Title: Forecasting the volatility of the yield curve
Authors: Mads Markvart Kjaer - Aarhus University (Denmark) [presenting]
Bezirgen Veliyev - Aarhus University (Denmark)
Bent Jesper Christensen - Aarhus University (Denmark)
Abstract: The ability of various forecasting methods to forecast future volatilities of the yield curve is examined. We document that standard affine term structure models are not able to capture all relevant information from the yield curve about future volatility and we find evidence for unspanned stochastic volatility even when we take the information from the yield curve into account. Overall, we conclude that it is difficult to significantly out-perform a random walk and that time-series approaches provide lower root-mean-squared forecast errors. We further find that changing the starting point of the out-of-sample has a large impact on the conclusions and, hence, we find switching behavior of the predictive-ability of term structure models regarding future volatility.