| The outbreak of the global financial crisis in 2008 and attendant economic recession attached great importance and consideration to the significance of leverage about stability of country's economy.In addition,in recent years,China's macro-leverage has been rising rapidly,the excessive debt of companies,households or governments and the price bubble especially the housing price bubble are the two main aspects of financial risk at this stage.Therefore,grasp the internal correlation among leverage ratio,asset prices and economic growth accurately can not only predict and control economic growth,but also provide ideas for deleveraging and effective prevention of financial risk.Based on the situation of economic growth and related research,the relationship among them may have nonlinear correlations and time-varying characteristics Thence,Due to the mixed quarter and month data,this study innovatively fuses the Mixing Frequency VAR(MF-VAR)and Markov-Switching VAR(MS-VAR)to construct the Mixing Frequency Markov-Switching VAR(MF-MS-VAR)model and study the time-varying characteristics and structural differences among leverage,asset prices and economic growth from the level of leverage aggregate and structure,it also discusses the asymmetric effects and transmission mechanism among them in detail due to mixed Granger causality test and analysis of the impulse response,as well compares the performance of the three models from fitness,dynamic prediction and statistically aspectsThe empirical research results show that:(1)The MF-MS-VAR model at leverage aggregate level is about 86.081%and 20.665%better in economic growth fitness than MF-VAR and MS-VAR models respectively,and the dynamic prediction errors are reduced by about 18.291%and 6.473%,the D-M statistics also reveal there exists significant difference among prediction methods,that is,the MF-MS-VAR model performs better;(2)the average effect among leverage ratio,asset price and economic growth is time-varying,while the coefficient of mutual influence is not dependent on state transformation,which reflects the stability of internal correlation of economic variables;(3)the MF-MS-VAR model can well identify the state characteristics,and the rate of total leverage and economic growth are featured by counter-cyclical characteristics,while the growth rate of total leverage and asset price are characterized by the heterogeneity of asset types,and the state transition probability shows that there is a decrease trend in leveraged growth and the total leverage macro can be dominated;(4)the mixing Granger causality test shows that the low-frequency variables and low-frequency variables,high-frequency variables and high-frequency variables are mutually Granger internally and transmission time are lagging by one period,while the Granger causality between high-frequency variables,namely stock and property price,and the low-frequency variables,namely economic and leverage growth rate,is asymmetric and show heterogeneity in asset types;(5)the action path,convergence and influence degree of total leverage,asset price and economic growth are time-varying and non-linear,and is subject to the influence of asset types;(6)the time-varying effect of leverage structural analysis shows that macro sector leverage exists significant structural differences from qualitative dimensions,the average effect and state transition characteristic among variables are time-varying and sector heterogeneity,in particular,the government and corporate leverage growth is more likely towards a low state,while residents sector of leverage growth tends to rise Based on the conclusions,the study finally proposes specific recommendations from five aspects pay attention to the linkage and dynamic control of the asset market,firm the financial supervision and credit system construction,promote the asset and liability side deleveraging process,accurately demonstrate the direction of structural deleveraging and strive to stabilize the market economic environment. |
PDF Full Text Request