Modeling Yield-Factor Volatility

The term structure of interest rates is often summarized using a handful of yield factors that capture shifts in the yield curve. Despite their wide application in financial economics, very little is known on the time-series properties of the yield-factor volatilities. We examine three common yield-factors: the level of short-term interest rates, the slope and curvature in the yield curve. We model the volatility dynamics in these yield factors using both GARCH and level effects and find that both are needed to adequately model yield-factor volatility. The level effect is routinely used when modeling volatility in short-term interest rates and we find that the level of the short-rate is useful in modeling the volatility of the slope and curvature too. We also examine the effect of volatility on the dynamics of the yield-factors and find that the GARCH-based volatility of the short-rate is negatively related to future interest rates and positively related to the slope of the yield curve. This volatility-in-mean effect is much weaker when a level effect is introduced. We also examine regime switching models that recognize different economic regimes and find that this dramatically improves the model's fit. Interestingly, the level effect is strengthened and the GARCH effects is weakened somewhat. The Bayesian information criteria suggests that the correct model is a regime-switching model with level effec