Forecasting the Interest-rate Term Structure: Using the Model of Fong & Vasicek, the Extended Kalman Filter and the Bollinger Bands

Abstract:     
In this paper, we consider the issue of forecasting the interest-rate term structure and we present a solution. We apply the Extended Kalman Filter (EKF) to the Fong & Vasicek model to deal with the issue of computing the hidden stochastic volatility. We also introduce Bollinger bands as a variance reduction technique used to improve the Monte Carlo simulation performance. Our results suggest that the forecasting technique using the unobservable component approach (EFK) to obtain values of the stochastic volatility is superior to another stochastic volatility model such as GARCH (1,1). In addition, the performance is improved when we introduce Bollinger bands.

Number of Pages in PDF File: 21

Keywords: Term structure of interest rate, Extended Kalman Filter, Monte Carlo simulation, Root Mean Square Error, forecasting, stochastic volatility, Bollinger bands, Fong and Vasicek

JEL Classification: C15, C63, G13

working papers series