Abstract:  Under the background of the 70th anniversary of the founding of New China, it is of great practical significance to accurately identify the dynamic correlation mechanism between China's foreign trade and economic growth. This paper chooses the annual sample data from 1952 to 2018, and measures the long-term nonlinear cointegration relationship and short-term dynamic correlation mechanism of import, export trade and economic growth by constructing a nonlinear autoregressive distribution lag model and a time-varying parameter vector autoregressive model with stochastic volatility. The empirical results show that: firstly, in the perspective of long-term equilibrium, there is a significant nonlinear cointegration relationship between export, import trade and output. in which the increase in export and import trade has less effect on the pull of output than its decline, and the export trade has a greater effect on output than import trade turns to a positive pull effect. Thirdly, at important time points such as the implementation of reform and opening up in 1978, the accession to the WTO in 2001, and the outbreak of the US financial crisis in 2008, export trade showed a gradual weakening effect on economic growth, while import trade shows a gradual increasing promoting effect. Therefore, in order to maintain sustained and healthy economic growth, China should continue to implement the strategy of optimizing and upgrading the structure of foreign trade. At the same time, we should firmly implement the strategy of innovation-driven and expanding domestic demand, replace the factor-driven in the past with innovation-driven, so as to make economic development rely more on domestic demand, especially consumption-driven, and gradually leave the traditional development mode to realize the economic transformation of our country.

Key Words:  Export Trade; Import Trade; Economic Growth; NARDL Model; SVTVP-VAR Model