Liquidity Spiral And Its Risk Contagion Mechanism Research

As the life of modern financial market,liquidity and its dynamic changes may directly affect the financial system stability and macro-economic fluctuation.In recent years,a series of economic and financial crisis have gradually highlighted the foundational position of liquidity in maintaining financial system stability.From 2007 subprime crisis,2010 European sovereign debt crisis to the 2015 stock market crash in China,all of these show that there exist serious imbalance in liquidity between supply and demand,excess and shortage.The local and a single form of financial fluctuation may become a systemic and diversified liquidity crisis under the liquidity spiral.Study about liquidity spiral and its risk contagion have gradually become current hot issues in financial research field,but they haven’t been deeply studied in the existing research.Therefore,the paper has deeply studied the formation mechanism,dynamic effects,states switching and risk contagion,etc.,and constructs risk supervision theory system of liquidity spiral,to further improve the level of financial system stability.Firstly,the paper further extends the theoretical content of liquidity spiral,based on the characteristics of financial system stability,liquidity spiral has been divided into three faces,including liquidity imbalance between supply and demand in macro-level,the dynamic correlation among the monetary liquidity,market liquidity and finding liquidity in meso-level,and interaction between "margin spiral" and "loss spiral".The formation mechanism of liquidity spiral has been explained by constructing model and numerical analysis.Meanwhile,on the basis of multi-dimension characteristics in liquidity spiral,the paper constructs the comprehensive measurement of uni-dimension liquidity and aggregating measurement of multi-dimension liquidity by using the principal component analysis(CPA)and distance-entropy method.The empirical study suggests that the overall liquidity in financial system takes February 2011 as a branch-line from Jan 2006 to Dec 2015,and shows an obvious two-stage feature,the overall liquidity volatility cycle has been extended after the branch,and the orientation of monetary policy has a significant effect on some characteristics of the overall liquidity,such as mean and volatility.Secondly,the paper has revealed the dynamic effects between multi-dimension liquidity in liquidity spiral based on financial system stability perspective:(1)The nonlinear causality relations among monetary liquidity,market liquidity and funding liquidity have been discussed via Bootstrap rolling window Granger causality test,and shows that the short-run stability of parameters among them are weak while their long-run stability are strong,the Granger causality relationships between these three liquidity have structural changes.(2)The dynamic relevant structures of multi-dimension liquidity have been exploited by use of DCC-MVGARCH model,and finds that the dynamic correlation between monetary liquidity and market liquidity is positive on the plateau while becoming negative during crisis,there exists a long-term and stable positive correlation between funding liquidity and monetary liquidity,the correlation between monetary liquidity and funding liquidity is relevant weak in crisis while its volatility increases in plateau.(3)According to the performance of liquidity spiral cycle,the cycle interaction mechanism of multi-dimension liquidity has been studied via spectral analysis,and suggests that the main cycle of market liquidity is significantly less than monetary liquidity and funding liquidity,all liquidity show a higher cycle consistency in low-frequency region,and have some asymmetry in cycle gain spectrum,while market liquidity is obviously lag behind monetary liquidity and funding liquidity.(4)The paper constructs two new indexes of investor sentiment and financial market stability,and analyzes the time-varying effect mechanism of investor sentiment and market liquidity on financial system stability with TVP-SV-SVAR model,and finds that the impulse response of investor sentiment in different time is gradually weakening year by year and has time-delay effect,and the dependence of financial market on market liquidity is gradually increasing.In addition,the time-varying effects of investor sentiment and market liquidity on financial market stability are different between the plateau and crisis.Thirdly,the paper reveals the state-switching mechanism of liquidity spiral from three facets,such as switching magnitude,switching time and switching types,and gives some empirical evidence in liquidity spiral process by taking subprime crisis and European sovereign debt crisis for examples.The state-switching mechanism of single dimension liquidity by constructing AR-MS-GARCH model has been exploited,and suggests that there exist significant switching effects for three dimension liquidity,and they have higher stability while moving within states,but we find the existence of asymmetry in the state-switching,and we also construct a new change-points detection index for market liquidity.Then,we discuss the state-switching mechanism of the liquidity system which consists of multi-dimension liquidity via TVTP-MS-VAR method,and shows that there also exists a significant switching effect for the liquidity system,and the transfer probability of liquidity system is time-varying and asymmetric,but the switching time is related to financial events or policies changes.Meanwhile,we construct two new indexes of stock market liquidity and futures market liquidity based on "time-scale"、"price-scale" and investors’ trading preference,and analysis the dynamic effects of monetary policy to two kinds of market liquidity with MS-VAR model,and the results show that the expansion of monetary policy is favorable to increase market liquidity while the shrinkage of monetary policies leads to decrease,and the impacts of M2 and rate differentiate from extent significantly in different financial markets and states.Finally,the paper explains the formation mechanism of liquidity risk from three facets of micro,macro and behavior level.Then,based on SIR model by considering infection delay time,a complex dynamics model of inter-bank liquidity risk contagion with the complex network theory,and the evolutionary law of liquidity risk contagion has been systematic analyzed by numerical simulation.By setting the liquidity shock distribution to general distribution,Beta distribution and IFR distribution family,the corresponding(weak)contagion probability of liquidity risk has been estimated.Additionally,a systemic stress analysis on both the whole bank system and a single bank(namely China Construction Bank)has been made,and the liquidity risk level of Bank of China by use of La VaR method has been measured by use of LaVaR method,and the macro-prudential supervision framework of liquidity risk and policy suggestions are provided.In sum,this study not only theoretically forms a systematic liquidity spiral analysis framework and follows the front direction of liquidity theory development,but also provides feasible models and approaches for the supervisors to improve the financial system stability,so it has higher theoretical value and practical significance.