Flexible Tails for Normalising Flows, with Application to the Modelling of Financial Return Data
This work addresses the challenge of capturing extreme events in financial data, which is crucial for risk management and forecasting in finance, though it appears incremental as it builds on existing normalizing flow techniques.
The authors tackled the problem of modeling multivariate heavy-tailed distributions, particularly for financial returns, by proposing a transformation based on extreme value theory that can be integrated into normalizing flows, resulting in models capable of generating synthetic data with extreme shocks.
We propose a transformation capable of altering the tail properties of a distribution, motivated by extreme value theory, which can be used as a layer in a normalizing flow to approximate multivariate heavy tailed distributions. We apply this approach to model financial returns, capturing potentially extreme shocks that arise in such data. The trained models can be used directly to generate new synthetic sets of potentially extreme returns