An Application Of ARIMA Model To The Prediction Of The Premiums Of Life Insurance In China

The insurance industry of China has kept rapid developing pace since resuming prosecution, and the insurance premium is quickly increasing year by year. Some materials show that the total insurance premium of China in 2005 reach 492.7 billions RMB, increased 14% compared with 2004. The proportion of life insurance premium in total insurance premium is also increasing. It has exceeded 74% for 4 years. According to it we can figure out that the life insurance premium has certain essentiality not only to the whole life insurance market but also to insurance companies. The amount of insurance premium significantly decides the investment direct, investment intensity, investment mode and the company's future solvency etc. Meanwhile, the insurance premium can make precision scale and estimate on the growth and management performance of the life insurance industry. The insurance premium is also the important index of development which showing the development. So if the company can get more information on developing trend and quotient of the life insurance premium market, they can make better decision when they are making and revising policy and project. It will be beneficial to the company's self-market developing and management performance.The article makes use of time series data from 1982 to 2005, applying the ARIMA model and ARIMA transfer function model to build the forecasting model of the life insurance premium and demonstrating the relation among the gross domestic product, inflation-rate, market-rate and insurance premium. We can make below conclusion. First, in ARIMA model,the insurance premium model-ARIMA(1.1.0) indicates that current period insurance premium will be effect by last time's insurance premium, existing lag effect. Second, ARIMA transfer function model shows that inflation-rate has little effect on insurance premium. The gross domestic product and market-rate has certain influence on insurance premium. The gross domestic product increasing has positive correlation with insurance premium, and the rate has negative correlation with the life insurance premium. So when we build forecasting model of the insurance premium, we should consider both factors of the gross domestic product and rate.