Quantile regression modeling for Malaysian automobile insurance premium data

Mohd Fadzli Mohd Fuzi, Noriszura Ismail, Abdul Aziz Jemain

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Quantile regression is a robust regression to outliers compared to mean regression models. Traditional mean regression models like Generalized Linear Model (GLM) are not able to capture the entire distribution of premium data. In this paper we demonstrate how a quantile regression approach can be used to model net premium data to study the effects of change in the estimates of regression parameters (rating classes) on the magnitude of response variable (pure premium). We then compare the results of quantile regression model with Gamma regression model. The results from quantile regression show that some rating classes increase as quantile increases and some decrease with decreasing quantile. Further, we found that the confidence interval of median regression (τ = O.5) is always smaller than Gamma regression in all risk factors.

Original languageEnglish
Title of host publication2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium
PublisherAmerican Institute of Physics Inc.
Volume1678
ISBN (Electronic)9780735413252
DOIs
Publication statusPublished - 25 Sep 2015
Event2015 Postgraduate Colloquium of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology, UKM FST 2015 - Selangor, Malaysia
Duration: 15 Apr 201516 Apr 2015

Other

Other2015 Postgraduate Colloquium of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology, UKM FST 2015
CountryMalaysia
CitySelangor
Period15/4/1516/4/15

Fingerprint

quantiles
automobiles
regression analysis
ratings
confidence

Keywords

  • Automobile Insurance
  • Premium
  • Quantile Regression
  • Ratemaking

ASJC Scopus subject areas

  • Physics and Astronomy(all)

Cite this

Fuzi, M. F. M., Ismail, N., & Jemain, A. A. (2015). Quantile regression modeling for Malaysian automobile insurance premium data. In 2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium (Vol. 1678). [060014] American Institute of Physics Inc.. https://doi.org/10.1063/1.4931341

Quantile regression modeling for Malaysian automobile insurance premium data. / Fuzi, Mohd Fadzli Mohd; Ismail, Noriszura; Jemain, Abdul Aziz.

2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium. Vol. 1678 American Institute of Physics Inc., 2015. 060014.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Fuzi, MFM, Ismail, N & Jemain, AA 2015, Quantile regression modeling for Malaysian automobile insurance premium data. in 2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium. vol. 1678, 060014, American Institute of Physics Inc., 2015 Postgraduate Colloquium of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology, UKM FST 2015, Selangor, Malaysia, 15/4/15. https://doi.org/10.1063/1.4931341
Fuzi MFM, Ismail N, Jemain AA. Quantile regression modeling for Malaysian automobile insurance premium data. In 2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium. Vol. 1678. American Institute of Physics Inc. 2015. 060014 https://doi.org/10.1063/1.4931341
Fuzi, Mohd Fadzli Mohd ; Ismail, Noriszura ; Jemain, Abdul Aziz. / Quantile regression modeling for Malaysian automobile insurance premium data. 2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium. Vol. 1678 American Institute of Physics Inc., 2015.
@inproceedings{4fe9169bd88d48b5b605e5a050c6c2fb,
title = "Quantile regression modeling for Malaysian automobile insurance premium data",
abstract = "Quantile regression is a robust regression to outliers compared to mean regression models. Traditional mean regression models like Generalized Linear Model (GLM) are not able to capture the entire distribution of premium data. In this paper we demonstrate how a quantile regression approach can be used to model net premium data to study the effects of change in the estimates of regression parameters (rating classes) on the magnitude of response variable (pure premium). We then compare the results of quantile regression model with Gamma regression model. The results from quantile regression show that some rating classes increase as quantile increases and some decrease with decreasing quantile. Further, we found that the confidence interval of median regression (τ = O.5) is always smaller than Gamma regression in all risk factors.",
keywords = "Automobile Insurance, Premium, Quantile Regression, Ratemaking",
author = "Fuzi, {Mohd Fadzli Mohd} and Noriszura Ismail and Jemain, {Abdul Aziz}",
year = "2015",
month = "9",
day = "25",
doi = "10.1063/1.4931341",
language = "English",
volume = "1678",
booktitle = "2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium",
publisher = "American Institute of Physics Inc.",

}

TY - GEN

T1 - Quantile regression modeling for Malaysian automobile insurance premium data

AU - Fuzi, Mohd Fadzli Mohd

AU - Ismail, Noriszura

AU - Jemain, Abdul Aziz

PY - 2015/9/25

Y1 - 2015/9/25

N2 - Quantile regression is a robust regression to outliers compared to mean regression models. Traditional mean regression models like Generalized Linear Model (GLM) are not able to capture the entire distribution of premium data. In this paper we demonstrate how a quantile regression approach can be used to model net premium data to study the effects of change in the estimates of regression parameters (rating classes) on the magnitude of response variable (pure premium). We then compare the results of quantile regression model with Gamma regression model. The results from quantile regression show that some rating classes increase as quantile increases and some decrease with decreasing quantile. Further, we found that the confidence interval of median regression (τ = O.5) is always smaller than Gamma regression in all risk factors.

AB - Quantile regression is a robust regression to outliers compared to mean regression models. Traditional mean regression models like Generalized Linear Model (GLM) are not able to capture the entire distribution of premium data. In this paper we demonstrate how a quantile regression approach can be used to model net premium data to study the effects of change in the estimates of regression parameters (rating classes) on the magnitude of response variable (pure premium). We then compare the results of quantile regression model with Gamma regression model. The results from quantile regression show that some rating classes increase as quantile increases and some decrease with decreasing quantile. Further, we found that the confidence interval of median regression (τ = O.5) is always smaller than Gamma regression in all risk factors.

KW - Automobile Insurance

KW - Premium

KW - Quantile Regression

KW - Ratemaking

UR - http://www.scopus.com/inward/record.url?scp=85006197390&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006197390&partnerID=8YFLogxK

U2 - 10.1063/1.4931341

DO - 10.1063/1.4931341

M3 - Conference contribution

VL - 1678

BT - 2015 UKM FST Postgraduate Colloquium: Proceedings of the Universiti Kebangsaan Malaysia, Faculty of Science and Technology 2015 Postgraduate Colloquium

PB - American Institute of Physics Inc.

ER -