Risk, hedging and control via signatures and machine learning
Contact person: Giulia Di Nunno
Keywords: Risk measures, signatures, deep forward backward systems, sustainability
Research group: Risk and Stochastics
Department of Mathematics
The identification of risk, its quantification, and its hedging and control are naturally of paramount importance in both engineering, climate, and financial linked activities. A monetary evaluation of risk is necessary for decision making and evaluation of investments and portfolios. The volumes treated are usually of such high dimensionality that heavy computational methods are required with the downside in the tradeoff between effectiveness and precision in risk evaluations. Then the question is how to effectively reduce on the dimensionality and still keep the structure of the data set. Also the definition and the computation of risk measures both cross-sector and environmentally sensitive are still challenging aspects in the strive of capturing and comparing different data streams in the risk evaluation. This would also impact on the overall risk quantification aimed at ESG reporting.
We are interested in research proposals that are directed towards the theoretical and the computational aspects of risk quantification, hedging and minimisation.
The research proposals may span several methodological approaches within stochastic control, optimisation, and computational methods, as well as a number of application domains.
Methodological research topics:
- Static and dynamic risk measurement and stochastic control methods
- Multi-dimensional and/or systemic, and/or set-valued risk measures, static and dynamic frameworks.
- Signature methods for the determination of characteristics of prices and/or climate factors, e.g. temperatures, rainfalls.
- Deep forward backward systems for the computation of risk measures, risk minimisation, risk indifferent pricing
Application domains:
- Energy systems, banking systems, insurance
- Sustainability
External partner:
- Gjensidige
