Research on Earthquake Insurance Losses Fitting in China Based on G-h Distribution
|School||Ocean University of China|
|Keywords||g-h distributions g-h VaR Earthquake Insurance|
As for the earthquake disasters, China is one of the most serious countries all over the world, the earthquakes always cause great property and person damage to the people, and influence the stability and health of the economy development. However, the earthquake insurance mechanism is not so complete, the loss and the compensation have great difference compared to the high frequency of the disaster. In the face of the great earthquake loss, government allowance and society salvation is not enough for the reconstruction. So it is very important and essential to push the establishing and perfection of the earthquake insurance, which is an important method to disperse risk and compensate the loss, and it is also very sufficient in lightening the government’s financial pressure and strengthens the anti-earthquake ability.Other than the policy-leading to completing the earthquake insurance system completing, the insurance actuarial science is also needed to offer the technical support of which the basis and core is earthquake loss distribution. What’s more, a better fitting of the loss distribution can also provide support to the insurance pricing, the reserved funds’ collection and the insurance supervision. So it is very meaningful both theoretically and actually. In this context, the author introduces the g-h distribution which is generally applied in the financial investing yield field and rarely in the insurance field into the earthquake loss data fitting. This thesis’s main work is to fit the earthquake insurance loss data; it can be separated into4parts:Firstly, the thesis introduces the concept and meaning of the earthquake insurance and combs the research condition all over the world to lay the root for the following research.Secondly, an important factor to develop the earthquake insurance is the actuarial science, of which the premise is better fitting the earthquake loss distribution. So the thesis introduces some common distribution model and finds that they are somewhat powerless in the earthquake loss distribution fitting which is characterized with fat tail and sharp peak through the author’s analysis, then the author introduces the g-h distribution.Thirdly, the thesis introduces the concept, quality, measuring methods of the parameters, and the calculation of the VaR based on g-h distribution of the g-h distribution in detail, g-h distribution is developed from the normal distribution and can describe the distribution more accurate with the deflecting parameter g and tail parameter h. On this basis, the author discusses the applying prospect of the g-h distribution in the insurance field.Finally, the thesis selects the43-year-long earthquake loss data over100million from1969to2011and uses it to make an empirical research. The author firstly analyzes the character of the earthquake loss data distribution, then builds a model using the g-h distribution. The empirical result indicates that the g-h distribution perfectly fits the earthquake loss data.The focus and innovate points are reflected as followed:introducing the g-h distribution into the analysis of the earthquake risk and tests its applicability through the empirical research; fitting the loss distribution of China’s earthquake disaster, preparing for the actuarial pricing of the earthquake insurance; introducing the VaR based on g-h distribution into the earthquake research; discussing the applying prospect of the g-h distribution in earthquake insurance.