A Bayesian Hierarchical Model with Spatial Variable Selection: The Effect of Weather on Insurance Claims

Scheel, Ida; Ferkingstad, Egil; Frigessi, Arnoldo; Haug, Ola; Hinnerichsen, Mikkel; Meze-Hausken, Elisabeth

doi:10.1111/j.1467-9876.2012.01039.x

Cited by 31 publications

(48 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The more recent studies from the Norwegian Computing Center replace the GLMs with Bayesian Poisson hurdle model, modify the list of predictors (omit linear trend and Fourier series, shorten the window for aggregated precipitation to 3 days), and eliminate long‐term forecasts based on climate models. In particular, Scheel et al () provide one‐week ahead forecasts (the period of 1 week is selected based on trustworthiness of the weather data used in such forecasts); Scheel and Hinnerichsen () build scenario forecasts by increasing the 2001 meteorological and hydrological covariates by 5, 18, or 30%.…”

Section: Home Insurancementioning

confidence: 99%

Insurance risk assessment in the face of climate change: Integrating data science and statistics

Lyubchich

Newlands

Ghahari

et al. 2019

WIREs Computational Stats

View full text Add to dashboard Cite

Local extreme weather events cause more insurance losses overall than large natural disasters. The evidence is provided by long‐term observations of weather and insurance records that are also a foundation for the majority of insurance products covering weather related damages. The insurers around the world are concerned, however, that the past records used to assess and price the risks underestimate the risk and incurred losses in recent years. The growing insurance risks are largely attributed to climate change that brings increasingly more alterations and permanent impact on all aspects of human life and welfare. From floods to hail to excessive wind, adverse atmospheric events are a poignant reminder of how vulnerable our society is across a broad range of threats posed by environmental extremes. Indeed, as climate change effects become more pronounced, we face a new era of risk with increasing weather related damages and losses. This in turn, coupled with challenges of massive climatic data, requires developing innovative analytic approaches that transcend traditional disciplinary boundaries of statistical, actuarial and environmental sciences. Nevertheless, the multidisciplinary nature of climate risk assessment and its impact on insurance is often overlooked and neglected. We highlight the most recent developments and interdisciplinary perspectives on diverse statistical and machine learning methodology for modeling and assessing climate risk in agricultural and home insurances, with a particular focus on noncatastrophic events. This article is categorized under: Applications of Computational Statistics > Computational Climate Change and Numerical Weather Forecasting Statistical and Graphical Methods of Data Analysis > Multivariate Analysis Data: Types and Structure > Massive Data

show abstract

Section: Home Insurancementioning

confidence: 99%

Insurance risk assessment in the face of climate change: Integrating data science and statistics

Lyubchich

Newlands

Ghahari

et al. 2019

WIREs Computational Stats

View full text Add to dashboard Cite

show abstract

“…Among such recent studies is the analysis of Norwegian house insurance dynamics by Haug et al (2011) and Scheel et al (2013) who develop a Bayesian hierarchical approach to explain insurance losses due to extreme weather events at a local geographic scale. Scheel et al (2013) consider only a leave-one-out type of prediction, e.g., using the data of 1996-2006, except those for 2001, and predicting the number of claims in 2001. Cheng et al (2012) propose a rainfall index and study the relationship between this index and insurance data.…”

Section: Motivationmentioning

confidence: 99%

Evaluating the Impact of Climate Change on Dynamics of House Insurance Claims

Soliman¹,

Lyubchich

Gel

et al. 2015

Machine Learning and Data Mining Approaches to Climate Science

View full text Add to dashboard Cite

The adverse effects of climate change bring increasingly more alterations to all aspects of human life and welfare, and one of the sectors that is particularly affected by changing climate is the insurance sector. Indeed, the year 2013 brought a record number of claims and substantial losses due to weather-related damages, and in the USA and Canada alone, the extreme weather events cost the insurance industry more than 3 billion dollars. The objective of this paper is to provide statistical data-driven insight on the (non)linear relationship between weatherrelated house insurance claims and atmospheric variables and to predict future claim dynamics accounting for changes in extreme precipitation. In this paper we propose to employ a flexible Generalized Autoregressive Moving Average (GARMA) model for count time series of claims, develop a new method to compare tails of the observed and projected extreme precipitation, and evaluate the impact of climate change on a number of house insurance claims in the GARMA framework. We illustrate our approach by studying insurance dynamics in four Canadian cities.

show abstract

“…A zero-truncated model, the so-called hurdle model, was proposed by Mullahy (1986) and also extended to complex data settings, e.g. by Scheel et al (2013). In the context of spatial count data, Agarwal et al (2002; 2006), Gschössl and Gzado (2008) and Ugarte et al (2004) used a zero-inflated spatial Poisson model.…”

Section: Introductionmentioning

confidence: 99%

Disease mapping of zero-excessive mesothelioma data in Flanders

Neyens

Lawson

Kirby

et al. 2017

Annals of Epidemiology

View full text Add to dashboard Cite

Purpose To investigate the distribution of mesothelioma in Flanders using Bayesian disease mapping models that account for both an excess of zeros and overdispersion. Methods The numbers of newly diagnosed mesothelioma cases within all Flemish municipalities between 1999 and 2008 were obtained from the Belgian Cancer Registry. To deal with overdispersion, zero-inflation and geographical association, the hurdle combined model was proposed, which has three components: a Bernoulli zero-inflation mixture component to account for excess zeros, a gamma random effect to adjust for overdispersion and a normal conditional autoregressive random effect to attribute spatial association. This model was compared with other existing methods in literature. Results The results indicate that hurdle models with a random effects term accounting for extra-variance in the Bernoulli zero-inflation component fit the data better than hurdle models that do not take overdispersion in the occurrence of zeros into account. Furthermore, traditional models that do not take into account excessive zeros but contain at least one random effects term that models extra-variance in the counts have better fits compared to their hurdle counterparts. In other words, the extra-variability, due to an excess of zeros, can be accommodated by spatially structured and/or unstructured random effects in a Poisson model such that the hurdle mixture model is not necessary. Conclusions Models taking into account zero-inflation do not always provide better fits to data with excessive zeros than less complex models. In this study, a simple conditional autoregressive model identified a cluster in mesothelioma cases near a former asbestos processing plant (Kapelle-op-den-Bos). This observation is likely linked with historical local asbestos exposures. Future research will clarify this.

show abstract

A Bayesian Hierarchical Model with Spatial Variable Selection: The Effect of Weather on Insurance Claims

Cited by 31 publications

References 37 publications

Insurance risk assessment in the face of climate change: Integrating data science and statistics

Insurance risk assessment in the face of climate change: Integrating data science and statistics

Evaluating the Impact of Climate Change on Dynamics of House Insurance Claims

Disease mapping of zero-excessive mesothelioma data in Flanders

Contact Info

Product

Resources

About