上海证券市场回报率分析——基于ARIMA模型和GARCH模型

本文分析中国上海证券市场回报率。分别通过APdMA模型和GARCH模型，发现若用APdMA模型分析和建立时间序列模型，一次自回归项是不够的，需要高次项，在大多数情形，若运用GARCH模型，则GARCH（1，1）就能够很好的拟合数据。     

基于Expectile和Realized GARCH模型的波动率预测

Realized GARCH模型是预测波动率的经典模型之一,最小化非对称二次损失函数的Expectile对收益率尾部分布更加敏感,我们在Realized GARCH模型的基础上引入Expectile提出Expectile-Realized GARCH模型。以沪深300指数的高频收益率为例建模分析,对比不同模型下的波动率预测效果,发现Expectile-Realized GARCH模型较Realized GARCH模型对波动率预测能力更好。其中,当风险水平为95%时,对应的Expectile-Realized GARCH波动率预测能力最好。     

向量GARCH族模型计算中国金融市场投资组合风险价值(VaR)的比较研究与评述

对由上证综合指数、深证成分指数、上证基金指数、上证国债指数计算的日自然对数收益率组成的数据矩阵,分别建立了残差服从正态分布、t分布的向量ARCH、向量GARCH、纯对角GARCH、BEKK、常条件相关GARCH、主成分GARCH和EWMA模型,基于这些模型,计算了风险价值(VaR),进而通过比较计算结果,得出BEKK—t模型测算中国金融市场投资组合的风险价值(VaR)效果最好等的结论.     

基于MCMC抽样的金融贝叶斯半参数GARCH模型研究

GARCH模型是研究金融资产收益的重要模型,然而现有参数GARCH模型依然不能有效刻画金融资产收益偏态厚尾特性且存在模型设定风险。本文在非参数分布和GARCH模型基础上,建立半参数GARCH模型以提高模型的有效性;同时在贝叶斯框架内发展有效MCMC抽样解决模型的参数估计难问题,并利用DIC4研究模型比较问题;最后通过模拟研究和实证研究考察MCMC抽样的有效性,检验半参数GARCH模型在刻画金融资产收益特性和风险价值预测方面的实际效果。     

基于模糊GJR-GARCH模型的波动率估计

基于模糊数学和模糊时间序列分析理论,在模糊GARCH与GJR-GARCH模型的基础上建立模糊GJR-GARCH模型,并用遗传算法估计了该模型的参数。实证发现沪深两市的收益波动率具有明显的非对称性,相对于普通的GARCH、GJR-GARCH和模糊GARCH模型,模糊GJR-GARCH模型能更好的处理收益率的波动群聚性、时变性和非对称性,具有更好的估计精度。     

基于三类模型的四大银行股票收益率预测研究

运用时间序列分析的预测方法,对四大银行的股票日对数收益率序列进行拟合与预测分析,分别构建ARMA模型、GARCH模型以及ARMA-GARCH组合模型,通过模型比较,实证分析表明:在拟合效果上,ARMA-GARCH模型的拟合优度优于ARMA模型和GARCH模型;在预测效果上,ARMA模型的预测效果最优,ARMA-GARCH模型次之.     

基于双因子已实现GARCH模型的波动率预测研究

准确地预测金融市场的波动率对市场管理者和参与者而言都是至关重要的。本文在标准已实现GARCH模型基础上,将条件方差乘性分解为长期方差和短期方差两部分,分别构造包含杠杆函数的长期方差方程和短期方差方程,用以捕捉波动率的长记忆性和短期微观波动。运用上证综指和日经指数的日收盘价、已实现方差和已实现核波动此类高频数据进行实证分析,结果表明:与标准已实现GARCH模型相比,两指数的双因子已实现GARCH模型在样本内表现出更大的似然估计值;通过样本外误差函数分析和DM检验,双因子已实现GARCH模型也取得更好表现。     

GARCH模型对上海股市的一个实证研究

GARCH模型是近20年发展起来的时间序列模型，它反映了经济变量之间特殊的不确定形式：方差随时间变化而变化，所以其在金融市场的预测与决策方面有着重要的作用。本文详细介绍了GARCH模型以及其主要变形，并建立了基于t分布和正态分布假设的GARCH（1，1）模型对股票市场进行了风险分析。结果表明，基于t分布的假设能更准确地拟和GARCH（1，1）模型。     

