As one of the most popular derivatives to hedge interest rate risk, the variation of interest rate swap spread has been studied since its advent. Nevertheless, the variables in theory are regarded as determinant risk factors showing limited explanatory power. To investigate further, we conducted the cointegration test on each pair of variables which are considered in the financial and macroeconomic sense, and extend the classic Generalized Autoregressive Conditionally Heteroscedastic (GARCH) model by combining the Error correction model (ECM). Our testing results suggest that the changes in the swap spread negatively correlated with the slope of the yield curves of Treasury Securities and positively correlated with the implied volatility of...