CUSUM test for general nonlinear integer-valued GARCH models: comparison study

This study considers the problem of testing a parameter change in general nonlinear integer-valued time series models where the conditional distribution of current observations is assumed to follow a one-parameter exponential family. We consider score-, (standardized) residual-, and estimate-based CUSUM tests and show that their limiting null distributions take the form of the functions of Brownian bridges. Based on the obtained results, we then conduct a comparison study of the performance of CUSUM tests through the use of Monte Carlo simulations. Our findings demonstrate that the standardized residual-based CUSUM test largely outperforms the others.

     

Structured additive regression for overdispersed and zero‐inflated count data

In count data regression there can be several problems that prevent the use of the standard Poisson log‐linear model: overdispersion, caused by unobserved heterogeneity or correlation, excess of zeros, non‐linear effects of continuous covariates or of time scales, and spatial effects. We develop Bayesian count data models that can deal with these issues simultaneously and within a unified inferential approach. Models for overdispersed or zero‐inflated data are combined with semiparametrically structured additive predictors, resulting in a rich class of count data regression models. Inference is fully Bayesian and is carried out by computationally efficient MCMC techniques. Simulation studies investigate performance, in particular how well different model components can be identified. Applications to patent data and to data from a car insurance illustrate the potential and, to some extent, limitations of our approach. Copyright © 2006 John Wiley & Sons, Ltd.     

Modeling carbon spot and futures price returns with GARCH and Markov switching GARCH models

This paper makes use of spot and futures market data to carry out a thorough analysis of the dynamics of carbon price returns in the European Union Emission Trading Scheme for the whole first commitment period from 2008 to 2012. Understanding the properties of carbon price returns is especially crucial for industries which have to comply with an emission trading system and other market participants such as risk managers and speculators. We therefore seek to develop accurate models which capture the behavior of carbon price returns comprehensively. We apply a broad spectrum of GARCH model specifications, using different distributions for model innovations. As both time series, spot and futures price returns, exhibit asymmetric behavior in their variance, we additionally take Markov regime switching models for the variance equation into consideration. Empirical results demonstrate that AGARCH, NARCH and GJR fit the data best. We further show that, in the error term of any model, fat-tailed distributions—in particular the generalized error distribution—significantly improve the fit. Additionally, as futures returns seem to carry informational content concerning subsequent spot returns, we propose a sound, yet parsimonious, spot returns model, well-suited to capturing the dynamics. Finally, the most appropriate models for spot and futures price returns are tested in an out-of-sample environment, and further checked for robustness in data subsets. Subsequently a model for each market is proposed.     

Multivariate count data generalized linear models: Three approaches based on the Sarmanov distribution

Starting from the question: What is the accident risk of an insured individual?, we consider that the customer has contracted policies in different insurance lines: motor and home. Three models based on the multivariate Sarmanov distribution are analyzed. Driven by a real data set that takes into account three types of accident risks, two for motor and one for home, three trivariate Sarmanov distributions with generalized linear models (GLMs) for marginals are considered and fitted to the data. To estimate the parameters of these three models, we discuss a method for approaching the maximum likelihood (ML) estimators. Finally, the three models are compared numerically with the simpler trivariate Negative Binomial GLM and with elliptical copula based models.     

Two-component Poisson mixture regression modelling of count data with bivariate random effects

Two-component Poisson mixture regression is typically used to model heterogeneous count outcomes that arise from two underlying sub-populations. Furthermore, a random component can be incorporated into the linear predictor to account for the clustering data structure. However, when including random effects in both components of the mixture model, the two random effects are often assumed to be independent for simplicity. A two-component Poisson mixture regression model with bivariate random effects is proposed to deal with the correlated situation. A restricted maximum quasi-likelihood estimation procedure is provided to obtain the parameter estimates of the model. A simulation study shows both fixed effects and variance component estimates perform well under different conditions. An application to childhood gastroenteritis data demonstrates the usefulness of the proposed methodology, and suggests that neglecting the inherent correlation between random effects may lead to incorrect inferences concerning the count outcomes.     

Factor models for multivariate count data

We develop a general class of factor-analytic models for the analysis of multivariate (truncated) count data. Dependencies in multivariate counts are of interest in many applications, but few approaches have been proposed for their analysis. Our model class allows for a variety of distributions of the factors in the exponential family. The proposed framework includes a large number of previously proposed factor and random effect models as special cases and leads to many new models that have not been considered so far. Whereas previously these models were proposed separately as different cases, our framework unifies these models and enables one to study them simultaneously. We estimate the Poisson factor models with the method of simulated maximum likelihood. A Monte-Carlo study investigates the performance of this approach in terms of estimation bias and precision. We illustrate the approach in an analysis of TV channels data.     

