金融时序分析中动态波动模型的检验

商家名称	信用等级	购买信息	订购本书
		金融时序分析中动态波动模型的检验
		金融时序分析中动态波动模型的检验

基本信息·出版社：首都经济贸易大学出版社
·页码：132 页
·出版日期：2009年06月
·ISBN：9787563816774
·条形码：9787563816774
·版本：第1版
·装帧：平装
·开本：32
·正文语种：英语

内容简介 《金融时序分析中动态波动模型的检验》主要讲述了：波动模型是分析金融数据、估算金融风险的主要工具之一，受到了学界和业界的广泛关注。虽然金融波动模型的参数估计和实证方法备受重视，但对模型的检验方法的研究却被人们严重地忽视了。众所周知，模型的前提条件关系到模型的应用以及结论的正确与否。遗憾的是，由于仅有的几篇关于模型检验的文献都是基于沃尔德检验或尤度比检验的思想，面临着不可定义参数检验的难题，无法保证其检验结果的正确性。
《金融时序分析中动态波动模型的检验》借鉴量子力学中的德鲁塔函数的思想，率先建立拉格朗日乘数检验统计量，以规避不可定义参数的检验问题。各章的主要内容分别如下：
第一章，介绍金融波动模型及其相互关系；
第二章，在随机波动模型的基础上，提议检验EGARCH模型的拉格朗日乘数检验统计量，并通过计算机仿真和实证分析，验证该检验统计量的检验能力；
第三章，在Jump—GARCH模型的基础上，提议检验跳跃现象存在与否的拉格朗日乘数检验统计量，并应用计算机仿真，验证该检验统计量的正确性；
第四章，在Jump—GARCH（t）模型的基础上，提议检验跳跃现象的拉格朗日检验统计量，并用计算机仿真和实证分析加以验证；
第五章，分别在Jump—EGARCH模型和Jump—EGARCH（t）模型的基础上，提议检验跳跃现象的拉格朗日乘数检验统计量，并通过计算机仿真和实证分析加以验证；
第六章，在Jump-SV模型的基础上，提议检验跳跃现象的拉格朗日乘数检验统计量，并通过计算机仿真和实证分析验证该检验统计量的检验效率。
编辑推荐 《金融时序分析中动态波动模型的检验》是由史秀红所编著，首都经济贸易大学出版社出版发行的。
目录
1 Financial Volatility Models
1.1 Stylized Facts
1.1.1 ARCH-type Mode]
1.1.2 Stochastic Volatility (SV) Model
1.1.3 Jump Process
1.2 The Relationships of the Three Models
1.2.1 ARCH-type and SV Models
1.2.2 ARCH-type and SV Models with Jump Components
1.2.3 Purpose for Testing
1.2.4 Purpose of This Book
1.3 Methodology
1.3.1 Lagrange Multiplier Test
1.3.2 Dirac's Delta Function
1.4 Structure of This Book
References

2 Testing for EGARCH against Stochastic
Volatility Models
2.1 Introduction
2.2 Model and Test Statistic
2.3 Conclusions
Appendix
References

3 Testing for GARCH against Jump-GARCH Models
3.1 Introduction
3.2 Model and the Lagrange Multiplier Test Statistic
3.3 Simulation
3.4 Conclusions
Appendix A
Appendix B
Appendix C
References

4 Testing for Jumps in the GARCH(t) Jump Processes
4.1 Introduction
4.2 Model and Lagrange Multiplier Test Statistic
4.3 A Monte Carlo Experiment and an Empirical Example
4.4 Algebraic Details
References

5 Testing for Jumps in the EGARCH Process
5.1 Introduction
5.2 Lagrange Multiplier Test for Jump-EGARCH with Gaussian Innovations
5.3 Jump-EGARCH with Student-t Innovations




5.4 One-sidedTest
5.5 AMonteCarloExperimentandanEmpiricalExample
References

6 TestsforJumpsinStochasticVolatilityProcesses
6.1 Introduction
6.2 TestingforSimpleSVagainstSVwithJumpsinReturns
6.2.1 SVMoldewithJumpsinReturns
6.2.2 TestStatistic
6.3 TestingforJumpsintheVolatilityCorrelatedwithJumpsinReturns
6.3.1 Model
6.3.2 TestStatistic
6.4 TestingforJumpsinVolatilityIndependentofJumpsinReturns
6.4.1 TheModel
6.4.2 TestStatistic
6.5 EmpiricalExamplesandMonteCarloExperiment
Appendix
References

后记
……
文摘 插图：


This chapter considers the EGARCH model with jumps （Jump-EGARCH）and proposes the Lagrange multiplier （LM） test statisticfor no jumps. The null hypothesis Of no jumps is defined by either zerojump variance or zero jump 'probability in the EGARCH-jump modelwith Gaussian innovations. It is shown that, in the former case wherethe jump probability is a nuisance parameter unidentified under thenull; the nuisance parameter is cancelled out in the LM test statistic,and hence the test is nuisance parameter-free. In the later case weshow that the same nuisance parameter-free test statistic is obtained, ifvariance jump is small.
The derivative of the likelihood function of Jump-EGARCH withStudent-t innovations cannot be computed straightforwardly, since thelikelihood under the null is expressed as an integral of the jump-sizedensity with all probability mass at zero. We obtain the score functionapplying the derivative formula of Dirac's delta function to thedegenerate jump density. This formula is only a direct application ofintegration by parts to a degenerate density function. We show that theLM test for,this model is a nuisance-parameter free. The one-sideversion of the LM test is also proposed using nonnegative constraint onthe jump variance.
The rest of this chapter is organized as follows. The LM teststatistic is proposed: for the Jump-EGARCH model with Gaussianinnovations in Section 5.2. Section 5.3 considers Jump-EGARCHwith Student-t innovations. In Section 5.4 the one-side version of LMtest is obtained. In Section 5.5 the actual level and power of the test

查看更多 下一篇