A Study On The Liquidity Spillover Effect Based On Multivariate Garch Model

Along with the global economy integration process is accelerated, the world economy is blustery capriciousness. Our country following after the reform and opening, goes out of a extraordinary30years, after ninetys began to walk marketization road, GDP is take the double-digit annual growth rate high speed growth, making the world countries alike. However, in the big stride forward in the process, the road is bumpy, the1998Asian economic crisis,in2008the United States detonated upon a credit crisis of the global financial crisis, are profound to China-the emerging economies gave us a great test. Meanwhile, especially after the financial crisis, liquidity risk transfering effect between countries in the exposure to China and the countries all over the world, there is no guard against cases, the lack of liquidity to the countries all over the world economy as like dominoes fall in succession.This paper is based on the background, and make a research about the securities market liquidity spillover effect putting forward to China, Japan, the United States and Britain the four big economies in the world, we are considering the liquidity measures, sample data, and a certain model, from the level of liquidity and liquidity risk spillover effect two point of view, to measure the liquidity of spillover effect between them. The liquidity level spillover effect of characterization, we build VAR vector autoregressive model, and on this basis granger causality test, impulse response analysis and variance decomposition analysis the spillover effect between the variables. Liquidity risk spillover effect of characterization, we from liquidity index second order moment estimation, the final choice of the multivariate GARCH model representative BEKK, DCC-GARCH, GO-GARCH three model, and in the high dimension cases on the liquidity risk of empirical analysis, from the transverse and longitudinal before and after each model of crisis and between the conclusion are analyzed and compared, on the one hand, application of multivariate GARCH model analysis liquidity risk spillover effect, on the other hand, multivariate GARCH model of three kinds of comparison analysis and comparison, in order to further promote existing dimension disaster situation multivariate research.The empirical study after get the following conclusion, before the crisis between China and the United States securities market liquidity are granger causality, the China securities market as the Japanese stock market one-way granger British relationship; After the crisis, DIL_FS100, DIL_RJ225are DIL_HS300granger reason, the other between market liquidity and didn’t exist significant causal relationship. The market for a standard deviation from their own new interest response, always can quickly remove to zero, the other variables from a standard difference innovation response, they are zero as the center and concussion, and after ten days after returning to zero. Whether before or after the crisis crisis, countries in the securities market liquidity wave contribution factors for themselves the market itself. The risk of fluctuations, crisis after a higher degree of volatility persistence, and by countries in dealing with the crisis of the economic policies of the consistency, between the countries liquidity risk spillover effect are performance for a certain degree of positive correlation, and the SP500and FS100is positively related to the highest degree, in principle, the mobility of systemic risk still exist.