RCA models with GARCH innovations

Rapid developments of time series models and methods addressing volatility in computational finance and econometrics have been recently reported in the financial literature. The non-linear volatility theory either extends and complements existing time series methodology by introducing more general structures or provides an alternative framework (see Abraham and Thavaneswaran [B. Abraham, A. Thavaneswaran, A nonlinear time series model and estimation of missing observations, Ann. Inst. Statist. Math. 43 (1991) 493–504] and Granger [C.W.J. Granger, Overview of non-linear time series specification in Economics, Berkeley NSF-Symposia, 1998]). In this work, we consider Gaussian first-order linear autoregressive models with time varying volatility. General properties for process mean, variance and kurtosis are derived; examples illustrate the wide range of properties that can appear under the autoregressive assumptions. The results can be used in identifying some volatility models. The kurtosis of the classical RCA model of Nicholls and Quinn [D.F. Nicholls, B.G. Quinn, Random Coefficient Autoregressive Models: An Introduction, in: Lecture Notes in Statistics, vol. 11, Springer, New York, 1982] is shown to be a special case.     

Second-order generalized estimating equations for correlated count data

Generalized estimating equations have been widely used in the analysis of correlated count data. Solving these equations yields consistent parameter estimates while the variance of the estimates is obtained from a sandwich estimator, thereby ensuring that, even with misspecification of the so-called working correlation matrix, one can draw valid inferences on the marginal mean parameters. That they allow misspecification of the working correlation structure, though, implies a limitation of these equations should scientific interest also be in the covariance or correlation structure. We propose herein an extension of these estimating equations such that, by incorporating the bivariate Poisson distribution, the variance-covariance matrix of the response vector can be properly modelled, which would permit inference thereon. A sandwich estimator is used for the standard errors, ensuring sound inference on the parameters estimated. Two applications are presented.     

Doubly stochastic models with GARCH innovations

A rapid development of time series models and methods addressing volatility in computational finance and econometrics are recently reported in the financial literature. This paper considers doubly stochastic volatility models with GARCH errors. General properties for process mean, variance and kurtosis are derived as these results can be used in model identification.     

Hierarchical mixture models for zero-inflated correlated count data

Count data with excess zeros are often encountered in many medical, biomedical and public health applications. In this paper, an extension of zero-inflated Poisson mixed regression models is presented for dealing with multilevel data set, referred as hierarchical mixture zero-inflated Poisson mixed regression models. A stochastic EM algorithm is developed for obtaining the ML estimates of interested parameters and a model comparison is also considered for comparing models with different latent classes through BIC criterion. An application to the analysis of count data from a Shanghai Adolescence Fitness Survey and a simulation study illustrate the usefulness and effectiveness of our methodologies.     

Sample selection models for count data in R

We provide a detailed hands-on tutorial for the R package SemiParSampleSel (version 1.5). The package implements selection models for count responses fitted by penalized maximum likelihood estimation. The approach can deal with non-random sample selection, flexible covariate effects, heterogeneous selection mechanisms and varying distributional parameters. We provide an overview of the theoretical background and then demonstrate how SemiParSampleSel can be used to fit interpretable models of different complexity. We use data from the German Socio-Economic Panel survey (SOEP v28, 2012. doi: 10.5684/soep.v28) throughout the tutorial.     

A conditional-SGT-VaR approach with alternative GARCH models

This paper proposes a conditional technique for the estimation of VaR and expected shortfall measures based on the skewed generalized t (SGT) distribution. The estimation of the conditional mean and conditional variance of returns is based on ten popular variations of the GARCH model. The results indicate that the TS-GARCH and EGARCH models have the best overall performance. The remaining GARCH specifications, except in a few cases, produce acceptable results. An unconditional SGT-VaR performs well on an in-sample evaluation and fails the tests on an out-of-sample evaluation. The latter indicates the need to incorporate time-varying mean and volatility estimates in the computation of VaR and expected shortfall measures.     

Large deviations for generalized compound Poisson risk models and its bankruptcy moments

We extend the classical compound Poisson risk model to the case where the premium income process, based on a Poisson process, is no longer a linear function. For this more realistic risk model, Lundberg type limiting results on the finite time ruin probabilities are derived. Asymptotic behaviour of the tail probabilities of the claim surplus process is also investigated.     

