Adaptive Learning And Optimal Monetary Policy Of China

Modern economy is a system of economic agents interact with each other, but also is a system of expectation feedback of interaction between private sector and central bank. Because of the economic system, all kinds of dynamic are often changing from micro to macro level, as a rational economic agent to pursuit the utility of maximization, adjustment of consumption or investment decision-making is often based on its expectations on macro economic variables. Monetary policy in order to achieve its goals must consider the formation mechanism, influencing factors and dynamic characteristics of public expectations.Different economic agent makes different expectation for macroeconomic cognition and for the future economic trend, observation and adjustment has always been associated with the anticipation and decision process of economic agents, so the public must have long-term or even lifelong in observation and learning process dynamically. As another expectation formation mechanism besides rational expectation, adaptive learning theory is widely used in the fields of macroeconomic analysis, including monetary policy. Moreover,the economic agents learn from each other more through the transfer of information, in the microeconomy agents' learning mechanism,learning theory appears process rational assumptions instead of the traditional model-substantive rationality, providing a more solid micro foundation for the analysis of optimal monetary policy, contributing to the effect of monetary policy better, with better realizing the goal of monetary policy.At present, both the theoretical research on monetary policy or operating in practice, the analysis of monetary policy is mainly based on rational expectations, but the monetary policy research under adaptive learning assumption is relatively insufficient, as a modest departure from rational expectations and more close to the economic actual expectation formation mechanism, under the adaptive learning asumption, whether the public to learn and how to learn, how does the formation mechanism of inflation expectations on inflation affect the persistence through adaptive learning? Facing to the public who use the adaptive learning as expected the formation mechanism, how to choose and how to face the problem about the optimal monetary policy choice and practice in china?This paper will study and analyse the micro foundation of the monetary policy from learning and expectations formation angle, analyse the macro foundation of monetary policy from the angle that adaptive learning affect the infiation,starting with the expectation, learning and the optimal monetary policy theory analysis.and assuming that Chinese public use adaptive learning to form expection, establishing the Chinese Phillips curve under adaptive learning, combined with the data of1996-2013, with Chinese public under adaptive learning to make empirical test on inflation persistence effect, I think the short-run inflation persistence is mainly resulted from the inflation expectation under adaptive learning, the optimal monetary policy in China should be forward-looking and robust policy, and actively respond to inflation, while strengthening the management of expectations, stabilize monetary and financial environment.Then, combining with the analysis framework and operating characteristics of the current monetary policy of china, according to the emprical analysis in chapter5,1analysed the optimal monetary policy rules under the representative traditional monetary framework under adaptive learning. Specifically, based on the traditional monetary policy analysis framework,relaxing the assumption of rational expectations, introducing the adaptive learning theory for agent to form expectation, I got the simple single parameter optimal monetary policy rule under adaptive learning through the model analysis; then I analysed the flexible inflation targeting rules under adaptive learning with traditional monetary policy framework.At the same time, thinking about that the application of learning theory in the optimal monetary policy in China also belongs to the research field relatively new, this paper tries to summarize systematically the related theory, supplement the related literature ot the optimal monetary policy analysis under adaptive learning.The fulltext is divided into seven chapters, the basic structure and the main contents are as follows: The1st chapter is the introduction, mainly introduces the issues background? research purpose, research methods and research significance. The shortcomings of the rational expectations theory in the research on optimal monetary policy, as well as the problems of China monetary policy facing, research methods will be used and the possible innovation in this paper.The2nd chapter is literature review of this paper, to study,analyse and review the related domestic and foreign literatures about the development process of the learning theory, and the application of the research in optimal monetary policy under adaptive learning. Pointing out that the adaptive learning can give another interpretation in inflation volatility and persistence, to provide a new perspective closer to real economic explanation for expectation management and researching angle for central bank and the optimal monetary policy analysis.The3rd chapter systematically analyses on expectations and learning theory, introduces and reviews the main expectations theory and its history of development; and based on the game learning theory, this chaper devided the agents'learning theroies into individual learning and social leaning, and briefly reviews the development of the Phillips curve, to make the prepration foe research of the adaptive learning, inflation and the optimal monetary policy research in the following chapters.In the4th chapter, I comprehensive got the expectations formation mechanism under adaptive learing through mathematical derivation and solution, and the three learning algorithms on the mainstream of the detailed analysis is derived and solved, finally I carried on the review and summary of adaptive learning mechanism and learning algorithm.In the5th chapter,I set up Chinese public expectation formation equation and corresponding China Phillips curve under adaptive learning hypothesis, and using the macroeconomic data related to test the Phillips curve of china under adaptive learning, and make the comrehensive analysis for public