Eskişehir Teknik Üniversitesi Bilim ve Teknoloji Dergisi B - Teorik Bilimler 2667-419X Eskisehir Technical University 10.20290/estubtdb.609469 SOME COMPOSITE MODELS AND THEIR APPLICATION TO TURKISH MOTOR INSURANCE DATA SOME COMPOSITE MODELS AND THEIR APPLICATION TO TURKISH MOTOR INSURANCE DATA https://orcid.org/0000-0001-8180-995X Yiğiter Ayten HACETTEPE ÜNİVERSİTESİ https://orcid.org/0000-0002-8916-8509 Gençtürk Yasemin HACETTEPE ÜNİVERSİTESİ https://orcid.org/0000-0002-8537-513X Hamurkaroğlu Canan KARABÜK ÜNİVERSİTESİ 08 23 2019 7 2 121 132 07 23 2018 02 23 2019 Copyright © 2010, Eskişehir Teknik Üniversitesi Bilim ve Teknoloji Dergisi B - Teorik Bilimler 2010 Eskişehir Teknik Üniversitesi Bilim ve Teknoloji Dergisi B - Teorik Bilimler

It is important for aninsurance company to predict the future claims in order to evaluate premiums,to determine the reserve necessary to meet its obligation and probabilities ofruin, etc. as the claim data is highly positively skewed and has heavy tail, nostandard parametric model seems to provide an acceptable fit to both small andlarge losses. Composite models that use one standard distribution up to athreshold and other standard distribution thereafter are developed and it is seenthat these composite models provide better fit than the standard models whenclaim data involve small and high claims.  The aim of this study is toinvestigate the use of the composite models namely Exponential-Pareto,Weibull-Pareto and Lognormal-Pareto to model the Turkish Motor Insurance claimdata. From the results obtained, it is concluded that the compositeWeibull-Pareto model provides better fit to Turkish Motor Insurance claim datathan the all other models considered.

It is important for aninsurance company to predict the future claims in order to evaluate premiums,to determine the reserve necessary to meet its obligation and probabilities ofruin, etc. as the claim data is highly positively skewed and has heavy tail, nostandard parametric model seems to provide an acceptable fit to both small andlarge losses. Composite models that use one standard distribution up to athreshold and other standard distribution thereafter are developed and it is seenthat these composite models provide better fit than the standard models whenclaim data involve small and high claims.  The aim of this study is toinvestigate the use of the composite models namely Exponential-Pareto,Weibull-Pareto and Lognormal-Pareto to model the Turkish Motor Insurance claimdata. From the results obtained, it is concluded that the compositeWeibull-Pareto model provides better fit to Turkish Motor Insurance claim datathan the all other models considered.

 Composite models Standard parametric models Turkish motor insurance loss data Composite models Standard parametric models Turkish motor insurance loss data [1] Cooray K, Ananda MMA. Modeling Actuarial Data with a Composite Lognormal-Pareto Model. Scandinavian Actuarial Journal 2005; 5: 321-334. [2] Ciumara R. An Actuarial Model Based on Composite Weibull-Pareto Distribution. Mathematical Report-Bucharest 2006; 8 (4): 401-411. [3] Preda V, Ciumara R. On Composite Models: Weibull-Pareto and Lognormal-Pareto: A Comparative Study. Romanian Journal of Economic Forecasting 2006; 2: 32-46. [4] Teodorescu S, Vernic R. A Composite Exponential-Pareto Distribution. An. Şt. Univ. Ovidius Constanta 2006; 14(1): 99-108. [5] Scollnik DPM. On Composite Lognormal-ParetoModels. Scandinavian Actuarial Journal 2007; 1: 20-33. [6] Teodorescu S, Vernic R. Some Composite Exponential-Pareto Models for Actuarial Prediction. Romanian Journal of Economic Forecasting 2009; 4: 82-100. [7] Vernic R, Teodorescu S, Pelican E. Two Lognormal Models for Real Data. An. Şt. Univ. Ovidius Constanta 2009; 17(3): 263-279. [8] Nadarajah S, Bakar SAA. New composite Models for the Danish Fire Insurance Data. Scandinavian Actuarial Journal 2014; 2014(2): 180-187. [9] Maghsoudi M, Abu Bakar S. A, Hamzah NA. Composite Weibull-Inverse Transformed Gamma Distribution and Its Actuarial Application, Proceedings of the 21st National Symposium on Mathematical Sciences 2014. [10] Calderin-Ojeda E, Kwok CF. Modelling Large Claims with Composite Stoppa Models. Scandinavian Actuarial Journal 2015, Available from http://dx.doi.org/10.1080/03461238.2015.1034763. [11] Abu Bakar SA, Hamzah NA, Maghsoudi M. Nadarajah S. Modelling Loss Data Using Composite Models. Insurance: Mathematics and Economics 2015; 61: 146-154. [12] Klugman SS, Panjer HH, Willmot GE. Loss Models From Data to Decisions, New York: John Wiley&Sons, Inc., 1998.