当前位置:主页 > 经济论文 > 技术经济论文 >

中国学习型通胀预期的实证研究

发布时间:2019-06-06 20:49
【摘要】:现代宏观经济理论认为经济学与自然科学的核心区别在于经济人所做出的前瞻性决定。在经济运行的每个环节,预期起到了重要的作用。公众的通货膨胀预期是影响实际通货膨胀率的关键因素,通过有效的货币政策手段管理公众这一预期,是实现物价稳定这一货币政策目标的有效途径。当引导和管理市场通货膨胀预期已成为发达国家提升货币政策效果的重要手段时,中国政府也应与时俱进,推出自己的通胀预期管理措施。而预期管理政策有效性的关键在于知晓、秉承通货预期与真实通货膨胀的动态反应路径。预期理论经历了静态预期、外推型预期、适应性预期、理性预期。前三类预期将预期视为经济体系的外生变量,且没有坚实的经济学理论基础。而理性预期作为一种均衡的概念,假定公众知晓一切的计量经济模型,任何时刻均可对经济系统做出无偏的估计。但是理性预期有如下缺陷:首先,公众对真实经济结构有完全认知,信息获取与处理能力强大到始终形成与真实经济一致的无偏结果,这明显与事实不符。其次,没有对公众的预期形成机制进行具体描述,而只注重对预期结果的一系列规定。最后,大部分理性预期模型都存在多重均衡解的问题,到底出现哪种均衡存在任意性。学习型预期放松了理性预期假设暗含的一系列严格的条件,认为公众的预期形成机制是从感知运转模型向实际运转模型的映射过程,可以避免理性预期模型存在的多重均衡解问题。基于此,本文于三个层面(理性预期多重均衡解、学习型预期形成机制及稳态条件、学习规则)系统地阐述了适应性学习预期理论,并据此测度中国学习型通胀预期,进一步地对学习型通货膨胀预期的特征进行分析与实证检验。在实证分析中,考虑央行货币政策的平滑倾向与通货膨胀滞后因素,构建状态空间模型,利用卡尔曼滤波推导出公众通胀预期的学习规则,测度中国学习型通货膨胀预期,并结合韦达定理求解出学习规则的稳态值。进一步地,根据理性预期假设满足的无偏性、有效性及认知偏差为0且无自相关的三个性质,通过设置不同显著性水平,检验学习型通胀预期的近理性特征。
[Abstract]:Modern macroeconomic theory holds that the core difference between economics and natural science lies in the forward-looking decisions made by economic men. In every link of economic operation, expectation plays an important role. Public inflation expectation is the key factor affecting the actual inflation rate. Managing the public expectation through effective monetary policy means is an effective way to achieve the monetary policy goal of price stability. When guiding and managing market inflation expectations has become an important means for developed countries to improve the effectiveness of monetary policy, the Chinese government should also keep pace with the times and introduce its own inflation expectation management measures. The key to the effectiveness of expected management policy lies in knowing, adhering to the dynamic response path between currency expectation and real inflation. The theory of expectation has experienced static expectation, extrapolation expectation, adaptive expectation and rational expectation. The first three types of expectations regard expectations as exogenous variables of the economic system, and there is no solid economic theoretical basis. Rational expectation, as a concept of equilibrium, assumes that the public knows all econometric models and can make unbiased estimates of the economic system at any time. However, rational expectations have the following defects: first of all, the public has a complete understanding of the real economic structure, and the ability to obtain and process information is so strong that it always forms unbiased results consistent with the real economy, which is obviously inconsistent with the facts. Secondly, there is no specific description of the expected formation mechanism of the public, but only a series of provisions on the expected results. Finally, most rational expectation models have the problem of multiple equilibrium solutions, and which equilibrium is arbitrary. Learning expectation relaxes a series of strict conditions implied by rational expectation hypothesis, and holds that the expected formation mechanism of the public is the mapping process from perceptual operation model to actual operation model. The problem of multiple equilibrium solutions in rational expectation model can be avoided. Based on this, this paper systematically expounds the theory of adaptive learning expectation at three levels (multiple equilibrium solution of rational expectation, formation mechanism and steady-state condition of learning expectation, learning rule), and measures the expectation of learning inflation in China. Further analysis and empirical test of the characteristics of learning inflation expectations. In the empirical analysis, considering the smooth tendency of central bank monetary policy and the lagging factors of inflation, the state space model is constructed, and the learning rules of public inflation expectation are deduced by using Kalman filter to measure the learning inflation expectation of China. Combined with Weida theorem, the steady state value of the learning rule is solved. Furthermore, according to the three properties of unbiased, validity and cognitive deviation of 0 and no autocorrelation, the near rational characteristics of learning inflation expectation are tested by setting different significance levels.
【学位授予单位】:东北财经大学
【学位级别】:硕士
【学位授予年份】:2016
【分类号】:F822.5

【参考文献】

相关期刊论文 前10条

1 管河山;邹清明;罗智超;;时间序列平稳性分类识别研究[J];统计与信息论坛;2016年04期

2 刘立;;以建设金普新区为契机加快大连东北亚国际航运中心建设[J];决策咨询;2015年01期

3 杨钊;;滨海新区航运保险业发展对策研究[J];港口经济;2013年09期

4 崔惠贤;张碧波;;基于GEM模型的上海航运保险产业竞争力研究[J];上海金融学院学报;2013年02期

5 李新伟;;上海培育航运再保险市场研究[J];保险研究;2012年09期

6 何惠珍;;我国航运保险的发展现状、制约因素及对策建议[J];学术交流;2012年01期

7 袁t焧,

本文编号:2494589


资料下载
论文发表

本文链接:https://www.wllwen.com/jingjilunwen/jiliangjingjilunwen/2494589.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户d9cd4***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com