Modeling on interest rate behavior of RMB: Comparative research based on Vasicek, CIR and CKLS model

Under the background of the marketization of RMBpsilas interest rate, the capacity to capture the change of interest rate will be the core ability of financial institutions. After comparing the features of classic single factor interest rate models, Vasicek, CIR and CKLS model, this paper processes the three continuous time differential equations into discrete forms by Euler equation. The authors respectively compile SAS programs of LS and GMM for these three interest rate model, and obtain their parameter values, by taking 1271 daily closing quotation data of the Chinese inter-bank collateralised repo rate within 7 days (R007). The empirical results show that the Vasicek and CIR models are much more suitable to explain the behavior of RMBpsilas interest rate, other than the CKLS model, which is the most complicated model within the three interest rate models. Furthermore, the explanation power of CIR model for RMBpsilas interest rate is the strongest in our studies.