.
[24] Ellerman, A.D., CONVERY, F.,&DE PERTHUIS, C.. Pricing Carbon: The European UnionEmissions Trading Scheme[M]. Cambridge:Cambridge University Press,2010:32-158.
[25] Convery F. J., Redmond L.. Market and price developments in the European Union Emissions TradingScheme[J]. Rev Environ Econ Policy.2007,1(1):88–111.
[26]魏一鸣，刘兰翠，范英，等.中国能源报告（2008）：碳排放研究[M].北京：科学出版社,2008:167-171.
[27] Linares P, Santos F. J., Ventosa M, Lapiedra L.. Impacts of the European emission tradingscheme directive and permit assignment methods on the Spanish electricity sector[J]. Energy J.2006,27(1):79–98.
[28] Smale R, Hartley M, Hepburn C, Ward J, Grubb M. The impacts of CO2emissions trading onfirm profits and market prices[J]. Clim Policy.2006,6(1):29–46.
[29] Kara M, Syri S, Lehtil A, Helynen S, Kekkonen V, Ruska M, et al. The impacts of EU CO2emissions trading on electricity markets and electricity consumers in Finland[J]. Energy Econ.2008,30(2):193–211.
[30] Zachmann G, von Hirschhausen C. First evidence of asymmetric cost passthrough of EUemissions allowances: examining wholesale electricity prices in Germany[J]. Econ Lett.2008,99(3):465–469.
[31] Daskalakis G, Markellos RN. Are electricity risk premia affected by emission allowanceprices? Evidence from the EEX, Nord Pool and Powernext[J]. Energy Policy.2009,37(7):2594–2604.
[32] Streimikiene D, Roos I. GHG emission trading implications on energy sector in BalticStates[J]. Renew Sustain Energy Rev.2009,13(4):854–862.
[33] Haar LN, Haar L. Policy-making under uncertainty: commentary upon the European Union EmissionsTrading Scheme[J]. Energy Policy.2006,34(17):2615–2629.
[34] Daskalakis G, Markellos R N. Are the European Carbon Markets Efficient?[J]. Review of Futures Markets.2008,17(2):103-128.
[35] Charles A, Darn O, Fouilloux J. Testing the martingale difference hypothesis in CO2emissionallowances[J]. Economic Modelling.2010,28(1-2):27-35
[36]汪文隽.欧盟排放权配额交易市场的价格行为及市场效率[D].中国科学技术大学，2011
[37]张跃军，魏一鸣.国际碳期货价格的均值回归:基于EU ETS的实证分析[J].系统工程理论与实践.2011(02):34-45
[38]凤振华，魏一鸣.欧盟碳市场系统风险和预期收益的实证研究[J].管理学报.2011(03):451-455
[39]魏一鸣，王恺，凤振华，等.碳金融与碳市场——方法与实证[M].北京：科学出版社,2010:112-122
[40]刘维泉，郭兆晖. EU ETS碳排放期货市场风险度量——基于SV模型的实证分析[J].系统工程.2011(10):14-23.
[41] Klugman, S. A., Parsa, R. Fitting Bivariate loss distributions with copulas[J].Insurance Math.Econom.1999(24:)139-148
[42] Patton, A. J. Modelling Asymmetric Exchange Rate Dependence[J]. International EconomicReview.2006a,47(2):527-555
[43] Patton, A. J. On the Out-of Sample Importance of Skewness and Asymmetric Dependence forAsset Allocation[J]. Journal of Financial Econometrics.2004,2(1):130-168
[44] Sklar, A. Fonctions de Repartition a n Dimensionset Leurs Marges[J]. Publications de l'Institut de Statistiquede l' Universite de Paris,1959(8):229-231
[45] Joe, H. Multivariate Models and Dependence Concepts[M]. Chapman&Hall.1997
[46] Nelsen, R. An Introduction to Copulas[M]. Springer. New York.1999
[47] Embrechts, P., Lindskog, F. and McNeil, A. Modeling Dependence with Copulas andApplications to Risk Management[C]. Handbook of Heavy Tailed Distributions in Finance.2003:329-384
[48] Patton, A. J. Estimation of multivariate models for time series of possibly different lengths[J].Journal of Applied Econometrics.2006b,21(2):147-173
[49] Jondeau, E., Rockinger, M. The Copula-Garch model of conditional dependencies: Aninternational stock market application[J]. Journal of International equity markets[J]. Journalof Financial and Quantitative Analysis.2008,43(3):787-816
[50] Longin, F. From Value at Risk to Stress Testing:the Extreme Value Approach[J]. Journal ofBanking and Finance.2000(24):1097-1130
[51] Frey, R. and McNeil, A. Modeling Dependent Defaults[J]. Journal of Risk.2003,6(1):59-92
[52] Glasserman, P., Heidelberger, P. and Shahabuddin, P. Portfolio Value-at-risk with Heavy-tailed Risk Factors[J]. Mathematical Finance.2002(12):239-270
[53] Bouye, E., Gaussel, N. and Salmon, M. Investigating Dynamic Dependence Using Copula[C].Manuscript. Financial Econometrics Research Center.2002
[54] Embrechts, P., Lindskog, F. and McNeil, A. Modeling Dependence with Copulas andApplications to Risk Management[J]. Handbook of Heavy Tailed Distributions in Finance.2003:329-384.
