动态尾部金融网络结构与传染性度量

基于Lasso高维分位数回归模型,本文构建了中国金融系统的尾部网络结构,并定义了总体网络以及行业间、行业内和金融机构间等多层金融网络的尾部风险传染度,解构其传染机制与关联特征,评估各机构双向系统重要性(接收端和发射端)。同时,本文提出了一个最优滚动窗宽选择标准方法,以优化滚动样本技术下的动态网络结构。结果表明,所有层级尾部风险传染效应(总体系统、行业间、行业内和机构间)在经济金融极端困境时期,呈现明显增强及剧烈震荡特征,2015年中国股灾期间尤甚。跨行业传染效应日益严峻,银行与保险间表现出较强关联性,房地产机构与其他金融机构间均表现出较高传染性,跨业监管值得关注。接收与发射最多尾部风险传染的金融机构仍然是银行与证券类机构。系统中超过50%的金融机构倾向于接收风险传染,一旦出现系统性冲击,整个金融系统的稳定性将遭受重创,因此,应加强此类机构应对外界冲击的能力。此外,基于新滚动窗宽选择标准法的动态模型的估计性能明显优于传统方法。研究结论有助于理解中国金融系统的网络结构和传染机制,对宏观审慎监管体系的建立提供了依据。     

Estimation and variable selection in single-index composite quantile regression

In this article, a new composite quantile regression estimation approach is proposed for estimating the parametric part of single-index model. We use local linear composite quantile regression (CQR) for estimating the nonparametric part of single-index model (SIM) when the error distribution is symmetrical. The weighted local linear CQR is proposed for estimating the nonparametric part of SIM when the error distribution is asymmetrical. Moreover, a new variable selection procedure is proposed for SIM. Under some regularity conditions, we establish the large sample properties of the proposed estimators. Simulation studies and a real data analysis are presented to illustrate the behavior of the proposed estimators.     

Approximation of IMSE-optimal Designs via Quadrature Rules and Spectral Decomposition

We address the problem of computing integrated mean-squared error (IMSE) optimal designs for interpolation of random fields with known mean and covariance. We assume that the mean squared error is integrated through a discrete measure and restrict the design space to its support. We show that the IMSE and its approximation by spectral truncation can be easily evaluated, which makes their global minimization affordable. Numerical experiments are carried out that illustrate the effectiveness of the approach.     

Optimized Sampling Strategies for Identifying Modes in Length-frequency Distributions

An approach to evaluate sampling strategies to detect modes in length-distributions is presented. Distributions based on various numbers of samples (S) and sample sizes (n) were simulated from the original cod and capelin data in Icelandic waters, incorporating within sample correlations. A peak was discerned if the difference between any simulated and original distribution did not exceed a pre-specified Δ, given a probability. This was achieved with numerous combinations of S and n, and the optimal choice will depend on the sampling costs. Variance-equivalence curves also illustrate the difference between demanding precision of mean lengths versus precision of length distributions.     

Problems in sampling the Native American and Alaska Native populations

Selecting a scientific national sample of Native Americans and Alaska Natives is difficult for at least four reasons: (1) they are a small proportion of the total population, (2) they are not so segregated that geographic oversampling can reach most of the population, (3) criteria for deciding who is a member of the Native American and Alaska Native population are not well defined, and (4) they are so culturally diverse that subclass estimates may be of equal or greater importance than overall estimates. The goal of this paper is to identify key design choices a statistician must face when selecting a sample of these populations, and to discuss alternative strategies. I also provide commentary on sampling plans that have been used for previous surveys, and explain why these plans, while sufficient for the purposes of their particular surveys, do not provide general solutions. Of particular importance is a careful definition of the study population. For example, some studies would focus on the reservation population, others would focus on persons listed on tribal rolls, and still others on persons eligible for coverage by the Indian Health Service. The numbers of persons included in all of these groups is far short of the two million Native Americans and Alaska Natives counted in the 1990 Census. I conclude that there is no sampling strategy that is appropriate for all surveys, but also provide suggestions for certain survey situations.     

Efficient construction of reversible jump Markov chain Monte Carlo proposal distributions

Summary. The major implementational problem for reversible jump Markov chain Monte Carlo methods is that there is commonly no natural way to choose jump proposals since there is no Euclidean structure in the parameter space to guide our choice. We consider mechanisms for guiding the choice of proposal. The first group of methods is based on an analysis of acceptance probabilities for jumps. Essentially, these methods involve a Taylor series expansion of the acceptance probability around certain canonical jumps and turn out to have close connections to Langevin algorithms. The second group of methods generalizes the reversible jump algorithm by using the so-called saturated space approach. These allow the chain to retain some degree of memory so that, when proposing to move from a smaller to a larger model, information is borrowed from the last time that the reverse move was performed. The main motivation for this paper is that, in complex problems, the probability that the Markov chain moves between such spaces may be prohibitively small, as the probability mass can be very thinly spread across the space. Therefore, finding reasonable jump proposals becomes extremely important. We illustrate the procedure by using several examples of reversible jump Markov chain Monte Carlo applications including the analysis of autoregressive time series, graphical Gaussian modelling and mixture modelling.     

Covariate selection for estimating the causal effect of control plans by using causal diagrams

Summary. Consider a case where cause–effect relationships between variables can be described by a causal path diagram and the corresponding linear structural equation model. The paper proposes a graphical selection criterion for covariates to estimate the causal effect of a control plan. For designing the control plan, it is essential to determine both covariates that are used for control and covariates that are used for identification. The selection of covariates used for control is only constrained by the requirement that the covariates be non-descendants of a treatment variable. However, the selection of covariates used for identification is dependent on the selection of covariates used for control and is not unique. In the paper, the difference between covariates that are used for identification is evaluated on the basis of the asymptotic variance of the estimated causal effect of an effective control plan. Furthermore, the results can be also described in terms of a graph structure.     

Optimal knowledge outsourcing model

Every organization controls its investments in the development and maintenance of internal knowledge (IK) as opposed to outsourcing this effort, namely, consuming external knowledge (EK). A number of factors involved in this decision, such as the IK learning curve, its associated holding cost, value deterioration rate, value of future IK or cost of purchasing EK. This study proposes a dynamic optimal control model for examining the properties of this problem. Optimal control strategies and steady-state conditions are identified for a number of special cases. Some insightful observations are obtained by studying the solution sensitivity to the underlying assumptions.     

Variance of the difference between two estimated responses

Minimization of the maximum and average variance of the difference between estimated responses are taken as design criteria for univariate polynomial regression models. An optimal design under the first criterion is derived for the second-order model and a class of designs nearly optimal under the second criterion is obtained for the general polynomial models.