A1236
Title: A simulation approach to robust risk management of derivative products
Authors: Bertrand Tavin - EMLYON Business School (France) [presenting]
Abstract: The purpose is to consider the problem of assessing and hedging the risk carried by a portfolio of non-standard derivative products managed with a family of parametric models. The problem is first formalized, and the framework is defined in which it can be solved by using a constrained simulation approach with respect to model parameters. The approach is suitable for an agent who may be agnostic with respect to a prior model and who may wish to account for expert views on the range of possible model parameters. Instead of breaking them into several sub-problems, the proposed methodology has the important advantage of answering the risk measurement and robust hedging questions in one step. Namely, the agent needs to run the simulation just once to get the desired answers. Numerical results obtained with recent market data when applying the method to a portfolio of variance swaps and forward-start options valued with models incorporating stochastic volatility and jumps are presented. In addition, the method can easily accommodate additional features such as cardinality constraints for the robust hedging strategy or transaction costs.
