城市保险业经济环境指标的偏最小二乘分析

PLS Regression Analysis of Economic Environment Factors of Urban Insurance Development

本文借助偏最小二乘回归对中国部分城市保险业发展规模和水平及其经济环境进行了分析和建模。偏最小二乘回归模型不仅可以有效的解决自变量之间存在的多重共线性问题,还是一种多因变量对多自变量的回归建模,本文分别采用人身保险和财产保险的保费收入和保险密度作为因变量集,选取城市人均地区生产总值,城镇居民人均可支配收入,人均储蓄年末余额,单位面积货运量等组建自变量集。我们发现,自变量集和因变量集存在明显的相关关系;每一自变量对因变量的解释能力较为均匀,说明经济环境指标的选取的科学有效性;其中人均地区生产总值和地区生产总值的解释能力最强。分析结果与一般理论和认识十分吻合,准确找出对城市保险发展有重要影响的经济因素,有利于我们正确认识城市保险业发展的动力、潜力、规模和水平,以便对该城市保险业的发展阶段和发展战略有客观科学的把握和筹划。

This article constructs a partial least square regression model to analysis economic environment factors of urban insurance development. Partial least square regression can not only avoid harmful effects due to multicollinearity, but also make model linear regression between multi - dependent variables and multi - independent variables. The result shows that the independent variables we chose in the model are quite effective and appropriate; there among, GDP and GDP per capita contribute most. The conclusion can offer some objective perception and beneficent suggestions for policymakers in terms of local insurance development.