中国季度宏观经济模型理论及应用

发布时间：2018-05-31 15:56

本文选题：宏观经济模型 + ECM模型　；参考：《中国社会科学院研究生院》2014年硕士论文

【摘要】：近年来，国际经济形势日趋复杂，而且随着市场经济程度的不断加深，我国经济发展也面临更多新的问题与挑战。无论是政府的决策管理还是人们对经济形势的判断都更加困难。寻找恰当的分析方法和工具成为人们关心的问题，而宏观经济模型恰好符合了人们的这一现实需求。宏观经济模型是对现实经济的仿真和模拟，是基础数据、经济理论以及计量方法等的有效结合，在一定程度上已成为研究宏观经济问题的有效工具。国外宏观经济模型的研制已有较长的历史。在美国等发达国家，宏观经济模型在较早时期就已广泛使用。我国宏观经济模型的研制起步较晚，且其中以年度模型为主。近年来，随着数据的丰富，研究技术的不断成熟以及现实经济分析的需要，我国模型研制者逐渐将研究重点转向更具时效性的季度模型。本文在对国内外宏观经济模型研制历史进行梳理的基础上，结合我国步入社会主义市场经济阶段以来经济的运行情况，以宏观经济的收入、消费等不同方面为切入点，以支出法国内生产总值为最终落脚点，通过不同行为方程和恒等式构建联立方程模型，建立了符合我国经济运行机制和特色的季度宏观经济模型，并基于此模型对未来经济走势进行了预测和模拟。论文主要分为五章。第一章为绪论，简要介绍研究内容和目的。第二、三、四章为核心和重点，均围绕本文构建的季度模型展开，分别介绍了模型的概况、构建过程和应用情况。第五章对论文进行了总结，并指出了模型可以进一步改进和提升的地方。本文建立的宏观经济模型以季度数据为基础，样本区间从1992年第1季度至2012年第4季度。在建模之前，本文对原始数据进行了一系列处理，最终得到87个初始变量用于模型构建。在规模上，本季度模型由27个行为方程和37个恒等式构成，共涉及宏观经济的收入、消费、投资、政府财政、对外贸易、金融和价格等七个主要模块。在对宏观经济模块进行分析的过程中，本文同时梳理了各模块涉及到的相关经济理论，如生产理论、消费理论以及国际贸易理论等，这些构成了行为方程估计的理论基础；同时，也运用数据和图形对1992年以来我国经济各主要方面的运行状况进行了分析，这些构成了行为方程估计的现实依据。由于大部分经济变量为非平稳时间序列，本模型主要基于协整理论和误差修正模型(ECM)构建行为方程，由此刻画经济行为的长期均衡以及短期波动特征。但在方程的具体估计过程中，并未单独设置误差修正项，而是对其一般形式进行了估计，然后依据动态建模理论逐步约化得到包容原始模型的最简形式。本文在行为方程之外，又根据模型需要构造了相应的恒等式，以此将各个行为方程进行连接，从而实现对联立方程系统的求解。在众多分析软件中，，本季度模型选用Eviews6.0进行实证分析。模型构建均通过编程实现。整个建模过程共编写了四个Eviews程序，分别用于原始变量频率转换，行为方程估计等。程序的编写大大提高了后期模型求解及情景模拟的速度和效率。为检验模型质量和效果，实现其现实指导意义，本文对2013年至2015年的主要宏观经济变量进行了预测，并就汇率、世界贸易量两个外生变量以及M1供应量这一内生变量进行了不同情景的政策模拟。基于政策模拟的结果，本文建议，未来一段时期，我国应避免人民币在短期内大幅升值，并充分利用国际贸易环境复苏机会，继续保持对外贸易对经济增长的促进作用；同时保持货币供应量与经济发展相适应，以此维护价格水平的稳定。总体来讲，本季度模型较好的模拟了中国经济的整体运行情况，对于经济预测以及宏观经济政策分析具有一定的参考意义。
[Abstract]:In recent years, the international economic situation is becoming more and more complex, and with the deepening of the market economy, the economic development of our country is also facing more new problems and challenges. It is more difficult for the government to make decision management or to judge the economic situation. The economic model coincided with the realistic demand of the people. The macroeconomic model is the simulation and Simulation of the real economy, the effective combination of the basic data, the economic theory and the measurement method, and to a certain extent, it has become an effective tool for the study of macroeconomic problems.
The development of the macroeconomic model of foreign countries has a long history. In the developed countries such as the United States, the macro economic model has been widely used in the earlier period. The development of the macro-economic model of our country is late and the annual model is the main one. In recent years, with the rich data, the development of research technology and the need of real economic analysis. Our model developers gradually shift their research focus to a more timeliness quarterly model.
On the basis of the history of the development of the macro-economic model at home and abroad, and on the basis of the economic operation of our country since the stage of entering the socialist market economy, this paper takes the income of macro economy and consumption as the breakthrough point, and takes the gross domestic product of France as the ultimate foothold, and through the different behavior equations and the identity structure. A joint equation model is built, and a quarterly macroeconomic model conforms to the economic operating mechanism and characteristics of our country. Based on this model, the future economic trend is predicted and simulated.
