Estimating copula densities, using model selection techniques

Recently a new way of modeling dependence has been introduced considering a sequence of parametric copula models, covering more and more dependency aspects and thus giving a closer approximation to the true copula density. The method uses contamination families based on Legendre polynomials. It has been shown that in general after a few steps accurate approximations are obtained. In this paper selection of the adequate number of steps is considered, and estimation of the unknown parameters within the chosen contamination family is established, thus obtaining an estimator of the unknown copula density. There should be a balance between the complexity of the model and the number of parameters to be estimated. High complexity gives a low model error, but a large stochastic or estimation error, while a very simple model gives a small stochastic error, but a large model error. Techniques from model selection are applied, thus letting the data tell us which aspects are important enough to capture into the model. Natural and simple estimators of the involved Fourier coefficients complete the procedure. Theoretical results show that the expected quadratic error is reduced by the selection rule to the same order of magnitude as in a classical parametric problem. The method is applied on a real data set, illustrating that the new method describes the data set very well: the error involved in the classical Gaussian copula density is reduced with no fewer than 50%.     

Estimating value at risk of portfolio by conditional copula-GARCH method

Copula functions represent a methodology that describes the dependence structure of a multi-dimension random variable and has become one of the most significant new tools to handle risk factors in finance, such as Value-at Risk (VaR), which is probably the most widely used risk measure in financial institutions. Combining copula and the forecast function of the GARCH model, this paper proposes a new method, called conditional copula-GARCH, to compute the VaR of portfolios. This work presents an application of the copula-GARCH model in the estimation of a portfolio’s VaR, composed of NASDAQ and TAIEX. The empirical results show that, compared with traditional methods, the copula model captures the VaR more successfully. In addition, the Student-t copula describes the dependence structure of the portfolio return series quite well.     

包含股指期货的投资组合之风险研究——Copula方法在金融风险管理中的应用

本文运用Copula方法研究了含股指期货的投资组合的风险度量问题.由于股指期货和股票现货之间存在很大的相关性,因此在度量组合的风险时,各资产间的相关结构起到了关键作用,但这一相关结构很难用线性的相关系数去刻画,本文采用Copula模型来描述相关结构。而后,我们构建了基于Copula理论的风险度量指标PVaR,并验证了不同Copula模型的拟合效果.我们利用沪深300指数的数据来研究股指期货和现货的相关结构,并使用了多种Copula函数结合不同的边际分布假设进行了模拟,说明了Copula方法在风险度量尤其是包含了股指期货的投资组合的风险度量上具有较高的精确性.     

含状态转换的Copula模型及其应用

Copula在金融资产风险分析中被广泛应用,本文考虑了一种含状态转换的Copula模型。该模型的特点是,Copula函数在一个状态内是固定的,在不同状态之间的转移过程是一个不可观测的马氏链。其不同的状态可以用来描述金融资产相关性的波动。我们运用该模型对我国股票市场进行了实证研究,结果表明该模型可以更好地衡量其相关性模式。     

FGM Copula的另一种拓展

研究形如C(u,v)=uv+ψ(v)u(1-u),ψ(v)为v的非多项式函数的三类广义FGM Copula,计算它们的线性相关系数r,和谐性度量Kendall'sτ和Spearman'sρ.结果显示其中的两类广义FGM Copula与FGM Copula的性质接近,但是其τ与ρ的取值范围在某种程度上比FGM Copula的大,所以应用范围更广,适应性更强.另一类的线性相关系数r为0,但是τ和ρ均不为零,说明和谐性度量能够描述变量间非线性的相关关系.     

A scalar product for copulas

We introduce a scalar product for n-dimensional copulas, based on the Sobolev scalar product for W1,2-functions. The corresponding norm has quite remarkable properties and provides a new, geometric framework for copulas. We show that, in the bivariate case, it measures invertibility properties of copulas with respect to the ∗-operation introduced by Darsow et al. (1992). The unique copula of minimal norm is the null element for the ∗-operation, whereas the copulas of maximal norm are precisely the invertible elements.     

沪深股市收益率的尾部相关函数

尾部相关性是相关性分析中重要的一类,利用度量尾部相关性的指标χ,χ-以及尾部相关函数ρ(θ)来分析尾部相关性,并给出ρ(θ)的一种非参数估计方法.通过这两种方法研究上证综合指数和深证成分指数日收盘指数对数收益率在损失情况下的尾部相关性,结果表明两市指数日对数收益率具有很强的尾部相关性.     

A topological proof of Sklar’s theorem

We present a proof of Sklar’s Theorem that uses topological arguments, namely compactness (under the weak topology) of the class of copulas and some density properties of the class of distribution functions.     

Asymptotic efficiency of the two-stage estimation method for copula-based models

For multivariate copula-based models for which maximum likelihood is computationally difficult, a two-stage estimation procedure has been proposed previously; the first stage involves maximum likelihood from univariate margins, and the second stage involves maximum likelihood of the dependence parameters with the univariate parameters held fixed from the first stage. Using the theory of inference functions, a partitioned matrix in a form amenable to analysis is obtained for the asymptotic covariance matrix of the two-stage estimator. The asymptotic relative efficiency of the two-stage estimation procedure compared with maximum likelihood estimation is studied. Analysis of the limiting cases of the independence copula and Fréchet upper bound help to determine common patterns in the efficiency as the dependence in the model increases. For the Fréchet upper bound, the two-stage estimation procedure can sometimes be equivalent to maximum likelihood estimation for the univariate parameters. Numerical results are shown for some models, including multivariate ordinal probit and bivariate extreme value distributions, to indicate the typical level of asymptotic efficiency for discrete and continuous data.     

Archimedean Copula数据拟合

给出了选择较优Archimedean Copula相依结构的一般过程,并结合中国股市的实际数据作了分析,通过不同的标准得到了拟合深圳成份A股与深圳成份B股指数的较好的Archimedean Copula,而且还发现利用Copula刻画相依结构比传统的线性相关系数具有更多的优越性.