基于若干个统计模型的宏观经济监测比较

发布时间：2018-02-13 05:03

本文关键词： 宏观经济监测向量自回归模型隐马尔科夫模型马尔科夫开关模型　出处：《复旦大学》2010年硕士论文　论文类型：学位论文

【摘要】： 宏观经济周期监测,是对宏观经济进行全面系统的分析和判断。通过对表征经济活动过程和现状的一系列指标进行的监督和量测,对经济活动未来可能发生的转折和重大变化提出警报。对宏观经济周期进行监测,其意义在于：一、正确评价当前宏观经济运行的状态,揭示经济运行中的各种隐患和问题,恰当地反映经济形式的冷热程度,并能承担短期经济形式分析的任务；二、反映宏观经济的调控效果,判断宏观经济措施是否运用恰当,是否起到了平抑经济波动幅度的效果等；三、为企业开展经济活动提供参考。企业经营决策者可以利用宏观经济监测与预警体系提供的信息,及时预见即将出台的宏观调控措施。使企业经营调整适应政府的调控措施；四、为政府宏观调控决策提供依据。宏观经济监测与预警系统可以给政府调控措施、内容及出台时机提供必要的信息。传统的宏观经济监测方法大都是逐步回归分析法,这是基于所选取的经济预警指标之间的关系为线性的假定。而在现实情况中,指标之间的关系往往是非线性的。所以,本文引入隐马尔科夫模型(HMM)与马尔科夫开关模型(MSM)对宏观经济监测与预警,提出与传统依赖于指标之间线性关系的向量自回归模型(VAR)相区别的宏观经济监测与预警体系,并从拟合准确度与预测准确度两方面对各个模型的监测预警效果进行比较。实证分析结果表明,与传统线性模型相比,二阶隐马尔科夫模型这两方面的综合表现较其他模型,包括向量自回归模型,有更优的表现。本文包括四章。第一章,引言部分,主要阐述宏观经济监测预警系统的研究背景、意义及其发展历程,对前人应用过的各种研究方法进行评述。第二章,在遵循全面性、灵敏性、重要性、系统性、相对稳定性、数据的可得性和可操作性等原则的基础上从国家统计局公布的指标中选取一部分指标构建中国的宏观经济监测预警指标体系。第三章,描述本文采用的研究方法——向量自回归模型、隐马尔科夫模型和马尔科夫开关模型。第四章,实证分析,收集每期的宏观经济评价指标数据,比较不同模型对数据的拟合准确度与预测准确度。第五章是结论部分,是对本文结果的总结和对未来的展望。
[Abstract]:Macroeconomic cycle monitoring is a comprehensive and systematic analysis and judgment of the macro-economy through the monitoring and measurement of a series of indicators that characterize the process and status of economic activity. Warning of possible future shifts and major changes in economic activity. The significance of monitoring the macroeconomic cycle lies in: first, correctly evaluating the current state of macroeconomic operation, revealing all kinds of hidden dangers and problems in economic operation, and properly reflecting the degree of cooling and heat in the economic form. And can undertake the task of short-term economic form analysis; second, reflect the effects of macroeconomic regulation and control, judge whether macroeconomic measures are properly used and whether they have played a role in stabilizing the extent of economic fluctuations; third, Providing a reference for enterprises to carry out economic activities. Business decision makers can make use of the information provided by the macroeconomic monitoring and early warning system to foresee in time the forthcoming macro-control measures to enable enterprises to adjust their operations to the government's regulatory measures; fourth, Macroeconomic monitoring and early warning system can provide necessary information for government regulation and control measures, contents and timing. The traditional macroeconomic monitoring methods are mostly stepwise regression analysis, which is based on the assumption that the relationship between the selected economic warning indicators is linear. In reality, the relationship between indicators is often nonlinear. This paper introduces Hidden Markov Model (HMMM) and Markov switching Model (MSM) to macro-economic monitoring and early warning, and proposes a new system of macro-economic monitoring and early warning, which is different from the traditional VAR-based vector autoregressive model, which depends on linear relationship between indicators. The results of empirical analysis show that compared with the traditional linear model, the combined performance of the two models is better than other models. Including vector autoregressive model, there is better performance. This paper includes four chapters. Chapter one, introduction, mainly describes the research background, significance and development process of macroeconomic monitoring and early warning system. Based on the principles of sensitivity, importance, systematicness, relative stability, availability and maneuverability of data, a part of the indicators published by the National Bureau of Statistics is selected to construct China's macroeconomic monitoring and early warning index system. This paper describes the research methods used in this paper: vector autoregressive model, hidden Markov model and Markov switch model. Chapter 4th, empirical analysis, collects macroeconomic evaluation index data for each period. Chapter 5th is the conclusion part, which is the summary of the results of this paper and the prospect of the future.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：F224;F016

【参考文献】