Project
Project description
Fundamental challenges of modern stochastic finance are the construction of calibrated asset price models as well as the estimation of the inherent model risk. In the last decade, martingale optimal transport (MOT) has developed into a prime tool to deal with these challenges. It provides a systematic way to assess and quantify model risk given market data on European options, and has further important applications, e.g. to Skorokhod embedding, martingale inequalities, and model calibration. While being a powerful tool, a significant shortcoming of MOT is that it is currently limited to problems involving data on options with European exercise structure while it excludes American options. Since most of the traded options on individual stocks are of American type, this is a severe gap.
In this project we aim to overcome this shortcoming and make techniques of martingale transport applicable also to financial derivatives with American exercise structure based on adapted Wasserstein distance. In fact, as long as one is interested only in European derivatives, the classical weak topology / Wasserstein distance is sufficient to compare different processes. However, as soon as American derivatives are considered, the evolution of information plays a crucial role and hence it is necessary to use the adapted Wasserstein distance which adequately takes the role of information into account. Based on this insight, our main objective is to develop martingale transport into a comprehensive method to systematically deal with the problems of model calibration and model risk given market data on European as well as American option prices.
Important dates
The project started on October 1st, 2025.