The Empirical Application Of Bayesian Spatial Quantile Regression Method

Since Koenker and Bassett has put forward the important influence on the thought of quantile regression in 1978, its core is to modeling of the conditions of the dependent variable quantile, making up for the defects of ordinary least squares method in the analysis of regression, and it has incomparable advantages than frequency school. It based on the conditions of the dependent variable quantile regression, which was carried out on the independent variable so that all under the quantile regression model could be obtained. Quantile regression has made great developments, especially the rapid popularization of the computer in recent years, the research and application of quantile regression model is becoming one of the forefront and hottest issues in statistical and modern econometrics.This paper revolves around the quantile estimates and takes quantile regression model as the main line, researched the spatial estimation method and application of quantile regression model and used the method above to make an empirical analysis of the stock market risk. Moreover, we carried on the space statistical method and space descriptive statistics of the space regression model estimation based on the quantile regression, and used the model to make an empirical analysis of per capital income and urban-rural income gap of Guizhou Province.In the part of empirical research, the first one took China’s stock market as the research object, and the data source was between September 2013 and September 2015 stock index day’s closing price. The empirical results showed that the extreme risk of China’s stock market was influenced by the international market, and for the extreme risk of Shanghai A shares are mainly from its own market. The second one took 77 counties of Guizhou Province as the research object, and the data source was 2005, 2007 and 2009 three year’s economic data. The empirical results showed that per capita GDP and income between urban and rural areas was spatial lag autocorrelation and spatial error autocorrelation. To analyze the causes of the urban-rural income gap is to deepen the understanding of the gap between rich and poor, to strengthen the regulation of income inequality for the government.