s曲线与分量回归分析-以大中华地区保险场为研究对象.doc

S曲線與分量迴歸分析-以大中華地區保險市場為研究對象 摘要 保險滲透率enetration)，是保費收入佔GDP的比重，是衡量一個國家保險普及率相當重要的指標。保費wiss Re)，取得其2002~2008年間的保險滲透率以及保險密度資料，分成保險業、壽險業、非壽險業3種類型。並以最小平方法(OLS)、分量迴歸(Quantile regression)模型、Logistic模型進行分析，判斷出保險滲透率與人均GDP之間關係是否顯著，並檢查各分量下之關係，及判斷有無符合Enz(2000)提出的人均GDP與保險滲透率間的S曲線關係，最後以大中華地區4國家與實證結果印證。結果發現不論是在最小平方法或分量迴歸模型下，收入與滲透率間呈現正向顯著關係；而Logistic模型建立出來的圖形中，都符合S曲線關係。 關鍵詞：  The Application of S -Curue and Quantile Regression Anglysis in the Great China Area Insurance Markets. Student：HongZheng Li Advisor：Dr. Ming-Sun Horng Department of Risk Management and Insurance National Kaohsiung First University of Science and Technology Abstract Insurance penetration, the proportion of premium revenue on GDP, represents how much of total income people in the nation spend on insurance, and has been considered a crucial indicator to measure the insurance coverage of one country. Insurance density, the so-called “Premiums per capital”, denotes how much an individual in the nation spend purchasing insurance. Due to the fact that the insurance revenue can be influenced by the population, the economic scales, level of peoples income, and some other factors, the insurance penetration has long been used internationally to measure the developing conditions of the real insurance market in every country. In this paper, from the Swiss Re Bank we obtained the data of the insurance penetration and the insurance density from 2002 to 2008, separating them into three types, the insurance, life insurance and non-life insurance, and analyzed them with ordinary least squares (OLS), quantile regression model and logistic model, identifying whether the relationship between the insurance penetration and GDP per capita is significant, screening the relationship between each quantile, determining whether there exists the S curve relationship, suggested by Enz (2000), between GDP per capita and insurance penetration, and finally ending with the empirical results for testing 4 countries in the Grea