The Theory And Application Of Panel Unit Root And Co-integration

A panel data set is one that follows a given sample of individuals over time, and thus provides multiple observations on each individual in the sample. It involve two dimensions: a cross-sectional dimension, and a time-series dimension, so it contains the homogeneity and specification of cross-sectional dimension, and makes the result more stable, effective and credible. This article firstly makes a clear summary of panel data theory, especially of the unstable panel data model. From static panel data model to dynamic panel and then to unstable panel, it reviews the various test methods and conclusions of unstable panel data, and sum up the relations of every part. Secondly, along with the growth of export and import of China's trade, China's trade surplus is mounting up fast. Aim to solve the problems of China's trade surplus and based on the theory of trade Surplus, this paper analyzes the inside and outside influence factors of China's trade surplus with province level panel unit root and panel co-integration under Chinese opening economy. The paper shows that there is a panel co-integration among net export, wage, traffic construct, consumption, capital asserts investment, export drawback, FDI, processing trade export, high tech import and real effective exchange rate. Comparatively the inside factors can interpret the reason of trade surplus better; the real effective exchange rate plays the most important role in the trade surplus. At last, the paper suggests that we should pay more attention to inside factors to deal with trade surplus.