Modeling loss data using composite models

We develop several new composite models based on the Weibull distribution for heavy tailed insurance loss data. The composite model assumes different weighted distributions for the head and tail of the distribution and several such models have been introduced in the literature for modeling insurance loss data. For each model proposed in this paper, we specify two parameters as a function of the remaining parameters. These models are fitted to two real insurance loss data sets and their goodness-of-fit is tested. We also present an application to risk measurements and compare the suitability of the models to empirical results.     

Optimal decorrelated score subsampling for generalized linear models with massive data

In this paper, we consider the unified optimal subsampling estimation and inference on the lowdimensional parameter of main interest in the presence of the nuisance parameter for low/high-dimensional generalized linear models(GLMs) with massive data. We first present a general subsampling decorrelated score function to reduce the influence of the less accurate nuisance parameter estimation with the slow convergence rate. The consistency and asymptotic normality of the resultant subsample estimat...     

Modeling frequency and severity of claims with the zero-inflated generalized cluster-weighted models

To facilitate applications in general insurance, some extensions are proposed to cluster-weighted models (CWMs). First, we extend CWMs to have generalized cluster-weighted models (GCWMs) by allowing modeling of non-Gaussian distribution of the continuous covariates, as they frequently occur in insurance practice. Secondly, we introduce a zero-inflated extension of GCWM (ZI-GCWM) for modeling insurance claims data with excess zeros coming from heterogeneous sources. Additionally, we give two expectation–optimization (EM) algorithms for parameter estimation given in the proposed models. An appropriate simulation study shows that, for various settings and in contrast to the existing mixture-based approaches, both extended models perform well. Finally, a real data set based on French auto-mobile policies is used to illustrate the application of the proposed extensions.     

Comparison of control charts for Poisson count data in health‐care monitoring

Statistical surveillance is a noteworthy endeavor in many health‐care areas such as epidemiology, hospital quality, infection control, and patient safety. For monitoring hospital adverse events, the Shewhart u‐control chart is the most used methodology. One possible issue of the u‐chart is that in health‐care applications the lower control limit (LCL) is often conventionally set to zero as the adverse events are rare and the sample sizes are not sufficiently large to obtain LCL greater than zero. Consequently, the control chart loses any ability to signal improvements. Furthermore, as the area of opportunity (sample size) is not constant over time, the in‐control and out‐of‐control run length performances of the monitoring scheme are unknown. In this article, on the basis of a real case and through an intensive simulation study, we first investigate the in‐control statistical properties of the u‐chart. Then we set up several alternative monitoring schemes with the same in‐control performances and their out‐of‐control properties are studied and compared. The aim is to identify the most suitable control chart considering jointly: the ability to detect unexpected changes (usually worsening), the ability to test the impact of interventions (usually improvements), and the ease of use and clarity of interpretation. The results indicate that the exponentially weighted moving average control chart derived under the framework of weighted likelihood ratio test has the best overall performance.     

Modeling heterogeneity for bivariate survival data by the compound Poisson distribution with random scale

We propose a bivariate Weibull regression model with heterogeneity (frailty or random effect) which is generated by compound Poisson distribution with random scale. We assume that the bivariate survival data follow bivariate Weibull of Hanagal (2004). There are some interesting situations like survival times in genetic epidemiology, dental implants of patients and twin births (both monozygotic and dizygotic) where genetic behavior (which is unknown and random) of patients follows a known frailty distribution. These are the situations which motivate us to study this particular model. We propose a two stage maximum likelihood estimation procedure for the parameters in the proposed model and develop large sample tests for testing significance of regression parameters.     

散度偏大计数数据回归模型的变量选择与模型比较

讨论了具有散度偏大特征计数数据的建模与拟合问题.针对导致数据散度偏大的原因和常用的几类候选模型的结构,分别给出了关于嵌套模型的模型与变量同时选择的Bayes方法和关于非嵌套模型的模型检验与比较方法,并在此基础上进一步完善,提出了较为系统完整的模型与变量选择方法.实际例子说明了方法的具体实现过程和有效性.     

Inferences in semi-parametric dynamic mixed models for longitudinal count data

This paper considers a semi-parametric mixed model for longitudinal counts under the assumption that for conditional on a common random effect over time the repeated count responses of an individual follow a Poisson AR(1) (auto-regressive order 1) non-stationary correlation structure. A step-by-step estimation approach is developed which provides consistent estimators for the non-parametric function, regression parameters, variance of the random effects, and auto-correlation structure of the model. Proofs for the consistency properties of the estimators along with their convergence rates are derived. A simulation study is conducted to examine first the estimation effects on parameters when the non-parametric function is ignored, and then an overall estimation study is carried out in the presence of the non-parametric function by including its estimation as well.