expectation formation mechanism through the empirical analysis.The6th chapter will be based on the effect on the Chinese inflation of adaptive learning, conducts the research to the Chinese optimal monetary policy. First I reviews the transformation period of china's specific monetary policy framework, operation characteristics and inflation control. Then basing on the conclusion in chapter5,under the traditional framework for the analysis of monetary policy and building learning economic system, I have analyzed he simple one parameter optimal monetary policy rule under the rational expectations and the adaptive learning, and the flexible inflation targeting rule under adaptive learning, to analyse the robust and the looking-froward optimal monetary policy under adaptive learning in China.The7th chapter is the conclusion, policy recommendations, research problems and research prospects of this paper.The main conclusions of this paper are as follows:(1) expectation is to be widely used in macroeconomics research, learning theories can be divided into individual and social learning theory, As the new development of the expectations theory, adaptive learning is the reasonable departure from rational expectations, and is a expectation formation mechanism closer to the economic reality.(2) adaptive learning has been widely introduced into forward-looking macroeconomic analysis including monetary policy,The introduction of the adaptive learning mechanism, so that the public can form reasonable expectations of inflation, can relieve the difficulty for monetary policy to regulate and control macroecnomic. The monetary policy rule is designed so that the private sector's expectation can convergence to rational expectation in the long term, and the characteristics in the process that the economy convergence to the rational expectations equilibrium (REE, the same below) under adaptive learning, are the two important directions in the study of monetary policy with adaptive learning.(3) the expectation updating under adaptive learning is on the agent get new information and then reestimate the parameters to achieve the updating of expectation, there are three popular parameter updating method (algorithm) for adaptive learning:Recursive least squares learning, stochastic gradient learning and constan gain learning. The dynamic process of adaptive learning depends not only on the learning algorithm, but also depends on the initial conditions. The initial value is very important for the learning algorithm of the model parameters.(4)Under the adaptive learning mechanism,comparing within different empirical results of Phillips curve from different gain parameter values, I found that short-term inflation persistence in China is mainly affected by the expectation under adaptive learning. China's public adaptive learning mechanism is more in line with the recursive least squares learning characteristics, namely the recent inflation expectation brings the maximum effect on the inflation, the public more correct their expectaion according to the recent prediction error,but not to the knowledge foundation of the economic structure.(5)In the special transitional stage of China, monetary policy may pay more attention to the shift from discretion to interest rate tool gradually,so that the inflation could be stable and stable economy, harmonious development could be achieved. Especially in the adaptive learning mechanism, actively anti-inflation, effective communication about the inflation target to the public, helps to anchor inflation expectations of the public, prevent the economy from sliding into stagflation. And consider the flexible inflation targeting under adaptive learning, monetary policy should be to ensure that the parameters meet the determinacy and stability conditions, and consider the convergence rate, so that the economy equilibrium converges to rational expectation equilibrium or on the minimum deviation from equilibrium, and enhance social welfare, reduce the cost of institution changing. In addition, the policy maker strengthens public education helps the public to improve the expected level, establish correct expectation and achieve the convergence for the economy to the rational expectations equilibrium.This paper probably lies in innovation:(1)In view of advanced adaptive learning and research, according to the theories of learning and its micro foundation,and the application of adaptive learning in the research of optimal monetary policy as the main line, from the agent's expectation updating in the adaptive learning mechanism, to the optimal monetary policy analysis and operation, I fistly systematicly sum up and analyse the theory and research of learning, expectation and optimal monetary policy, strengthened and improved the relevant research literature.(2) Under the perspective of adaptive learning, I propose a inflation expectation forming equation which affected by the factors of inflation inertia, the real estate price index, M2growth rate and the interbank interest rate growth rate. based on the observation data of these influence factors and adaptive learning mechanism,I calculated the public's expectations of inflation in China under adaptive learning with the constant gain learning and the recursive least squares learning algorithm. And I analyse the learning mechanism and characteristics of Chinese public. Combined with the related macro data, I finished the empirical tests of the Chinese Phillips curve under adaptive learning.(3) Perfoming the study of Chinese optimal monetary policy selection and operation practice applied with adaptive learning. According to the results of empirical testing, based on the traditional monetary policy analysis framework,I propose the optimal monetary policy rule under adaptive learning, provide a new interpretation for the Chinese robust and forward-looking monetary policy theoreticaly, I suggest that being about the policy operation under adaptive learning, besides anchorring public expectations by the positive anti-inflation policies, attention should also be paid to the inflation target communication and the guidance to public expectation, preventing the economy from sliding into stagflation. While the flexible inflation targeting under adaptive learning should ensure the stability and determinacy of rational expectations equilibrium, the speed of convergence, and strengthen education to the public, in order to achieve Chinese monetary policy target better.