基于随机方差扩大模型的对中国宏观经济统计数据的结构变化分析

The Analyses on Structure Changes in Chinese Macroeconomics Data Series based on RVAR Model

本文从经济系统的角度运用随机方差扩大模型对我国36个宏观经济时间序列趋势成分的结构变化特征进行了全面的分析,发现了数据结构变化的特点和规律。研究结论表明:趋势成分和序列本身具有类似的结构变化特征,但是结构变化的范围、幅度却存在显著性差异;我国宏观经济统计数据结构变化的主要特征受到相应序列的趋势成分、波动成分的组合方式及其构成比例的深刻影响;大部分结构变化点的出现或多或少都是以聚集成堆的形式出现的,它们之间存在深刻的内在联系,结构变化点的出现大多数与各种历史因素以及外部冲击有关;大部分发生结构变化原始序列与其趋势成分的波动性特征出现了显著差异,而结构变化点修正后序列及其趋势成分的波动幅度与范围极其相似。

This paper uses RVAR model to deeply analyze 36 Chinese macroeconomics data series from the view of economic system and finds out the features and rules of the structure changes. The results show： The trend components and the primitive series have the same features of structure changes, but there exist obvious differences between them on the ranges and scopes. The structure changes of Chinese macroeconomics data are mostly affected by the compounding way and the constitutive proportion of the trend components and the volatility components. The most of structure changes appear in the cluster form more or less, and there are deep internal relationship between them, and the structure changes are mostly caused by the historic factors and external shocks. Almost all of the original series and the trend components have different volatility features in the structure changes, but corrected series and the trend components are the most similar features.