Construction Of The Financial Index And The Asymmetric Effects Of Monetary Policy

Monetary policy is one of the important components of macroeconomic policies and it couldimpose substantial effects on macroeconomic variables such as price level and economic growthby way of a series of variables conduction. So, it has been a significant research direction to studythe effects of monetary policy on price and output. Basing on the hypothesis of monetaryneutrality and the theory of monetary policy transmission channels, there appear a growingnumber of studies on the asymmetric effect of monetary policies. These effects imply that thesame monetary policy, under different economic conditions, has a varying degree of impacts onthe output or price. In this paper, we design three financial indexes, which contain core inflationrate, financial condition index, prices early warning index, and analyze the asymmetry ofmonetary policy from different perspectives. The main works are as follows:Firstly, previous studies on the effects of China’s monetary policy are mainly based on thevector autoregressive (VAR) model focusing on Consumer Price Index (CPI), which makes themodel lack of sufficient information and causes the aimlessness of price effect of monetary policy.On the one hand, factor-augmented smooth transfer of regression (FAVAR) model, with theadvantages of introducing abundant information but bringing no difficulty of parametersestimation, could be adopted to make up the deficiency of the VAR model in terms ofmacro-economic variables selection. On the other hand, core inflation rate is defined from theview of monetary policy, representing a long and potential price change. But the measure methodsin presence on the inflation, statistics or based on models, are mainly discussed focusing on CPI.Although CPI reflects the main part of the rate of inflation, it repents not all. So, it is necessary tofind more price information than CPI to obtain core inflation rate.Adopting the dynamic factor model, we firstly employ six composite price indexes to getcore inflation rate, which could be better to reflect a long and potential price change. Core inflation rate is depicted using logistic smooth transfer vector auto regression (LSTAR), and thenits nonlinear dynamic feature is analyzed. Then we employ the dynamic factor model to extract afew common factors from many variables selected from Chinese macro economy, and move on toconstruct FAVAR models (including macro monetary policy tools and the common factors)according respectively to CPI, core inflation rate and the CPI typical indexes. And we adopt theimpulse response function to describe the dynamic effect of the monetary policy tools on eachvariable. The empirical results show that compared with the VAR model, the effect of monetarypolicy can be reflected more fully and effectively in FAVAR model. The monetary authoritiesshould focus their attention on core inflation rate rather than the CPI. More than that, because ofthe difference of CPI typical indexes to impulse response function from the money supply shocks,the authorities also needs to keep eyes on the heterogeneity of the monetary policy effects on theCPI typical indexes. After that, core inflation rate is taken as transfer variable to build the logisticsmooth transfer vector auto-egression(LSTVAR), with which we analyzes the asymmetric effectsof monetary policy. The empirical results suggest that, under different inflation conditionsdetermined by the core inflation rate, monetary policy has asymmetric effects on the output andprices, and that the monetary policy under low inflation condition are more effective than thatunder high inflation condition.Second, a series of international economic events, like the subprime mortgage crisis,European debt crisis, have highlighted the severe impact of the financial markets on the realeconomy. Financial conditions index (FCI) can reflect the operational states of the financialmarkets. In view of the disadvantages of monotonous measurement methods or lack of abundantinformation in the previous researches on China’s FCI and because of the lack of deep analysisabout FCI forecasting ability on the macroeconomic variables, it is necessary to build a moreeffective measure method on China’s FCI construction, based on this, to analyze the interactionbetween financial market and macroeconomic variables.11vital financial variables are firstly selected from the typical macro variables of China’seconomy which could be employed to represent interest rates, exchange rates, stock prices andreal estate prices, and so on. Then, dynamic factor model is applied to extract their commonfactors, which is used subsequently combined with VAR model to construct China’s financial conditions index (FCI) that could be used to represent China’s financial market condition. Then,China’s financial market and its relevant measurement with macro economy are tested from theprospective of frequency domain and time domain, respectively. The empirical results show that,in the long run, financial variables have a strong ability to predict on the real economy, leading thechanges of macroeconomic variables. However, in the short run, the characteristics of thevariability show up when the effect of China’s financial market on macro economy is consideredand tested. Therefore, different developing cycles are needed to consider by the government toevaluate the relevance of financial market and macro economy, so as to cope better with theimpact of the financial market on China’s macro economy. After that, FCI is taken as transfervariable to build the logistic smooth transfer vector auto-regression (LSTVAR) including FCI,output and price, with which we analyze whether the financial market reflected by FCI has anasymmetric impact on the macroeconomic variables in different financial conditions. Theempirical results show that: when in a good financial condition, FCI has a significantly positiveimpact on the output. In a deteriorating financial condition, however, FCI has an apparentlynegative or harmful effect on output. After that, we take FCI as the transfer variable to build thelogistic smooth transfer vector auto-regression (LSTVAR), with which to analyze whethermonetary policy has an asymmetric impact in different financial conditions. The empirical resultsshow that: in a good financial condition, expansionary monetary policy has a significantlypositive short-term impact on the output, but invalid in the long run. In a deteriorating financialcondition, however, expansionary monetary policies bring about a significantly negative effect onoutput.Third, in most of the studies in China about the asymmetric effects of monetary policy basedon LSTVAR model, the selection of the smooth transfer function is according to the statisticaltheory, attempting to set all variables and their lagged variables as transfer variables, and makingstandard depending on the significance of nonlinear test and the significance degree. But this wayof variable selection leads to the lack of economic interpretation.Since keeping price steady is one of the most important monetary policy targets, we employthe dynamic factor model to design the prices early warning composite index using prices earlywarning indicators. Then, use the Markov regime switching model to discuss the asymmetry of the prices early warning composite index. The empirical shows that: when the fluctuation of theprices early warning composite index is grouped into low inflation condition and high inflationcondition, both the fitting effect and interpreting ability about the price fluctuation get increasingaccordingly. Finally, with this index as the transfer variable in the smooth transition function, andbased on the LSTVAR model, we move on to analyze the asymmetry of monetary policy underthe conditions of different economic inflation. The empirical results prove the existence of theasymmetric effect of monetary policy on both output and prices. And, comparatively, under lowinflation, the effect of monetary policy on output and price is positive and more significant.