Research On Monitoring And Forecasting Of China’s Macro-economy In The New Normal

The fluctuation of the business cycle is an essential theoretical issue of macroeconomics.Under the modern market economy,severe cycle fluctuation might bring about a huge loss of economic growth efficiency.Therefore,the study of the regularity of the business cycle fluctuations not only could help to maintain the stable growth of economy,but also be one of the important contents of stable economic growth and scientific development.Monitoring and forecasting business cycle fluctuations accurately are both an indispensable part of the government’s macroeconomic regulation system and an important reference factor for business decision-making.Since the 21st century,with the continuous deepening of the reform of the domestic economic system and the integration of the global economy,the marketization and internationalization of China’s economic operations have gradually emerged.The severe impact of the international financial crisis caused by the US subprime mortgage crisis in 2008 on the real economy of each country has made government departments and economic circles once again realize the importance and urgency of studying and mastering the regularity of the economic cycle fluctuations and improving the economic monitoring and early warning system.At present,China’s economy is in an important period of strategic opportunities,but at the same time we will face more and more severe challenges.In order to maintain a stable,rapid growth of the economy and long-term sustainable development,it is urgent to explore and study in depth the structural changes and fluctuation characteristics of the economic growth cycle in China.Based on the experience and achievements of developed countries,this paper systematically studies the structural changes,fluctuation characteristics and monitoring and forecasting methods of China’s business cycle,narrowing the gap on relevant measurement methods between domestic research and developed countries,and improving timeliness and accuracy of the analysis and forecast of economic cycle fluctuations.The analysis and forecasting provide the scientific basis for the government to judge and grasp the economic trend in advance,and take timely and scientific measures to maintain the stable growth of the national economy.The main work and innovations in this paper can be summarized as follows:1.Analysis the fluctuation characteristics of each phase in China’s business cycle and the transition characteristics between different regimes.This paper not only uses the monthly coincident index based on mixed frequency dynamic factor model to comprehensively characterize the changes in economic climate,but also introduces the Markov Switching intercept and variance terms to the model,Which could provide a more comprehensive analysis of China’s economic cycle from the dual perspective of the economic growth and economic fluctuations.Compared with previous studies,the biggest feature of this paper is that Markov Switching intercept and variance term is introduced into the mixed frequency dynamic factor model.Therefore,the evolution dynamics of the intercept and variance fluctuation of the coincidence index can be analyzed in detail.This paper analyzes the phase characteristics of the economic cycle through the filter probability of the coincidence index under different regimes to thoroughly examine the fluctuation characteristics of China’s economic cycle.At the same time,we can analyze the transition characteristics of different phases of China’s economic cycle by combining the joint transition probability matrix of intercepts and variances,and attempt to explore the possible cyclical forms of China’s economy under the background of the new normal economy through the joint transition probability matrix.2.An Investigation of timing and characteristics of China’s Business Cycle fluctuations in the New Normal.Taking into account the possibility of structural changes in the characteristics of China’s economic cycle fluctuations,We use the four-regime Markov Switching mixed frequency dynamic factor model,whose transition probability that correspond to low volatility state is constrained to the absorbtion state,to reflect the cyclical evolution of economic growth rates and economic fluctuations,and to identify the starting point of the "new normal" of China’s business cycle fluctuations.On this basis,we examine the new features and new characteristics that China’s economic cycle fluctuations in the new normal.3.Monitoring and forecasting of China’s economic climate in real time.This paper uses the Markov Switching dynamic bi-factor model,by imposing some constraints on the joint transition probability matrix;to describe the time-varying leads of the leading index to the coincidence index at different phases of the business cycle.On the basis of identifying the characteristics of the periodic changes of the leading index and the coincidence index,We use the leading index to carry out a pre-warning for the changing trend of the coincidence index in the near future.When the economic structure changes or the economy fluctuates severely,the parameters estimated by the traditional Markov Switching model are often unreliable.Therefore,it is impossible to make reasonable inferences about phases of business cycle.Compared with the existing domestic research,the Markov dynamic bi-factor model constructed in this paper establishes a relationship between the leading index and the coincidence index,which could not only reasonably infer the periodicity of the coincidence index,but also could examine the joint transfer dynamics between the leading index and the coincidence index,and use the leading index to forecast the macro-economy in the near future.4.The construction of the best early warning indicator of China.Based on many leading indicators,the macroeconomic early-warning index is constructed by the NBER method,which is not based on a rigorous statistical model.The macroeconomic early-warning index constructed is not necessarily optimal because it ignores the criteria for measuring the predictive ability of leading indicators.The model-based macroeconomic early-warning index is based on the strict assumption of parameters distribution.Once the real data violates these assumptions,the validity of the model cannot be guaranteed.To this end,using the research methods of related literature,this paper first uses the ROC curve to examine the predictive ability of a single leading indicator in different horizon.Then,we use the Copula function to decompose the joint distribution function of all leading indicators into two separate parts of the marginal distribution and the Copula function.First,the non-parametric econometrics kernel method is used to estimate the marginal distribution of individual indicators,thereby reducing the probability of misspecification of the model.Second,using the Vine-Copula model to describe the dynamic dependencies between different indicators to avoid the"curse of dimensionality ".Finally,construct the macroeconomic early warning index of China by using the functional expressions of the joint distribution function of the leading indicators under the conditions of the occurrence and non-occurrence of predicted events.5.The real-time forecast research of China’s economic growth rate.Gross domestic product(GDP),as the most of important indiactor,has received widespread attention from policy makers and market participants.However,statistical departments usually publish quarterly and annual GDP data.At the same time,there is a certain time lag in the publication of data.In view of the shortcomings in the existing domestic literature in this field,combined with China’s actual data publishtion,this paper adopts a mixed frequency dynamic factor model to forecast the year-on-year growth of quarterly GDP in real time,and through the weights of the "news" to reflects the relationship between the new released data and forecast revision,and then measure the contribution of the each(group)new released indicator(s)to the forecast revision.