多元时间序列模型协方差矩阵的递归算法

本推导了多元时序横型的协方差矩阵与模型参数的关系式，并给出了计算多维时序过程自协方差矩阵的递归算法。     

基于因子收缩方法的高维协方差估计

高维协方差矩阵在经济、金融、生物等众多领域中有着广泛应用.基于收缩估计模型,构造样本协方差矩阵与因子模型协方差矩阵的凸线性组合,通过对因子模型的改进来提高模型估计精度.在构造因子模型时,引入因子选择准则(pc_p3(k))来确定因子个数:在确定最优权重α时,使用基于MSE(S)分解的思想求解.通过数据验证发现,相较于传统方法,提升了协方差矩阵估计精确性;在构造投资组合模型时,也可以有效降低投资风险.     

协方差矩阵扰动生长曲线模型岭估计的影响分析

本文研究了协方差矩阵发生扰动时，生长曲线模型岭估计的影响分析，建立了生长曲线模型、协方差矩阵扰动生长曲线模型岭估计之间的关系式。讨论了协方差扰动和数据删除对岭估计的影响，导出了度量影响大小的基于岭估计的广义Cook距离Di^*（k）最后，用实例说明了用Di^*(k）度量生长曲线模型的影响点是有效的。     

协方差阵估计的比较

本文在一般线性回归模型误差异方差情况下，通过计算机模拟对回归系数最小二乘估计的协方差矩阵的估计进行了比较。结果表明，当样本大小大于50时，回归系数的最小二乘估计具有较高的估计精度；其协方差矩阵的五种估计以普通最小二乘估计的协方差矩阵为最优。     

面板数据交互固定效应模型的协方差矩阵检验

本文研究了面板数据交互固定效应模型中协方差矩阵的检验问题.首先依据模型协方差矩阵迹的估计构造检验统计量,检验协方差矩阵是否为单位矩阵,或是单位矩阵的常数倍.然后在一定正则条件下,证明了检验统计量的渐近性质,并说明所提出的检验方法不依赖于误差分布.最后,通过模拟研究对本文的检验方法进行评价,说明所提检验方法在高维面板数据下仍然有效.     

增长曲线模型中一致最小风险无偏估计的存在性

考虑协方差阵任意，或具有均匀协方差结构，或具有序列协方差结构的正态增长曲线模型本文将文[19]在设计矩阵满秩，且仅估计回归系数矩阵的情形获得的结果推广到设计矩阵不必列满秩，且同时估计回归系数矩阵的线性可估函数和协方差阵（或有关参数）的情形；在凸损失函数类和矩阵损失函数下，给出存在一致最小风险无偏估计的充分必要条件．     

可分图精度矩阵的精确分布函数

本文给出了从可分图协方差矩阵的分布密度函数确定图精度矩阵分布密度函数的一般方法，得到了可分的Gaussian图模型中精度矩阵极大似然估计的分布密度函数表达式．当图协方差矩阵的分布密度分别服从超逆Wishart分布、超逆Г分布时，也得到了图精度矩阵分布密度函数的解析表达式．     

多元波动率模型的一些新进展

对多个资产收益率的协方差矩阵建立动态模型是一个非常重要的问题。本文就近些年来该方面研究的一些主要进展进行了综述,特别地介绍了几种基于数据降维技术发展起来的能够适用于高维情形的多元GARCH模型,另外,对于多元波动率的模型诊断与比较方法以及条件协方差矩阵的预测等方面的研究成果也作了分析。     

一类多元重复测量模型参数的似然比检验及其功效分析

在多元重复测量试验模型下,当受试对象观测矩阵的协方差矩阵∑为等方差等协方差结构时,给出了参数的似然比检验统计量.给出该检验在原假设下的渐近零分布和在备择假设下的渐近非零分布,并就检验的功效进行了分析.     

矩阵损失函数下回归系数矩阵的非齐次线性估计的可容许性

在二次矩阵损失函数下研究了协方差矩阵未知的多元线性模型中回归系数矩阵的可估线性函数的矩阵非齐次线性估计的可容许性,给出了矩阵非齐次线性估计在线性估计类中可容许的一个充要条件.     

Spectrum-Based Comparison of Stationary Multivariate Time Series

The problem of comparison of several multivariate time series via their spectral properties is discussed. A pairwise comparison between two independent multivariate stationary time series via a likelihood ratio test based on the estimated cross-spectra of the series yields a quasi-distance between the series. A hierarchical clustering algorithm is then employed to compare several time series given the quasi-distance matrix. For use in situations where components of the multivariate time series are measured in different units of scale, a modified quasi-distance based on a profile likelihood based estimation of the scale parameter is described. The approach is illustrated using simulated data and data on daily temperatures and precipitations at multiple locations. A comparison between hierarchical clustering based on the likelihood ratio test quasi-distance and a quasi-distance described in Kakizawa et al. (J Am Stat Assoc 93:328–340, 1998) is interesting.     

A type of matrix Padé approximant inspired by scalar component models

In this paper we define a type of matrix Padé approximant inspired by the identification stage of multivariate time series models considering scalar component models. Of course, the formalization of certain properties in the matrix Padé approximation framework can be applied to time series models and in other fields. Specifically, we want to study matrix Padé approximants as follows: to find rational representations (or rational approximations) of a matrix formal power series, with both matrix polynomials, numerator and denominator, satisfying three conditions: (a) minimum row degrees for the numerator and denominator, (b) an invertible denominator at the origin, and (c) canonical representation (without free parameters).     

