Thesis Type: Postgraduate
Institution Of The Thesis: Gazi Üniversitesi, Fen Bilimleri Enstitüsü, Turkey
Approval Date: 2019
Thesis Language: Turkish
Student: FEYZA ELİF AKKUŞ
Supervisor: MERAL EBEGİL
Open Archive Collection: AVESIS Open Access Collection
Abstract:In actuarial applications, insurance companies are required to determine premiums for insurance contracts. When determining premiums, it should be taken into consideration that the insured people in the portfolio have different past experiences. If a common premium is determined without taking into account the individual experience of the insured, this premium will not be fair. Because there are insured people in the portfolio who are in better and worse condition than this prime. Therefore, the individual experiences of the insured and the experiences of other insured persons in the portfolio should be given specific weighting and a premium value should be determined for both the insurer and the insured. Such premium value can be determined by weighting using credibility theory. In the credibility theory, weighting is done with the Z credibility factor. There are several methods for determining the credibility factor. These methods are called credibility models. Where the credibility factor takes a value of zero, it means that the premium is calculated according to the general average of the insured without taking into account the individual experience of the insured person, and this is the case where the credibility models do not work well. In this study, firstly credibility theory is mentioned. Then, the credibility models are and preliminary information is given for the simulation study which will be done in the following section. In the light of this preliminary information, in order to determine a fair premium in different cases in the Bayesian approach and Bühlmann-Straub credibility models, the values of Z credibility factor were examined by Monte Carlo simulation. In cases where both methods did not work well, the results were interpreted