Empirical Research On The Liquidity Of China’s Bond Market

To measure the overall liquidity of China’s bond market in a quantitative way, this paper tries to extract the information of market liquidity change from pricing noise according to the relation between arbitrage capital and pricing error, from the aspect of market efficiency. We firstly calculate the market spot rate through bootstrapping method and build the dynamic term structure model based on AFNS model using the data of Inter-Bank market bond prices; Then, we estimate the model parameters through Kalman Filter method and Maximum Likelihood Estimation, and get the theoretical interest rate of treasury bond; At last, we take the difference between theoretical rate and market rate as market pricing error, and extract the noise representing market liquidity. To test the validity and stability of the liquidity noise, we take DNS model as the control group and make a comparison of different extracting methods. Additionally, we regress the noise on several market variables to retest the rationality of the noise as the measure of market liquidity.The empirical results show that, the noise extracted from AFNS model can mark the market liquidity abnormality effectively, and its feature of changing with policy and other liquidity indicators makes it a good measurement of liquidity change in treasury bond market, and even in overall bond market.