The biplot graphic technique for the representation of multivariate time series

The paper applies the graphical technique known as biplot to the treatment of multivariate time series. The focus is on the analysis of a matrix whose observations form a time series and it is shown how the effect of time can be segregated. A dynamic biplot representation of a particular market is presented and some characteristics of Israel's banking system are revealed by the biplot.     

Canonical correlation for principal components of time series

With contemporary data collection capacity, data sets containing large numbers of different multivariate time series relating to a common entity (e.g., fMRI, financial stocks) are becoming more prevalent. One pervasive question is whether or not there are patterns or groups of series within the larger data set (e.g., disease patterns in brain scans, mining stocks may be internally similar but themselves may be distinct from banking stocks). There is a relatively large body of literature centered on clustering methods for univariate and multivariate time series, though most do not utilize the time dependencies inherent to time series. This paper develops an exploratory data methodology which in addition to the time dependencies, utilizes the dependency information between S series themselves as well as the dependency information between p variables within the series simultaneously while still retaining the distinctiveness of the two types of variables. This is achieved by combining the principles of both canonical correlation analysis and principal component analysis for time series to obtain a new type of covariance/correlation matrix for a principal component analysis to produce a so-called “principal component time series”. The results are illustrated on two data sets.     

Change points in heavy‐tailed multivariate time series: Methods using precision matrices

We propose two robust data‐driven techniques for detecting network structure change points between heavy‐tailed multivariate time series for situations where both the placement and number of change points are unknown. The first technique utilizes the graphical lasso method to estimate the change points, whereas the second technique utilizes the tlasso method. The techniques not only locate the change points but also estimate an undirected graph (or precision matrix) representing the relationship between the time series within each interval created by pairs of adjacent change points. An inference procedure on the edges is used in the graphs to effectively remove false‐positive edges, which are caused by the data deviating from normality. The techniques are compared using simulated multivariate t‐distributed (heavy‐tailed) time series data and the best method is applied to two financial returns data sets of stocks and indices. The results illustrate the method's ability to determine how the dependence structure of the returns changes over time. This information could potentially be used as a tool for portfolio optimization.     

Testing nonparametric and semiparametric hypotheses in vector stationary processes

We propose a general nonparametric approach for testing hypotheses about the spectral density matrix of multivariate stationary time series based on estimating the integrated deviation from the null hypothesis. This approach covers many important examples from interrelation analysis such as tests for noncorrelation or partial noncorrelation. Based on a central limit theorem for integrated quadratic functionals of the spectral matrix, we derive asymptotic normality of a suitably standardized version of the test statistic under the null hypothesis and under fixed as well as under sequences of local alternatives. The results are extended to cover also parametric and semiparametric hypotheses about spectral density matrices, which includes as examples goodness-of-fit tests and tests for separability.     

Graphical modelling of multivariate time series

We introduce graphical time series models for the analysis of dynamic relationships among variables in multivariate time series. The modelling approach is based on the notion of strong Granger causality and can be applied to time series with non-linear dependences. The models are derived from ordinary time series models by imposing constraints that are encoded by mixed graphs. In these graphs each component series is represented by a single vertex and directed edges indicate possible Granger-causal relationships between variables while undirected edges are used to map the contemporaneous dependence structure. We introduce various notions of Granger-causal Markov properties and discuss the relationships among them and to other Markov properties that can be applied in this context. Examples for graphical time series models include nonlinear autoregressive models and multivariate ARCH models.     

Gauss inequalities on ordered linear spaces

Markov inequalities on ordered linear spaces are tightened through the α-unimodality of the corresponding measures. Modality indices are studied for various induced measures, including the singular values of a random matrix and the periodogram of a time series. These tools support a detailed study of linear inference and the ordering of random matrices, to include fixed and random designs and probability bounds on their comparative efficiencies. Other applications include probability bounds on quadratic forms and of order statistics on Rⁿ, on periodograms in the analysis of time series, and on run-length distributions in multivariate statistical process control. Connections to other topics in applied probability and statistics are noted.     

A Case Study of the Residual-Based Cointegration Procedure

The study of long-run equilibrium processes is a significant component of economic and finance theory. The Johansen technique for identifying the existence of such long-run stationary equilibrium conditions among financial time series allows the identification of all potential linearly independent cointegrating vectors within a given system of eligible financial time series. The practical application of the technique may be restricted, however, by the pre-condition that the underlying data generating process fits a finite-order vector autoregression (VAR) model with white noise. This paper studies an alternative method for determining cointegrating relationships without such a precondition. The method is simple to implement through commonly available statistical packages. This 'residual-based cointegration' (RBC) technique uses the relationship between cointegration and univariate Box-Jenkins ARIMA models to identify cointegrating vectors through the rank of the covariance matrix of the residual processes which result from the fitting of univariate ARIMA models. The RBC approach for identifying multivariate cointegrating vectors is explained and then demonstrated through simulated examples. The RBC and Johansen techniques are then both implemented using several real-life financial time series.     

Rationality,minimality and uniqueness of representation of matrix formal power series

In some multivariate time-series models a matrix power series is involved. These models can be identified as rational models if these series correspond to a matrix rational function. Moreover, it is necessary to answer some questions about minimality and uniqueness of representation. The main results of this paper fall within the sphere of matrix Padé approximation. On the basis of formal power series, matrix rational functions of arbitrary dimensions are characterized. Furthermore, we study certain minimality types, that are, global minimum degrees and row minimum degrees. In addition, given that the rational representation of the function for the same pair of degrees need not be unique, we have obtained conditions to study the uniqueness of said representation and, also, to find a “canonical” unique representation. Moreover, we consider an application to special series which is associated with time-series models; such series leads to new theoretical results relating to matrix Padé approximation. Finally, we comment on some illustrative examples.