Estimating risks of option books using neural-SDE market models

TL;DR

This paper introduces an arbitrage-free neural-SDE market model that generates realistic joint option dynamics and serves as an efficient risk simulation tool, outperforming traditional methods in VaR estimation accuracy and stability.

Contribution

It presents a novel neural-SDE market model capable of realistic joint option dynamics and improved risk assessment for portfolios.

Findings

More accurate VaR estimates with better coverage.

Enhanced computational efficiency over traditional methods.

Reduced procyclicality in risk evaluation.

Abstract

In this paper, we examine the capacity of an arbitrage-free neural-SDE market model to produce realistic scenarios for the joint dynamics of multiple European options on a single underlying. We subsequently demonstrate its use as a risk simulation engine for option portfolios. Through backtesting analysis, we show that our models are more computationally efficient and accurate for evaluating the Value-at-Risk (VaR) of option portfolios, with better coverage performance and less procyclicality than standard filtered historical simulation approaches.