Risk modelling for health insurance policies with modifications of the loss random variable

Risk modelling is of great importance for health insurance as it is for many other insurance branches. By using risk modelling and modelling the probable loss amounts and frequencies, the expected payment amount can be calculated for any portfolio for a given time period. By doing this, the premium calculations are run in a more secure atmosphere. In this thesis, first, some information about the past and current situation of health insurance in both Turkey and all over the World is given. Then, individual risk modelling and collective risk modelling approaches are studied and the loss random variable is described. In addition, the consequences of applying the modifications of the loss random variable such as limits, deductibles and coinsurance are observed. Finally, some commonly used frequency and severity distributions are explained in detail. In application, individual and collective risk modelling approaches are applied using the data of a private health benefit fund in Turkey for the year 2006. Following these, various limits and deductibles are applied to the policies and the effects of these on the loss random variable are investigated. Finally the present value of future health benefits is calculated and the analysis is concluded. Keywords: Health insurance, Individual risk modelling, Collective risk modelling, Limit, Deductible, Present value of future benefits