Document Type : Original Research Paper

Authors

• A. Omrani ¹
• M.R. Faghihi Habibabadi ²

¹ Department of Actuarial Science, Shahid Beheshti University, Tehran, Iran

² Department of Statistics, Faculty of Mathematical Sciences, Shahid Beheshti University, Tehran, Iran

Abstract

In their financial statements, insurance companies often use the chain ladder method to forecast claims reserves. The chain ladder method is based on accumulated data and years of claims development in the triangle of future obligations. This triangle is a summary of the datasets for individual claims. In this paper, the framework of state-dependent signed Poisson process, and statistical tools for recurrence events in single claims are used for a method of storage under the title of small-level random loss storage. Details of the time of claim occurrence, time of delay in claim reporting, times between payments and amounts of payments made, and information on the time of final settlement of claims are used in calculating the micro level reserve. To evaluate the new model, the data set of an Iranian insurance company has been considered; By using these data sets and simulating the damage reserve with the micro level model, it was shown that the use of the small level random damage reserve model has a close estimate to the actual amount of the required loss reserve for the coming years.

Keywords

• Loss Reserve‎
• Poisson Process
• ‎‎Survival Analysis‎
• ‎Non-Life Insurance‎
• ‎Predicting‎
• ‎Micro-Level