2021 Volume 64 Pages 46-70
In this paper, we propose a stochastic interest rate model with a Markovian regime-switching property in order to evaluate the interest-rate risk of various kinds of asset and liability portfolios synthetically. In the model, parameters in the short rate process depend on a latent state, and the state transits between some finite regimes with a Markovian property. The Monte Carlo simulation is used to generate many sample paths of the future short rates and evaluate the interest rate risk numerically. Since the term structures of interest rates in future are derived by using the no-arbitrage pricing method based on sample paths, interest rates with various maturities can be used for modeling assets and liabilities. Numerical examples show that the future interest rates will not proceed to decrease deeply to negative values, rather they will behave as if they had a lower limit, and that they will keep their present trends, while there exist some small probabilities under which the extremely upwarding interest-rate scenarios will happen. Additionally, from applying of this model to evaluating the interest rate risk of the non-maturity deposits, we obtain that the present increasing trend of the volume of the deposits will be kept in future, while that the drastic decrease of the volume will happen with small probabilities.