The Application of Additive Seasonal Model in the Analysis of Domestic Credit Scale

Abstract

To establish statistical model of measurement, this paper analyses the five major domestic economic data whether impact on domestic credit amount, for the purpose of the data cointegration test to eliminate the spurious regression result in error, and through the OLS regression estimate and error correction model is established for dynamic balance of feedback in the short term, then use VAR impulse response function to feedback the long-term dynamic equilibrium, the last three parameters exponential smoothing model is set up to predict the future scale of the credit. According to the conclusion, the total domestic credit scale is a series with cycle fluctuation, seasonal effect and long-term trend. Domestic money supply and domestic fiscal revenue are the main factors affecting the scale of domestic credit. In other words, the fluctuation of domestic credit model tends to return to the normal level, which is balanced and restricted by the medium and long term stable co-integration relationship of the model. To absorb the impact of domestic fiscal revenue or an increase in the domestic money supply, banks adjust for changes in the size of credit, and then show little change. In the short term, the two variables significantly affect the credit scale, but the long-term impact is not significant. Under the condition of constant economic environment, it will still show a steady rise in the future, and the trend of annual cycle is increasing month by month.