Xiaolin Liang, Maozai Tian

【Abstract】

Assets are used in the production and business operation activities by enterprises, and thus enterprises will make more assets. That is to say, it can generate the en-terprises` revenue. So, there must be a certain correlation between enterprise` revenue andssets. Based on the general linear fitting of the listed companies` total assets and revenuethis paper adopt quantile regression model to make an in-depth analysis of the relationshipbetween total assets and revenue. It turn out that, there is a positive correlation betweentotal assets and revenue. When enterprises improve a certain assets, the enterprises whichhave high quantile in revenu will make greater increase in revenue. Due to the outliers in thedata collected, the paper in section 5 discusses the problem about statistical diagnosis in linearquantile regression model, in analogy to the R square in general linear model, we get the Rquare in different quantiles, By removing the outliers, draw the quantile regression model ismore robust than general linear model. As to the data in this paper, the fitting effect in highquantile is better.

【Keywords】

correlation, revenue, total assets, quantile regression, statistical diagnosis