The Macroeconomic Factors And The Term Structure Latent Factors

Term Structure of Treasury bonds depicts the relationship between spot rate and remaining maturity. In a complete and effective bond market, bond rate as a benchmark interest rate is not only the premise for financial instruments pricing in capital market pricing, but also a very important factor for participants to judge the trends of the future economy and make the right decisions. Therefore, to study the factors that affect the movements in term structure of interest rates in bond market and what the relationship between these factors and macroeconomic factors not only has important theoretical significance more of its practical significance. However, we need an appropriate model to fit the term structure of interest rates before this study. So, this paper provide a no-arbitrage dynamic Nelson-Siegel model (AFDNS) to fit the spot rate in our inter-bank bond market, the model is a dynamic extension of the Nelson-Siegel model and is not only maintaining an ideal fit effects of the original model, but also taking no-arbitrage analysis into account, no-arbitrage analysis is a basic method in the financial study and also one of the most basic assumptions in risk-free arbitrage pricing of financial assets. So the use of the extended model to fit the term structure not only maintains the strict no-arbitrage conditions of the affine class model and gets analytic solutions of the bond pricing, besides the estimation results also show that the model can well reflect the dynamic characteristics of Treasury term structure in real market. After analyzing these characteristics, we found that our national spot rate curve volatile obviously, and change discontinuously, especially short-term interest rates.While the long-term interest rates move horizontally.Then, on the basis of this model, we further use the Kalman filter to obtain three main factors which can explain the changes in term structure.Those three factors are called "level factor""slope factor" and "curvature factor", we not only give theoretical explanation for these three latent factors but also do some relevant empirical analysis.The result show that the level factor represents the long-term interest rates, the slope factor represents short-long spreads, the curvature factor represents medium-term interest rates. Also the empirical results show that:level factor shows high persistence and less variance. From the perspective of the economic, it means long-term interest rate indicates relatively stable and the economic cycles have little impact on the long-term trend. While the slope factor has lower persistence and the curvature factor has the biggest variance.Finally, in order to study the relationship between Term Structure and macroeconomic, we first do correlation analysis among level factor,slope factor, curvature factor,industrial production growth rate, broad money supply growth rate,and the CPI. We found that the correlation between slope factor and three macro factors are more obvious than the level factor and the curvature factor. The slope factor is highly negatively correlated with the broad money supply growth rate and also highly positively correlated with the CPI. This result suggests that the CPI which can represent inflation factor and the broad money supply growth rate which can represent the monetary policy factors have a more significant effect on our Interest Rate Term Structure. In addition, we can also conclude that:the slope factor can be used as an indicator of the state’s macroeconomic performance metrics, reflecting changes in monetary policy and inflation. When the short-long spreads become big, it means that the market participants predict a big press on CPI, and the country will take loosen monetary policy, the same to the opposite condition. Considering the curve contains the information about the trend of the CPI, It is useful for the country to make policy. We then build a VAR structure that contain six factors and do impulse response analysis under this structure. The results show that there exits intertwined relationship between the latent factors and macroeconomic factors, and the shock response effect and correlation analysis conclusions are basically the same. In addition, we have chosen y (3m), y (60m), y (120m) maturity yield which represent short-term, middle-term and long-term curve and do variance decomposition to analyze how latent factors and macroeconomic factors influence the yield curve and to evaluate the importance of different impact. Variance decomposition results show that variance contribution rate of macroeconomic factor for predicting short-term yield curve is not strong, but the contribution rate gradually increased with the forecast period increasing. Variance contribution rate of slope factor is always on the dominant state, while variance contribution rate of level factor and curvature factor has been low, and the variance contribution rates of slope factor for the long term interest rate are higher than short-term interest rates.By introducing no-arbitrage conditions, we innovate the DRA model and improve and expand the original methods in this paper. The AFDNS model not only makes the model fit better, but also maintains the strict no-arbitrage conditions of the affine class model. Both with a layer of a solid theoretical foundation of economics, but also has a high practical value, can be considered as the country’s benchmark interest rate derivatives pricing models. Although some empirical results have higher accuracy than the original model, but unfortunately there are some empirical estimation result does not extend the original model. Therefore, we have yet to conduct a more comprehensive evidence to draw a regular conclusion of the same issue.