Statistical Analysis Of The Factors Affecting Premium Income In China's Insurance Industry

With the development of the information age,the processing of high dimensional data has become a hot issue in data analysis.In particular,A large number of candidate variables often appear in the initial stage of modeling.If there are more irrelevant variables or missing important variables in the model,the deviation between the predicted result and the true value will increase,and the model credibility will be greatly reduced.Therefore,the importance of variable selection in the modeling process is obvious.In this paper,using Lasso quantile regression,adaptive Lasso quantile regression and partial least squares regression for the empirical analysis.The empirical analysis of this article is divided into two.parts to study the influencing factors of premium income.The first part is based on the 31 provinces and.autonomous regions in China and mainly analyzes the relationship between the premium income of each region and the local economic,cultural and population factors.Based on the panel data model with individual random effects,this paper analyzes the Lasso quantile regression and adaptive Lasso quantile regression analysis of interprovincial panel data from 2007 to 2016.In addition,cross-sectional data for 2007,2011,and 2016 were selected for partial least-squares analysis.The second part is based on China's insurance companies premium income and the impact of internal factors,as the insurance companies in the past decade has increased,in order to ensure data consistency,This article specifically excludes some insurance companies,retaining only the life insurance companies that are in continuous operation from 2007 to 2016,a total of 46.Lasso quantile regression and adaptive Lasso quantile regression were used to analyze the personal insurance company's panel data.Empirical results from the inter-provincial panel data at the macro level show that GDP,the total retail sales of social consumer goods,fiscal revenue,and financial expenditure are all important contributors to premium income.The empirical results of the personal insurance company data at the micro level show that the return on assets,current asset turnover,fixed asset rate are factors that have a large impact on premium income within the company.