Systematic comparison of deep generative models applied to multivariate financial time series

TL;DR

This paper systematically compares deep generative models and traditional parametric models for generating multivariate financial time series, demonstrating the potential advantages of DGMs in realistic financial applications.

Contribution

It provides a comprehensive evaluation of DGMs against parametric models on synthetic and real data for multivariate FTS generation.

Findings

DGMs outperform parametric models on synthetic datasets with increasing complexity.

DGMs show improved performance in capturing distribution moments of FTS.

Application of DGMs enhances implied volatility trading strategies.

Abstract

Financial time series (FTS) generation models are a core pillar to applications in finance. Risk management and portfolio optimization rely on realistic multivariate price generation models. Accordingly, there is a strong modelling literature dating back to Bachelier's Theory of Speculation in 1901. Generating FTS using deep generative models (DGMs) is still in its infancy. In this work, we systematically compare DGMs against state-of-the-art parametric alternatives for multivariate FTS generation. We initially compare both DGMs and parametric models over increasingly complex synthetic datasets. The models are evaluated through distance measures for varying distribution moments of both the full and rolling FTS. We then apply the best performing DGM models to empirical data, demonstrating the benefit of DGMs through a implied volatility trading task.