New framework reveals global warming’s impact on risk that multiple regions experience hot and dry conditions simultaneously.

Abstract A time‐varying risk analysis is proposed for an adaptive design framework in nonstationary conditions arising from climate change. Bayesian, dynamic conditional copula developed modeling the dependence structure between mixed continuous and discrete multiattributes of multidimensional hydrometeorological phenomena. Joint Bayesian inference carried out to fit marginals illustrative example using adaptive, Gibbs Markov Chain Monte Carlo (MCMC) sampler. Posterior mean estimates...

Abstract This survey reviews the existing literature on most relevant Bayesian inference methods for univariate and multivariate GARCH models. The advantages drawbacks of each procedure are outlined as well approach versus classical procedures. paper makes emphasis recent non‐parametric approaches models that avoid imposing arbitrary parametric distributional assumptions. These novel implicitly assume infinite mixture Gaussian distributions standardized returns which have been shown to be...

In a changing climate arising from anthropogenic global warming, the nature of extreme climatic events is over time. Existing analytical stationary-based risk methods, however, assume multi-dimensional phenomena will not significantly vary To strengthen reliability infrastructure designs and management water systems in environment, multidimensional stationary studies should be replaced with new adaptive perspective. The results comparison indicate that current frameworks are no longer...

Abstract A multivariate generalized autoregressive conditional heteroscedasticity model with dynamic correlations is proposed, in which the individual volatilities follow exponential models and standardized innovations a mixture of Gaussian distributions. Inference on parameters prediction future are addressed by both maximum likelihood Bayesian estimation methods. Estimation Value at Risk given portfolio selection optimal portfolios under proposed specification addressed. The good...

This article designs a Sequential Monte Carlo (SMC) algorithm for estimation of Bayesian semi-parametric Stochastic Volatility model financial data. In particular, it makes use one the most recent particle filters called Particle Learning (PL). SMC methods are especially well suited state-space models and can be seen as cost-efficient alternative to Markov Chain (MCMC), since they allow online type inference. The posterior distributions updated new data is observed, which exceedingly costly...

Summary This paper describes a Bayesian approach to make inference for risk reserve processes with an unknown claim‐size distribution. A flexible model based on mixtures of Erlang distributions is proposed approximate the special features frequently observed in insurance claim sizes, such as long tails and heterogeneity. density estimation sizes implemented using reversible jump Markov chain Monte Carlo methods. An advantage considered mixture that it belongs class phase‐type distributions,...

Abstract In this paper, we consider Bayesian inference and estimation of finite time ruin probabilities for the Sparre Andersen risk model. The dense family Coxian distributions is considered approximation both inter‐claim claim size distributions. We illustrate that model can be well fitted to real, long‐tailed claims data compares with generalized Pareto main advantage using times sizes it possible compute making use recent results from queueing theory. practice, are much more useful than...

This paper describes a nonparametric approach to make inferences for aggregate loss models in the insurance framework. We assume that an company provides historical sample of claims given by claim occurrence times and sizes. Furthermore, information may be incomplete as censored and/or truncated. In this context, main goal work consists fitting probability model total amount will paid on all during fixed future time period. order solve prediction problem, we propose new methodology based...

Recently, the field of multiple hypothesis testing has experienced a great expansion, basically because new methods developed in genomics. These allow scientists to simultaneously process thousands tests. The frequentist approach this problem is made by using different error measures that control Type I rate at certain desired level. Alternatively, article, Bayesian hierarchical model based on mixture distributions and an empirical Bayes are proposed order produce list rejected hypotheses...

To account for asymmetric dependence in extreme events, we propose a dynamic generalized hyperbolic skew Student-t factor copula where the loadings follow autoregressive score processes. Conditioning on latent factor, components of return series become independent, which allows us to run Bayesian estimation parallel setting. Hence, inference different specifications one models can be done few minutes. Finally, illustrate performance our proposed returns 140 companies listed S&P500 index. We...