|| |Dear colleagues, | ||||||| LTI@UniTO (www.carloalberto.org/lti <www.carloalberto.org/lti>) and Collegio Carlo Alberto are pleased to invite you to the following webinar in Quantitative Finance, which will take place on Monday, December 14th at 12.45 via Zoom. Please register at the following link to get the Zoom meeting link and reminders: https://www.eventbrite.it/e/biglietti-bond-risk-premia-the-information-in-re... <https://www.eventbrite.it/e/biglietti-bond-risk-premia-the-information-in-really-long-maturity-forward-rates-131857554495> |* * *Title: “Bond risk premia: the information in really long-maturity forward rates”* * * *Speaker: Andrea Berardi (Università Ca' Foscari Venezia) * **Abstract:*Disentangling expectations of future interest rates from risk premia in the determinants of the term structure has long been a challenge, not only to financial economists but to policymakers. Although the great majority of empirical research on the term structure focuses on relatively short maturities (10 years or less), we show that distinguishing between expectations and risk premia is easier at long maturities and, in this paper, we include much longer term rates (20 years and more). In this region, differences in forward rates are little affected by expectations but, nonetheless, reflect significantly differences in risk premia. Key to extracting information about risk premia, is taking proper account of the influence of time-varying volatility of long-term rates and bond convexity. The impact of convexity on the term structure increases strongly with maturity and leads frequently to a negative slope (a “downward tilt”) in long-term forward rates. We employ a four-factor affine model with stochastic volatility that fits the dynamics of the yield curve and, in particular, the downward tilt in forward rates, well. Risk premia in our model are, on average, monotonically upward sloping and we find that volatility accounts for a significant fraction of their variation over time. We also find that including stochastic volatility results in less volatile estimates of term premia than in models with constant volatility. Our model is consistent with previously reported deviations from the pure Expectations Hypothesis, a result which contrasts with previous empirical evidence on the failure of stochastic volatility term structure models in this matter.* * * * * Best regards, Luca Regis -- Luca Regis Associate Professor Department of Economics and Statistics (ESOMAS) University of Torino sites.google.com/view/lucaregis Office: +39 011 670 6065 www.carloalberto.org/lti ||