| With the rapid development of China economy,the circulation and volume of Treasury bond market,especially in the last 10 years,has realized the explosive growth.Investors gradually increase the proportion of investment in Treasury bonds.The research and application of Treasury bond investment strategies enjoy a lot of concern.Term structure of interest rates is extremely important reference factor in asset pricing,portfolio management,so then term structure of interest rates has been a hot topic.Currently,research in this area can be divided into static fitting and dynamic prediction of the term structure.This paper selects the Shanghai Stock Exchange,the Shenzhen Stock Exchange and the Inter Band Market from January 2010 to December 2014 monthly Treasury bonds trading data,as inter-sample data,and then use Nelson-Siegel model,Svensson model and B-spline model to statically fit these inter-sample data.Fitting results show that the Nelson-Siegel model in China Treasury bond market have batter application effect.Parameters obtained by static fitting constitute a time series of the state factor vector in Dynamic Nelson-Siegel model.This paper use “Diebold and Li two-step method” to forecasting the term structure in next time point.By backward iteration method,monthly predictions of outof-sample data from January 2014 to December 2015 can be achieved.Based on these predictions,this paper mainly adopt the expected yield curve strategy and the duration vector immunization strategy to conduct empirical study.Finally,this paper find that by using the expected yield curve strategy to actively manage bonds portfolio,investors can effectively avoid volatility,realize investment return which is higher than the market basic return;also find to a certain extent,the immune effect of duration vector immunization strategy is better than the traditional immunization,but the greater amount of computation and complexity affect the usefulness of this strategy. |
PDF Full Text Request