CMStatistics 2017: Start Registration
View Submission - CFE
A1395
Title: Option pricing and hedging with one-step Kalman filtered factors in non-affine stochastic volatility models Authors:  Alex Badescu - University of Calgary (Canada)
Lyudmila Grigoryeva - University of Konstanz (Germany)
Juan-Pablo Ortega - University St. Gallen (Switzerland) [presenting]
Abstract: An innovative Kalman-based estimation technique is introduced for a non-affine auto-regressive stochastic factor model with non-predictable drift which allows to account for leverage effects. More specifically, we adopt a one-step unscented filtering approach which circumvents the use of the Kalman gain that produces a poor performance for this kind of models. New pricing and hedging strategies are proposed for contingent products that have this model for the underlying asset by introducing a volatility dependent exponential linear pricing kernel with stochastic risk aversion parameters. This technique proves to outperform standard GARCH and Heston-Nandi based strategies in terms of a variety of considered criteria in an empirical exercise using historical returns and options data.