[55] Rosenberg, J. and Schuermann, T. A General Approach to Integrated Risk Management withSkewed, Fat-tailed Risk[R]. FRB of New York Staff Report.2005:185
[56] Longin, F. and Solink, B. Extreme Correlation of International Equity Markets[J]. Journal ofFinance.2001(56):649-676
[57]张尧庭.我们应该选用什么样的相关性指标？[J].统计研究.2002(9):41-44
[58]张尧庭.连接函数（Copula）技术与金融风险分析[J].统计研究.2002b(10):48-51
[59]傅强，邢琳琳.基于极值理论和Copula函数的条件VaR计算[J].系统工程学报.2009(5):531-537
[60]韦艳华. Copula理论及其在多变量金融时间序列分析上的应用研究[D].天津大学,2004
[61]韦艳华，张世英.金融市场非对称尾部相依结构的研究[J].管理学报.2005(5):601-605
[62]李秀敏，史道济.金融市场组合风险的相关性研究[J].系统工程理论与实践.2007(2):112-117
[63]包卫军，徐成贤.基于SV-copula模型的相关性分析[J].统计研究.2008(10):100-104
[64] Bertero E, Mayer C. Structure and performance: global interdependence of stock marketsaround the crash of october1987[Z]. C. E. P. R..1989
[65] King M, Wadhwani S. Transmission of volatility between stock markets[J]. The Review ofFinancial Studies.1990(3):5-33
[66] Cavlo S, Reinhart C. Capital Flows to latin America: Is There Evidence of ContagionEffects[Z]. IMF.1996
[67] Baig T, Goldfaijn I. Financial market contagion in the Asian crisis[Z]. IMF.1998
[68] Boyer B H, Gibson M S, Loretan M. Pitfalls in tests for changes in correlation[Z].International Finance Discussion Paper No.597,1999
[69] Lin, W..Engle.R.F.&Ito, T. Do Bulls and Bears Move across Borders? InternationalTransmission of Stock Returns and Volatility [J]. Review of Financial Studies.1994,7(3):507-538
[70] Hamao,Y.. Masulis, R.W.&Ng. V. Correlations in Price Changes and Volatility acrossInternational Stock Markets[J]. Review of Financial Studies.1990.3(2):281-307
[71] Hartman. P., Straetmans,&C.G. de Vries. Asset Market Linkages in Crisis Periods[J]. Reviewof Financial Studies.2004,86(1):313-326
[72] Theodossiou&Lee. Mean and Volatility Spillovers Across Major National Stock Markets:Further Empirical evidence[J]. The Journal Financial Research.1993,16(4):337-350
[73] YA Liu&MS Pan. Mean and Volatility Spillover in the U.S. and Pacific-Basin StockMarkets[J]. Multinational Finance Journal.1997,1(1):47-62
[74]赵留彦，王一鸣. A、B股之间的信息流动与波动溢出[J].金融研究.2003(10):37-52
[75]张瑞锋.金融市场协同波动溢出分析及实证研究[J].数量经济技术经济研究.2006(10):141-149
[76]张瑞锋，张世英，唐勇.金融市场波动溢出分析及实证研究[J].中国管理科学.2006(10):14-22
[77]陈云，陈浪南，林鲁东.人民币汇率与股票市场波动溢出效应研究[J].管理科学.2009(3):104-1121
[78]万军，谢敏，熊正德.金融市场间波动溢出效应研究[J].统计与决策.2007(18),98-10
[79] Adrian, Brunnermeier(September2008),“CoVaR”,Federal Reserve Bank of New York StaffReports no.348
[80] Adrian, Brunnermeier (April2009),“CoVaR” working paper, Federal Reserve Bank of NewYork and Princeton University
[81] Adrian, Brunnermeier (August2009),“CoVaR” working paper, Federal Reserve Bank of NewYork and Princeton University
[82] Engle, R. F.. Dynamic conditional correlation: a simple class of multivariate generalizedautoregressive conditional heteroskedasticity models[J]. Journal of Business and EconomicsStatistics.2002,20(3):339-350
[83] Bollerslev T.. Modeling the coherence in short run nominal exchange rates: A multivariategeneralized ARCH Model[J]. Review of Economics and Statistics.1990,96(1):116-131
[84] Nelson, R. B. An introduction to Copulas [M]. New York:Springer-Verlag.1999
[85] McNeil, A.J., Frey, R., Embrechts, P. Quantitative Risk Management:Concepts, Techniquesand Tools[M]. Princeton University Press, New Jersey,2005
[86] Zivot, E., Jiahui, W. Modelling Financial Time Series with S-PLUS [M]. Second Edition.2005.http://www.insightful.com
[87] Capéraà, P., Fourgères, A. L., Genest, C. Bivariate Distributions with Given Extreme ValueAttractor [J]. Journel of Multivariate Analysis,2000(72):30-49
[88] Schweizer, B., Wolff, E.F. On nonparametric measures of dependence for random variables[J].Ann Statist.1981(9):879-885
[89] Boyer B H, Gibson M S, Loretan M. Pitfalls in tests for changes in correlation[Z].International Finance Discussion Paper No.597,1999
