The main profit sources of insurance company are underwriting profit and investment profit. In the world, the underwriting business of insurance companies is under deficit at present, the Mainland and Taiwan are no exception, so insurance fund application is very important; it effects profit and sustainable operation of insurance company. However, as a part of social security system, insurance company has the functiong of economic compensation and stabilizing social, therefore, the basic goal of insurance fund application is safety. However, sagety and profit are contracted. We can solve this contraction by investing government bonds. This article is based on this idea, by predicting interest rate of government bons, to improve profit of insurance fund application.This article is composed of six chapters. The first chapter introduces undergrand, significance, innovative points and deficiency of the article; class the both domestic and foreign researches of government, insurance fund application and neural network model. The second chapter introduces related concepts of government bond, the facators which influence interest rate of government bonds and government bonds maket of some countries and regions, laies a foundation for the research below. The third chapte introduces domestic and foreign present situation of insurance fund application, the importance of government bonds in insurance fund application, the tendency of interest rate of government bonds to provide supports for the research of insurance fund application. The forth chapter introduces theory of interest rate expectation and neural network model to lay a fuoudation for the fifth chapter. The fifth chapter introduces governmrnt bonds market of Taiwan, and forcasts the interest rate of ten year government bonds of Twaiwan by choosing reasonable variables and neural network model to provide reference for insurance to invest government bonds. The sixth chapter summarizes the main contents, points out the deficiency of the article and the research direction in the future. |