The thesis is divided into five chapters. The first chapter is the introduction, which briefly introduces the content and purpose of the study. The second, third, the four chapter is the core and the focus, which are all around the quarterly model constructed in this paper. The general situation of the model, the construction process and the application situation are introduced respectively. The fifth chapter summarizes the paper and points out that the model can be further improved and promoted. The place.
The macro economic model based on the quarterly data is based on the quarterly data. The sample interval is from the first quarter of 1992 to the fourth quarter of 2012. Before the modeling, the original data were dealt with a series of processing, and finally 87 initial variables were used for the model construction. On the scale, the model was composed of 27 behavior equations and 37 identities in this quarter. And macro economic income, consumption, investment, government finance, foreign trade, finance and price and other seven main modules. In the process of analyzing the macro-economic module, this paper also combs the related economic theories involved in each module, such as production theory, consumption theory and international trade theory, which constitute the behavior equation estimation. At the same time, we also use data and graphics to analyze the operating conditions of the main economic aspects of our country since 1992, which constitute the realistic basis for the estimation of the behavior equation.
Because most of the economic variables are nonstationary time series, the model is mainly based on the cointegration theory and the error correction model (ECM) to construct the behavior equation, which depicts the long-term equilibrium and short-term fluctuation characteristics of economic behavior. However, in the specific estimation process of the equation, the error correction is not set separately, but the general form is carried out. It is estimated that the simplest form of the tolerant original model is gradually reduced according to the dynamic modeling theory. In addition to the behavior equation, the corresponding identities are constructed according to the needs of the model. In order to connect the equation of behavior, the solution of the simultaneous equation system is realized.
In many analysis software, this quarter model selects Eviews6.0 for empirical analysis. The model construction is realized through programming. Four Eviews programs are written in the whole modeling process, which are used for frequency conversion of original variables and estimation of behavior equation respectively. The programming of the program greatly improves the speed and efficiency of the model solving and scenario simulation in the later period.
In order to test the quality and effect of the model and realize its practical significance, this paper forecasts the major macroeconomic variables from 2013 to 2015, and carries out a policy simulation of different scenarios on the exchange rate, two exogenous variables of the world trade volume and the endogenous variable of M1 supply.
Based on the results of the policy simulation, this paper suggests that in the coming period, China should avoid a substantial appreciation of the RMB in the short term, and make full use of the international trade environment recovery opportunities to continue to maintain the role of foreign trade in promoting economic growth, while maintaining the supply of money and economic development in order to maintain the stability of the price level.
In general, the model of this quarter is better to simulate the overall operation of China's economy, which has some reference significance for economic forecast and macroeconomic policy analysis.
【学位授予单位】：中国社会科学院研究生院
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F124;F224

【参考文献】