[90] Forbes, K. J., Rigobon, R.. No contagion, only interdependence: Measuring stock marketscomovements[J]. Journal of Finance.2002(57):2223-2261
[91] Longin, F., Solnik, B.. Extreme Correlation of International Equity Markets [J]. Journal ofFinance.2001,56(2):649-676
[92] Ledford, A. W., Tawn, J. A. Modelling dependance with in joint tail regions [J]. J. R. StatistSoc B.1996(2):475-499
[93] Beirlant, J. Statistics of extremes: theory and applications[M]. John Wiley&Sons Ltd.2004
[94] Hill, B. M. A simple general approach to inference about the tail of a distribution[J]. Ann.Statist.1975(3):1163-1174
[95] McNeil, A J., Frey, R. Estimation of Tail-Related Risk Measure for Heteroscedastic FinancialTime Series: An Extreme Value Approach[J]. Journal of Empirical Finance.2000(7):271-300
[96] Patie, P. Estimation of Value at Risk Using Extreme Value Theory [R].http://www.math.ethz.ch/~patie/VaREvT.2000.
[97] Beirlant, J., Dierckx, G., Goegebeur, Y. et al. Tail index estimation and an exponentialregression model[J]. Extremes.1999(2):177-200.
[98] Hosking, J.R.M., Wallis J.R. Parameter and quantile estimation for the generalize Paretodistribution[J]. Technometrics.1987(29):339-349
[99] Rao, C.R.. Linear Statistical Inference and its Applications[M].2th Edition. New York: JohnWiley.1973
[100] Smith, R. L. Maximum likelihood estimation in a class of non-regular cases[J]. Biometrika.1985,72(1):67-90
[101] Matthys, G., Beirlant, J.. Estimating the extreme value index and high quantiles withexponential regression models[J]. Statistica Sinica.2003(13):853-880Pickands, J. Statisticalinference using extreme order statistics [J]. Ann.Statist.1975(3):119-131
[102] Azzalini, A.. Statistical Inference Based on the Likelihood[R]. London:Chapman and Hall,1996
[103] Joe, H.&Xu, J.J. The Estimation Method of Inference Functions for Margins for MultivariateModels[R]. Dept. of Statistics University of British Columbia, Tech. Rept.166.1996
[104] Quesenberry, C. P.. Some Transformation Methods in Goodness-of-Fit[R]. Chapter6inGoodness-of-Fit Techniques, Edited by R. B. D'Agostino and M. A. Stephens. Marcel Dekker,Inc., New York.1986
[105] Klugman, S., Parsa, R.. Fitting bivariate loss distributions with copulas[J]. Insurance:Mathematics and Economics.1999(24):139-148
[106] Shao, J.. Mathematical Statistics[R]. New York:Springer-Verlag,1999
[107] Hu, Y. T., Kiesel R., Perraudin, W.. The estimation of transition matrices for sovereign creditratings[J]. Journal of Banking and Finance.2002(26):1383-1406
[108] Akaike, H.. A new look at the statistical model identification. IEEE Transactions onAutomatic Control.1974,19(6):716-723
[109] DuMouchel W.M. Estimating the Stable Index in Order to Measure Tail Thickness: ACritique[M]. Annals of Statistics11.1983:1019-1031
[110] Artzner P, Delbaen F, Eber J, Heath D. Coherent measures of risk[J]. Mathematical Finance.1999(3):203-228
[111] Acerbi C., Tasche D.. Expected shortfall: a natural coherent alternative to value at risk.Working Paper.2001Beirlant, J. Statistics of extremes: theory and applications[M]. JohnWiley&Sons Ltd.2004
[112] Bollerslev T.. Modeling the coherence in short run nominal exchange rates: A multivariategeneralized ARCH Model[J]. Review of Economics and Statistics.1990,96(1):116-131
[113] Jagannathan R., Runkle D. E.. On the relations between the expected value and the volatilityon the nominal excess returns on stocks[J]. Journal of Finance.1993,48(5):1779-1801
[114] Glosten L.R., Zakoian J. M.. Threshold heteroskedastic models[J]. Journal of EconomicDynamics and Control.1994,18(5):931-955
[115] World Bank.10Years of Carbon Finance Experience-Insights From Working with CarbonMarkets for Development&Global Greenhouse Gas Mitigation[EB/OL].http://web.worldbank.org/,2010-3-20