Testing For The Indeterminacy In China's Monetary Policy And Forecasting The Policy Effects

Research on monetary policy has been and still will be a hot spot in the realmof macroeconomic and macro finance. By applying various empirical analysis methods,scholars had conducted both positive and normal analysis to studies about rules, execu-tions and effects of monetary policy. As a result, this field has become more and morecomplete. However, as the situation of economy changes throughout times, some draw-backs in traditional macroeconometric methodology had arisen gradually. For example,as Lucas'critique points out, the parameters in simultaneous equations are invariantto the shift of policy regime, thus this kind of model cannot simulate how economicvariables react to policy shifts under different regimes. Sims'critique also pointed outthat, the specification of the parameters in traditional model heavily relied on some ex-perience which is hard to test or even untestable. Possibly, utilizing this kind of modelwould well record the past, while out-of-sample forecasts may be misleading and incred-ible. These existing flaws and problems significantly propelled both the development ofmonetary economics and the evolution of econometric theory.Given the extant literatures about monetary economics, the ones which applyingVAR scheme into empirical research take up a significant part. Similarly, VARs arewidely used in empirical analysis in China. The reason is that scholars do not needa lot of premises and empirical specifications, which bring a lot of conveniences forChinese researchers. Along with the advantages it brings are the shortcomings whileusing VAR model. First is about the availability of data. For instance, if we construct aVAR model for China economy, sometimes we will encounter problems such as qualityof data and variables lacking of explanatory ability. Secondly, VAR methodology willstill be criticized according to Lucas'critique. In addition, we cannot analyze welfareimplications only in virtue of VAR model. At present, Dynamic Stochastic GeneralizedEquilibrium (DSGE) model is an achievement of solving these problems at current stage.For a period of time, the advantages of DSGE model lie in the theoretical area;whereas the harbinger work by Smets and Wouters (2003) has greatly improved theforecasts and analysis generated by DSGE in virtue of Bayesian theory. Meanwhile, DSGE model can better depict the dynamics of economic system, thus widely beingaccepted by monetary authorities of different countries and becoming a main analysistoolkit for monetary policy. Under this circumstance, some scholars in China has be-gun to do some insightful research to make this model compatible in studying Chineseeconomy.Based on extant relative literature and research, this article treated monetarypolicy-leading indeterminacy as the theme in virtue of positive and normal analysis.Consequently, an analytical model for monetary policy is constructed based on determi-nacy in order to obtain forecasts and welfare analysis result. The aim of this analysisis to find out how can monetary policy affect macroeconomic dynamics and ultimatelyprovide beneficial references for monetary authorities.As a start point of this research, this article employs a likelihood function basedDSGE methodology, which is proposed by Lubik and Schorfheide (2004) to explore theproblem"whether the effects of China monetary policy are consistent with determi-nacy". In order to achieve this goal, three variables of most interest to policy makers– real GDP growth, inflation and the nominal interest rate– are chosen to constructa structural model. Besides these endogenous variables, this paper also introduces ex-ogenous sunspot shocks into the model. The sample period begins at 1992 Q1 and inthis period, interest rate is assumed to be the intervening variable. In order to separatedifferent impacts on monetary policy and fundamental variables derived from sunspotshocks and indeterminacy, this article adopts a ceteris paribus method and constructed4 different controlled groups for priors (i.e. Prior1 for all, Prior2 for sunspots, Prior3for indeterminacy, and a prior specification for determinacy). On the basis of this spec-ification, this article analyzes the propagation of shocks.The results about indeterminacy versus determinacy suggest that, According tothe past sample data in China, monetary policy induced"animal spirit"driven economicfluctuations, or we say indeterminacy exists in the economic system. The indeterminacyhas led sunspot shock have influence on endogenous variables, for instance, it raises GDPgrowth and inflation; however, it is not a major influential factor in changing the behav-ior of endogenous variable. Further comparison shows that indeterminacy significantlyalters the propagation mechanism of fundamental shocks– monetary policy shock, de-mand shock and supply shock. As for policy shock which is the mainly concerned,"pricepuzzle"arose under the circumstance of indeterminacy. That is to say, when monetaryauthority tightens monetary policy, this implementation would raise inflation rate. Con-sidering all the results displayed above, the effects of monetary policy would be affectedby indeterminacy, thus central banks need to react to the"distortion effect"generatedby indeterminacy; and, in the process of decision making, they should consider a specificpolicy to rule out indeterminacy.Based on the above results about indeterminacy and ruled out indeterminacy, thisarticle tries to develop a toolkit which involves forecasting the effects of monetary policyand analyzing the welfare implication of different regimes. Hence, ideas from Del Negro and Schorfheide (2004) are borrowed in this paper, and a tri–variable DSGE–VAR (4)model, which can be used for primary analysis, is constructed and estimated via usingChina economic data. Forecasts have been significantly improved in this new modelvis-`a-vis tradition models. A more important feature of this new model is that, wecan analyze the welfare implication of regime shifts from"easing"monetary policy to"moderately tightening", which correspond to the"soft–landing"period in 1996 and thetime before Subprime Crisis, 2008, respectively.By combining the results obtained from indeterminacy and welfare analysis, thisarticle suggests that central banks should implement a active other than a passive policyregime, ad hoc, aggressively react to inflation and implement stabilized policy. On theone hand, this type of policy can significantly lower the possibility of indeterminacy, andlead the economy system reach a unique equilibrium. After ruling out indeterminacy,the uncertainty of monetary policy would be lowered, and this is good for monetaryauthority to forecast and evaluate a policy. On the other hand, if aggregate welfare isassessed by the volatility of production gap and inflation gap, an aggressive monetarypolicy will reduce the volatility of these two gaps and increase the level of aggregatewelfare. Considering these results, this article argues that if being appropriately utilized,DSGE model can offer a sound analytical tool for policy-making and a reliable referencein the decision making process.