Abstract
When pricing a specific insurance premium, actuary needs to evaluate the claims cost distribution for the warranty. Traditional actuarial methods use parametric specifications to model claims distribution, like lognormal, Weibull and Pareto laws. Mixtures of such distributions allow to improve the flexibility of the parametric approach and seem to be quite well-adapted to capture the skewness, the long tails as well as the unobserved heterogeneity among the claims. In this paper, instead of looking for a finely tuned mixture with many components, we choose a parsimonious mixture modeling, typically a two or three-component mixture. Next, we use the mixture cumulative distribution function (CDF) to transform data into the unit interval where we apply a beta-kernel smoothing procedure. A bandwidth rule adapted to our methodology is proposed. Finally, the beta-kernel density estimate is back-transformed to recover an estimate of the original claims density. The beta-kernel smoothing provides an automatic fine-tuning of the parsimonious mixture and thus avoids inference in more complex mixture models with many parameters. We investigate the empirical performance of the new method in the estimation of the quantiles with simulated nonnegative data and the quantiles of the individual claims distribution in a non-life insurance application.
Keywords
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Atienza, N., Garcia-Heras, J., & Muñoz-Pichardo, J. M. (2006). A new condition for identifiability of finite mixture distributions. Metrika, 63, 215–221.
Bolance, C., Guillen, M., & Nielsen, J. P. (2003). Kernel density estimation of acutarial loss functions. Insurance Mathematics and Economics, 32, 19–36.
Bouezmarni, T., & Scaillet, O. (2005). Consistency of asymetric kernel density estimators and smoothed histograms with application to income data. Econometric Theory, 21, 390–412.
Charpentier, A., & Oulidi, A. (2009). Beta kernel estimation for value-at-risk of heavy-tailed distributions. Computational Statistics (forthcoming).
Chen, S. X. (1999). Beta kernel estimators for density functions. Computational Statistics and Data Analysis, 31, 131–145.
Devroye, L., & Gyorfi, L. (1985). Nonparametric density estimation: The L1 view. New York: Wiley.
Gustafsson, J., Hagmann, M., Nielsen, J. P., & Scaillet, O. (2009). Local transformation kernel density estimation of loss distributions. Journal of Business and Economic Statistics, 27(2), 161–175.
Marron, J. S., & Ruppert, D. (1994). Transformations to reduce boundary bias in kernel density estimation. Journal of the Royal Statistical Society, Series B, 56, 653–671.
Mclachlan, G. J., & Peel, D. (2000). Finite mixture models. New York: Wiley.
Sarabia, J. M., Castillo, E., Gómez-Déniz, E., & Vázquez-Polo, F. J. (2005). A class of conjugate priors for log-normal claims based on conditional specification. The Journal of Risk and Insurance, 72, 479–495.
Silverman, B. W. (1994). Density estimation for statistics and data analysis. New York: Chapman & Hall.
Titterington, D. M., Smith, A. F., & Makov, U. E. (1985). Statistical analysis of mixture distributions. Wiley series in probability and mathematical statistics. New York: Wiley.
Wand, M. P., Marron, J. S., & Ruppert, D. (1991). Transformations in density estimation. Journal of the American Statistical Association, 86, 343–361.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Sardet, L., Patilea, V. (2009). Nonparametric Fine Tuning of Mixtures: Application to Non-Life Insurance Claims Distribution Estimation. In: Fink, A., Lausen, B., Seidel, W., Ultsch, A. (eds) Advances in Data Analysis, Data Handling and Business Intelligence. Studies in Classification, Data Analysis, and Knowledge Organization. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-01044-6_25
Download citation
DOI: https://doi.org/10.1007/978-3-642-01044-6_25
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-01043-9
Online ISBN: 978-3-642-01044-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)