Continuous time counterpart of vector auto regression of term structure dynamics with jump diffusion processes

Abstract

The understanding of bond yields is important for several reasons such as forecasting, monetary and debt policies, derivative pricing and investment decisions. The existence of a huge literature on this subject is a clue on how lots of researchers are trying to improve modeling of bond yields. In this paper two of such improvements are presented and discussed. The first improvement to be discussed is by Ang and Piazzesi (2003) who have found out that inclusion of macro variables to the affine term structure models provides a better fit empirically. The second improvement by Das (1998) also provides a better fitting term structure model by modeling the underlying state variable following a jump process.