Analysis Of Characteristics Of China’s Inflation Dynamic Mechanism

After decades of continuous development, Phillips curve theory has become an important fundamental research tool for the theory of inflation mechanism in Western economics. As with many economic problems, any undisputed theoretical model has yet appeared. Hybrid Phillips curve model is the latest version of the Phillips curve, which taking into account the impact of both forward-looking and backward-looking factors on the formation of inflation. Based on the hybrid Phillips curve model, this paper fully absorbed the latest research results of the Western economic circles, and extended the model of the structure to improve the empirical methods. By using China’s quarterly data for the empirically test of the model, we provide a reference for the formulation of monetary policy based on the empirical results analysis.At the beginning of this article, we pointed out the background and significance of our topics, and explained the significance of research about the Phillips curve for the current monetary policy in China. The focus is on the hybrid Phillips curve model used in this article. We also explained the main idea of this article and our research methods. Then we made a derivation of the hybrid Phillips curve model, and pointed out the academic progress of the Western economies. This article focuses on two controversial issues in the model study.The first one is about the application of the expected inflation value. Since the inflation expectation is a unobservable variable, alternative method were used in the past empirical studies, take the next term inflation data as expected values. We used the least squares estimation method for the regression between the actual rate of inflation and the instrumental variables, thus residuals are removed from the next actual inflation rate and can be used as the expected inflation. This method solved the problem of expected inflation value’s application.The second problem is that the hybrid Phillips curve model implicitly assumes that the setting error of the model is not a related sequence. This assumption has been questioned. Through empirical study we found that serial correlation also exists in the empirical study of the mixed Phillips curve model in China. Therefore, this paper expanded the inflation lagged term order in empirical research to solve this problem.The empirical results show that the extended hybrid Phillips curve model well simulated the Chinese inflation dynamic, and our model did exclude serial correlation. Increased inflation lags are statistically significant, means that forward-looking factors play a greater role in China’s inflation formation.