CMStatistics 2017: Start Registration
View Submission - CFE
Title: Expected term structures Authors:  Ilaria Piatti - University of Oxford (United Kingdom) [presenting]
Andrea Buraschi - Imperial College Business School (United Kingdom)
Paul Whelan - Copenhagen Business School (Denmark)
Abstract: The properties of bond risk premia in the cross-section of subjective expectations are studied. We exploit an extensive dataset of yield curve forecasts from financial institutions and document a number of novel findings. First, the relation between subjective expectations and future realizations is positive, and this result holds for the entire cross-section of beliefs. Second, when predicting short term interest rates, primary dealers display superior forecasting ability when compared to non-primary dealers. When predicting long term rates, however, primary dealers have no information advantage. This suggests that a key source of variation in long-term bonds are risk premia and not short-term rate variation. Fourth, we show that consensus beliefs are not a sufficient statistics to describe the cross-section of beliefs. We build an aggregate measure of bond risk premia based on the beliefs of the most accurate agents, consistent with equilibrium models with disagreement and with Friedmans market selection hypothesis. Finally, we use these ex-ante subjective beliefs to evaluate several reduced-form and structural models. We find support for heterogeneous beliefs models and also uncover a number of statistically significant relationships in favour of alternative rational expectations models once the effect of heterogeneous beliefs is taken into account.