Linear Prediction Model for RBNS

The traditional claims reserving approaches are all based on the aggregated data and usually produce inaccurate predictions of the reserve because the aggregated data usually failed to use all the information of the individual claims. A linear prediction model based on the individual claims is proposed in this paper, in which only the second order moment of the data is required. This method is flexible and simple and hence easy to apply in reality. The proposed method is compared with the traditional chain ladder method via a simulation study as well, and the simulation results illustrate that the proposed approach performs well