Nonparametric Analysis Of Interbank Offered Rate

With the rapid development of financial marketization and globalization,studies on interest rate and its term structure models are becoming more and more important as one of key economic indicators.Compared with the domestic financial market,foreign financial markets gets long-term development and corresponding research becomes a relative perfect system.Due to reform and opening-up policy and the interest rate marketization,China's financial market has gotten rapid development,but the study of short-term interest rates and interest rate structure model is still relatively lacking.Most of researchers focus on parameter estimation to analyze the model and to give specific function form,which cannot accurately estimate drift and diffusion terms for interest rate models.To some extent parameter estimation may not describe the dynamic characteristics of interest rate accurately.In present nonparametric estimation methods are proposed and used to analyze the dynamic characteristics of interest model,but there are only precious few literatures.In this thesis short-term interest rate model is first used to analyze Hong Kong interbank offered rate(Hibor),and basic features during the last ten years are revealed.Notice that stationarity of the whole data process may not be guaranteed,so nonparametric statistical methods are established.Based on this model,functional nonparametric estimations of drift and diffusion terms and the local time of process are obtained by using Bandi's method.Empirical analysis shows that the data has a different feature before and after 1999,and that spatial density of this process appears to be bimodal.