A Study on Spatial Correlation Econometrics Analysis of the Establishment for Bank Branches

In light of spatial level investigation, which coupled with demographic attributes data of Taiwan’s towns and cities. The possible spatial aggregation phenomenon, with the distribution of random or scattered, and then analyzes the locations of distributed spatial aggregation to find out the characteristics of each region in Taiwan. By spatial autocorrelation analysis, the scope of empirical analysis results with the global Moran's I and local indicators of spatial association (LISA) which are based on the districts of towns and cities in Taiwan. To explore the spatial clustering phenomenon of the number of branches of financial institutions in Taiwan, it is found that the analysis results of the demographic attribute data of Sex ratio, the population density, old-age dependency ratio, aging index, and total population in Taiwan’s towns and cities based on the global spatial autocorrelation analysis illustrate clustered distribution in spatial, and the spatial relationship is positive with greater than negative correlations, which means that Taiwan’s 7760 villages are presented in concentrated distribution; and with the spatial autoregressive model has the most explanatory power. Furthermore, through the method of regional autocorrelation analysis comprehend the relationship between the elderly population and the spatial unit (the spatial unit and its adjacent units) with the bank branches establish. The aggregation strength correlation is divided into four quadrants: “HH”, “LL”, “HL”, and “LH”. It is found that the establishment of branches of financial institutions in the “HH” related areas of the elderly population is insufficient.