Interest rate models in continuous time for small and open economies

This paper sets out to investigate diffusions in the short Norwegian interest rate. The main objective is to find stochastic differential equations (SDEs) that fit the partly observed interest data series. The SDE approach is chosen due to two observed market features. Firstly, the Norwegian market is an open and small economy inducing national as well as international shocks. Secondly, the level series is probably not stationary. The paper employs efficient methods of moments (EMM) to fit a wide range of interest rate models with extentions using versions of SDEs. The paper finds that all international one-factor interest rate models are rejected mainly due to the leptokurtic behaviour of the time series. Two-factor models, allowing for stochastic volatility, improve the fit. However, only models specifying flexibility near the origin of the diffusion function reports success. Any feedback extensions are rejected. In small and open economies, the interest rate diffusions need stochastic volatility coupled with empirical sensitivity to flexibility near the origin of the diffusion function. The more advanced HJM and Libor Market models (two-factor models) must be applied to value interest rate derivatives.