Application Of MIDAS Models In Bond Markets

The Mixed Date Sampling(MIDAS)technique provides a general framework for the data analysis of different frequencies in econometric models.With the parameterized weight format,MIDAS can avoid the loss of information without the problem of overparameterization.The impact of macroeconomic policies cannot be ignored for the bond market.It is of great practical significance to analyze the impact of macroeconomic policies on the financial distress of listed companies,the volatility of high-yield bond market and the correlation between the stock and bond markets.In this paper,MIDAS is used to study the impact of macroeconomic policies on different aspects of bond markets from the following three parts.Firstly,combining Aalen additive hazard model with MIDAS,Aalen-MIDAS model is established to analyze the impact of corporate financial and macro factors sampled at different frequencies on the financial distress of China’s listed companies.It is found that 7 corporate financial indicators,consumer price index(CPI),producer price index(PPI)and corporate bond credit spreads have significant influence.The consumer price index is a protective factor,which can reduce the probability of financial distress.The ROC curve analysis shows that,compared with Aalen model,Aalen-MDIAS model can improve the performance of the prediction for financial distress.Secondly,GARCH-MIDAS model is used to analyze the impact of economic policy uncertainty,monetary policy,tax,government spending and financial regulation on the volatility of U.S.high-yield bond market.The empirical results show that the model can effectively extract the long-term components of the volatility.Economic policy uncertainty,monetary policy,tax,government spending and financial regulation are all important drivers of high-yield bond volatility.By establishing the same model on subsamples,the results are consistent with the whole sample,which ensures the robustness.Finally,DCC-MIDAS model is used to analyze the impact of macroeconomics on the correlation between U.S.stock and bond markets.In the first step,GARCH-MIDAS model is used to describe the impact of financial regulation on the long-term components of stock and bond volatility.In the second step,the dynamic and long-term correlations are obtained based on the DCC-MIDAS model and the impact of macro policies on correlation is studied by multiple regression analysis.It is found that the correlation between U.S.stock and bond markets shows negative effects.The monetary policy,entitlement programs and the trade policy show positive effects on the correlations,the positive changes in these policies can make the two market increasingly interlinked,while fiscal policy,national security and financial regulation show the negative effects.The robustness of the result is ensured by the use of the autoregressive term of the macro factors in the same model.