Towards Robust Representation of Limit Orders Books for Deep Learning Models

TL;DR

This paper examines the limitations of current limit order book data representations for deep learning, proposes modifications to improve robustness and accuracy, and demonstrates state-of-the-art results with simple models.

Contribution

It identifies issues with existing representations, introduces theoretically aligned modifications, and empirically shows improved robustness and performance.

Findings

Proposed representations enhance robustness against adversarial perturbations.

Modified data representations achieve state-of-the-art accuracy.

Simple neural networks benefit significantly from the proposed data modifications.

Abstract

The success of deep learning-based limit order book forecasting models is highly dependent on the quality and the robustness of the input data representation. A significant body of the quantitative finance literature focuses on utilising different deep learning architectures without taking into consideration the key assumptions these models make with respect to the input data representation. In this paper, we highlight the issues associated with the commonly-used representations of limit order book data from both a theoretical and practical perspectives. We also show the fragility of the representations under adversarial perturbations and propose two simple modifications to the existing representations that match the theoretical assumptions of deep learning models. Finally, we show experimentally how our proposed representations lead to state-of-the-art performance in both accuracy and robustness utilising very simple neural network architectures.