With the improvement of China's comprehensive national power, the share of tertiary industry in GDP has been increasing year by year, and it has started to take a dominant position gradually, playing an increasingly important role for economic growth. In this paper, an ARIMA model is developed to forecast the value of tertiary industry from three perspectives, namely, society, government and people. Secondly, this paper conducts smoothness test, Johansen co-integration test and Granger causality test on the variables, and then introduces a four-factor VAR model to explore the relationship, and performs variance decomposition and impulse response analysis on the results.