Data-Efficient Realized Volatility Forecasting with Vision Transformers

TL;DR

This paper explores using Vision Transformers to predict asset volatility from options data, demonstrating their ability to learn complex seasonal and nonlinear patterns, which could enhance financial forecasting models.

Contribution

It introduces a novel application of Vision Transformers to options data for volatility forecasting, a relatively unexplored area in financial machine learning.

Findings

ViT can learn seasonal patterns from IV surface

ViT captures nonlinear features in options data

Preliminary results show promise for model development

Abstract

Recent work in financial machine learning has shown the virtue of complexity: the phenomenon by which deep learning methods capable of learning highly nonlinear relationships outperform simpler approaches in financial forecasting. While transformer architectures like Informer have shown promise for financial time series forecasting, the application of transformer models for options data remains largely unexplored. We conduct preliminary studies towards the development of a transformer model for options data by training the Vision Transformer (ViT) architecture, typically used in modern image recognition and classification systems, to predict the realized volatility of an asset over the next 30 days from its implied volatility surface (augmented with date information) for a single day. We show that the ViT can learn seasonal patterns and nonlinear features from the IV surface, suggesting a promising direction for model development.