Tail-GAN: Learning to Simulate Tail Risk Scenarios

Key Findings

Our idea involves a set of user-defined benchmark trading strategies to guide the training of the scenario generator to match the properties of their loss distributions. Focusing on a finite set of trading strategies leads to a dimension reduction, reducing the task to learning K one-dimensional distributions. Building on the concept of dimension reduction through the PnLs of benchmark trading strategies, we introduce a novel data-driven approach to simulate high-dimensional time series that represent financial market scenarios.

Abstract

The estimation of loss distributions for dynamic portfolios requires the simulation of scenarios representing realistic joint dynamics of their components. We propose a novel data-driven approach for simulating realistic, high-dimensional multi-asset scenarios, focusing on accurately representing tail risk for a class of static and dynamic trading strategies. We exploit the joint elicitability property of Value-at-Risk (VaR) and Expected Shortfall (ES) to design a Generative Adversarial Network (GAN) that learns to simulate price scenarios preserving these tail risk features. We demonstrate the performance of our algorithm on synthetic and market data sets through detailed numerical experiments.