Empirical Research On The Liquidity Premium Of Corporate Bond Market In China

Ample liquidity is the prerequisite for normal operation of the financial market. In recent years, especially after the 2008 financial crisis, the European and American developed country market regulators and market participants find that although the market reached a scale and liquidity of the corporate bond market is still not optimistic. They even think that liquidity risk has been over credit risk and interest rate risk, become the greatest hazard to the corporate bond market. In our country, the government has put developing the corporate bond market as a strategy, but due to various reasons, China’s corporate bond market development is relatively slow, liquidity problems is more serious than developed markets. The research on dimension effect of liquidity measuring of corporate bond market, liquidity measuring of corporate bond market on different data frequency and liquidity premium helps to reveal the influence of the liquidity premium on corporate bonds pricing mechanism, has important academic and practical significance to the development of corporate bond market in China. Using data in Chinese corporate bond market, the paper study the dimension effect of liquidity measuring of corporate bond market, liquidity measuring of corporate bond market on different data frequency and liquidity premium.Firstly, the liquidity has the multidimensional connotation such as volume, speed and cost. Respectively using the intraday and daily data in Chinese corporate bond market, using different liquidity measure calculation case, different time interval, different trading places, and correlation coefficient analysis and the correlation between the difference statistical test method, this paper empirically studies the dimension effect of liquidity measuring of corporate bond market. The results show that there are significant dimension effects of the liquidity measuring in Chines corporate bond market. The relationship is weak among liquidity measures of volume dimension, speed dimension and cost dimension. The relationship between multiple dimension measure and three single dimension measures are stronger. The strongest relationship happens in the measures reflecting the same dimension.Secondly, using different liquidity measure calculation case, different time interval, different trading places, sign test, average errors analysis, Theil(1966)’s U statistical tests, the correlation coefficient analysis and other methods, this paper empirically studies the relationship of the liquidity measuring on intraday and daily data in Chinese corporate bond market. The research results show that there are significant differences between the liquidity measuring based on different frequency data. The liquidity measures on low-frequency data can’t predict the liquidity measures on high-frequency data. From the point of average errors, the time interval based on low-frequency data average overestimate about 1 days or so than the time interval based on high-frequency data. The Roll(1984) index based on low-frequency data significantly greater than the Roll(1984) index based on high-frequency data. The Amihud(2002) index based on low-frequency data significantly smaller than the Amihud(2002) index based on high-frequency data. The single-dimensioned liquidity measures on low-frequency data tend to underestimate liquidity level. The multiple-dimensioned liquidity measures on low-frequency data tend to overestimate liquidity level. Based on different data frequency of liquidity, however, there is no significant difference between "relative" measures. And there exist significant strong time-series and cross-sectional correlations between liquidity measuring on the two different data frequencies. This indicates that there is a stable relationship between the liquidity measures based on two kinds of data frequency. Conclusion shows that, on condition that we are unable to get the high frequency data, or the high-frequency data costs too much,we can use the liquidity measures on low-frequency data to give the correct relative order of level of liquidity of corporate bonds, and level of liquidity measured by low-frequency data have a solid relationship between real level of liquidity. But the differences between the liquidity measures on low-frequency data and the real level of liquidity are significant. Therefore, when involve the theoretical research and practice problem which need accurately mearsure the level of liquidity of corporate bonds, we should Prior use the high-frequency data.Finally, the levels and risk of liquidity has important theoretical and realistic significance for corporate bonds, the company’s cost of debt financing determining, the corporate bond market pricing, the corporate bond market regulation. Based on linear multifactor pricing model, considering both Fama-French stock three factor and bond two factor conditions, we use the Fama-MacBeth method analyzes the corporate bond market liquidity premium. Studies have shown that in addition to the default premium and term premiums, the corporate bond yields include both the level of liquidity premium and liquidity risk premium. In normal times, based on the liquidity risk of liquidity premium is smaller. But during the period of credit and/or a liquidity crisis, There is a positive feedback loop between liquidity and default, and credit risk can lead to liquidity premium and liquidity risk will lead to default premiums.