| This paper tests the determinants of475MTNs’ credit spread in2011-2013.Using three bond liquidity measures, trading volume, turnover and bid-ask spread, Ifind liquidity is one essential component of credit spread, but not as important ascredit. Generally, bid ask spread, trading volume and turnover are all significantliquidity proxies. Hausman test gives both trading volume and credit rating areexogenous, meaning they are not determined by system itself, so there is no need torun simultaneous equations. In single liquidity measure regression, two methods (GLSand Fama MacBeth regression) are used to find the credit spread factors.Liquidity’s impact exists in most cases, the coefficient is distinct but the effect islimited, smaller than credit factors.Most macro pricing factors like CPI, M2growth rate, repo rate are omitted dueto multicollinearity. Only term premium (the difference of10year treasury rate and2year treasury rate) is significant.Bond specific information like Debt/Asset is most essential in majority.1gradecredit rating’s impact on spread is about24BP by Fama MacBeth approach. Othervariables like coupon rate, CFO/sales, profit margin are less significant. |
