CMStatistics 2021: Start Registration
View Submission - CFE
Title: A factor-based calibration model: An application to the relation between sustainability and option-implied distributions Authors:  Giovanni Pianon - Ca' Foscari University of Venice (Italy) [presenting]
Abstract: Option-implied data represent an essential source of information for modelling the distribution of stock returns. However, as the scientific literature has shown, agents' risk preferences and irrationality can lead to a severe misalignment from the objective distribution. We propose a factor-based calibration model to correct for such a bias. By linking the calibration function parameters to a set of common and idiosyncratic factors, the model captures the time-varying nature of the forces driving the distortion of option-implied distributions. Moreover, thanks to copulas, it leverages the cross-sectional dependence of stock returns in implementing univariate calibration. A Bayesian inference approach and an efficient Monte Carlo posterior approximation allow us to deal with the model's high-dimensionality and intractability. We employed the model to perform density forecasting and to shed light on the underlying factors determining the bias of risk-neutral densities, investigating, in particular, the role of ESG ratings and other sustainability indicators