Modeling of interbank call loan rates with jump diffusion

• Main Author: Mariano, Redentor S.
• Format: text
• Language: English
• Published: Animo Repository 2004
• Subjects: Banks and banking; Mathematical models; Bank loans; Business Administration, Management, and Operations
• Online Access: https://animorepository.dlsu.edu.ph/etd_doctoral/95; https://animorepository.dlsu.edu.ph/context/etd_doctoral/article/1094/viewcontent/TG03933_F_Partial.pdf
• Institution: De La Salle University

The present study provides a model for interbank call loan rates (IBCLR) through jump diffusion modeling and link jumps with monetary interventions. Modeling the changes in IBCLR involves the evaluation of three sets of models: pure diffusion models, jump diffusion models, and other modeling approaches: stochastic volatility, ARCH, and regime shift models. From the four criteria Akaike info criterion, Schwarz criterion, standard error of regression and sum of squared residuals used in Generalized Methods of Moments (GMM), it can been seen that the jump diffusion model with positive and negative jumps fits better than the other models. The present study has proven that the BSP intervenes in the interbank market through the reverse repurchased agreement (RR/P). The BSPs changing of the RR/P rate automatically changes the IBCLR. This paper provides two arguments to link the changes in the RR/P rate with the jumps in IBCLR. First, all the changes in the RR/P rate are reflected in changes in IBCLR. There are a total 175 jumps in IBCLR and 470 changes in the RR/P rate. The jumps in IBCLR and the changes in RR/P coincide on 50 out of 175 days. Second, the BSP sets RR/P after the start of the interbank trading session. Thus, about 25.2% of the total variance in the IBCLR is explained by IBCLRs relationship with the RR/P rate.