Data driven approach to robust pricing, hedging and risk management and its dynamics in time

Thesis

Model-independent mathematical finance is a relatively young research field, which aims to quantify Knightian uncertainty on the choice of pricing models. This thesis extends the existing framework to a setting, which is dynamic in time and takes new sources of information (e.g. new statistical estimates or changed market prices) into account as soon as they emerge. In this context, new connections to the fields of of optimal transport on the one hand and statistics on the other are established, with the ultimate goal to develop a universal toolbox for the implementation of robust and time-consistent trading strategies and risk assessment.

Authors: Wiesel, JCW

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Robustness; (Robust) Statistics; Superhedging price; Wasserstein distance; (Martingale) Optimal Transport; Derivatives hedging; Neural networks; Risk measures; Mathematical Finance; Pricing-hedging duality; Fundamental Theorem of Asset pricing; Derivatives pricing; Statistical estimation
• Subjects: Machine learning; Mathematics; Statistics; Finance; Hedging (Finance); Derivative securities