基于随机矩阵理论决定多元GARCH模型最佳维度研究

基于随机矩阵理论(RMT)的降维技术能够通过去除噪声和只保留有用“信息”,而对相关矩阵估计中用来描述相关的主成分或因子的最佳使用数量做出确定.本文认为利用RMT对相关矩阵估计的降维操作来实现RMT对多元GARCH模型的有效降维是可能的.为说明基于RMT的降维技术用于多元GARCH模型的有效性,本文建立了两类将基于RMT的相关矩阵估计和波动率结合在一起的多元GARCH模型:滑动相关多元GARCH模型(SC-GARCH模型)和改进的O-GARCH模型(IO-GARCH模型).理论分析表明,这两类模型具有降维的相关结构,易于估计,并且利用RMT能确定出它们的理论最佳维度.实证研究中,本文建立了上海证券市场100只股票收益率的两类多元GARCH模型,并在马克维茨证券组合理论的框架下,考察了它们的协方差矩阵预测效果.结果表明这两类模型的预测效果很好.通过两类模型各个维度预测效果的比较可以看出.RMT能够为多元GARCH的降维提供有效的依据并且较准确地确定多元GARCH模型的最佳维度.理论和实证分析结果表明,基于RMT的降维技术是解决多元GARCH模型“维数灾祸”问题的有效手段.     

利用高频数据和copula度量资产组合市场风险:建模与实证

利用realized GARCH模型纳入高频信息,通过copula函数描述资产收益之间的复杂相依关系,构建资产组合市场风险度量的copula-realized GARCH模型,并以上证综指和深证成指的等权重组合进行实证分析。实证结果表明,用学生t-copula函数描述上证综指和深证成指收益序列之间的相依关系比较恰当,与常用的GARCH模型和GJR-GARCH模型相比,学生t-copula与realized GARCH相结合的模型提供相对可靠、更具效率的VaR估计以及更准确的ES估计。因此,资产组合市场风险度量有必要同时考虑高频信息和复杂相依关系。     

Power transformation and threshold modeling for ARCH innovations with applications to tests for ARCH structure

A new class of power-transformed threshold ARCH models is proposed as a threshold-asymmetric generalization of the nonlinear ARCH considered by Higgins and Bera [Internat. Econom. Rev. 33 (1992) 137]. This class is rich enough to include diverse nonlinear and nonsymmetric ARCH models which have been spelled out in the literature. Geometric ergodicity of the model and existence of stationary moments are studied. The model facilitates discussing ARCH structures and hence large sample tests for ARCH structures are investigated via local asymptotic normality approach. Semiparametric tests are also discussed for the case when the error density is unknown.     

ARCH族模型及其对深圳股票市场的实证分析

ARCH族模型是动态非线性的股票定价模型,它在金融和经济领域具有广阔的应用前景.在短短二十年时间里取得了迅速的发展,先后提出了GARCH、ARCH-M、TARCH、EGRACH等模型丰富了ARCH族模型.本文从ARCH族模型出发,简要介绍了其主要形式和特点,然后将ARCH族模型运用于我国深圳股票市场的实证分析,从实证结果中总结深圳股市的总体特征.     

Value at risk calculation through ARCH factor methodology: Proposal and comparative analysis

In this paper we put forward a new method to estimate value at risk (VaR), autoregressive conditional heteroskedastic (ARCH) factor, which combines multivariate analysis with ARCH models. Firstly, from a set of correlated portfolio risk factors, we derive a smaller uncorrelated risk factors set, by applying multivariate analysis. Secondly, we use ARCH schemes to model uncorrelated factors historical behaviour. Thirdly, we use the estimated models to predict future values for factors standard deviation. From them, VaR calculation is immediate. In this way, ARCH factor methodology overcomes the multivariate ARCH models drawbacks, which, in practice, make these unworkable for VaR calculation purposes. We apply the proposed methodology over a set of foreign exchange risk exposed portfolios, obtaining better results than those reached when J.P. Morgan’s Riskmetrics is used.     

基于MRS-GARCH模型的中国股市波动率估计与预测

基于误差项服从正态分布、t分布、广义误差分布的GARCH族模型和MRS-GARCH模型对中国股市波动的结构变化特征进行了实证研究。结果表明,中国股市存在显著的高、低波动状态,两种波动状态的ARCH和GARCH项系数存在较大差异;高、低波动状态均具有较长的持续时间,低波动状态的持续时间长于高波动状态的持续时间,且中国股市更易于从高波动状态转向低波动状态;MRS-GARCH模型预测效果总体上优于GARCH族模型,基于正态分布的MRS-GARCH模型短期预测效果较好。     

Arch model with Box–Cox transformed dependent variable

Box–Cox power transformation has been used traditionally to linearise otherwise nonlinear models. In this paper, Engle's linear ARCH specification is considered for a regression model in which the dependent variable is Box–Cox transformed. The consequent issues arising in both testing and estimation of the model are investigated. A Lagrange multiplier test is also developed to test Engle's linear ARCH model against this wider class of models. The usefulness of this generalisation is examined by applying it to the daily closing prices on the Bombay Stock Exchange Sensitive Index, and the findings strongly favour the proposed model.     

Asymptotics of rank order statistics for ARCH residual empirical processes

This paper gives the asymptotic theory of a class of rank order statistics {T_N} for two-sample problem pertaining to empirical processes based on the squared residuals from two classes of ARCH models. An important aspect is that, unlike the residuals of ARMA models, the asymptotics of {T_N} depend on those of ARCH volatility estimators. Such asymptotics provide a useful guide to the reliability of confidence intervals, asymptotic relative efficiency and ARCH affection. We consider these aspects of {T_N} for some ARCH residual distributions via numerical illustrations. Moreover, a measure of robustness for {T_N} is introduced. These studies help to highlight some important features of ARCH residuals in comparison with the i.i.d. or ARMA settings.     

On the ARCH model with random coefficients

In the usual ARCH model, the coefficients have a degenerate distribution, and it is thus constant over realizations. In this Note we introduce the ARCH model, involving coefficients that are independent random variables and may vary over realizations. Conditions for the existence of a stationary solution and conditions ensuring the existence of higher order moments are obtained. The covariance structures of such models are studied. To cite this article: A. Bibi, M. Bousseboua, C. R. Acad. Sci. Paris, Ser. I 346 (2008).     

The GARCH(1,1)-M model: results for the densities of the variance and the mean

This paper starts from the GARCH(1,1)-M model of Bollerslev [Generalized autoregressive conditional heteroskedasticity, Journal of Econometrics 31 (1986) 307–327], and investigates the limit diffusion form as it is presented in Nelson [ARCH models as diffusion approximations, Journal of Econometrics 45 (1990) 7–38]. The distribution for the conditional variance process is derived, and in the limit for t going to infinity is shown to coincide with the stationary distribution given in Nelson [ARCH models as diffusion approximations, Journal of Econometrics 45 (1990) 7–38]. In addition it is shown how the distribution for the complete model can be arrived at; explicit calculations are given in case the conditional variance is a martingale.     

Estimating Functions for Nonlinear Time Series Models

This paper discusses the problem of estimation for two classes of nonlinear models, namely random coefficient autoregressive (RCA) and autoregressive conditional heteroskedasticity (ARCH) models. For the RCA model, first assuming that the nuisance parameters are known we construct an estimator for parameters of interest based on Godambe's asymptotically optimal estimating function. Then, using the conditional least squares (CLS) estimator given by Tjøstheim (1986, Stochastic Process. Appl., 21, 251–273) and classical moment estimators for the nuisance parameters, we propose an estimated version of this estimator. These results are extended to the case of vector parameter. Next, we turn to discuss the problem of estimating the ARCH model with unknown parameter vector. We construct an estimator for parameters of interest based on Godambe's optimal estimator allowing that a part of the estimator depends on unknown parameters. Then, substituting the CLS estimators for the unknown parameters, the estimated version is proposed. Comparisons between the CLS and estimated optimal estimator of the RCA model and between the CLS and estimated version of the ARCH model are given via simulation studies.     

Tail dependence of random variables from ARCH and heavy-tailed bilinear models

Discussed in this paper is the dependent structure in the tails of distributions of random variables from some heavy-tailed stationary nonlinear time series. One class of models discussed is the first-order autoregressive conditional heteroscedastic (ARCH) process introduced by Engle (1982). The other class is the simple first-order bilinear models driven by heavy-tailed innovations. We give some explicit formulas for the asymptotic values of conditional probabilities used for measuring the tail dependence between two random variables from these models. Our results have significant meanings in finance.