Deep Limit Order Book Forecasting

TL;DR

This paper investigates the use of deep learning to predict high-frequency Limit Order Book mid-price changes on NASDAQ, introduces an open-source processing tool, and proposes a new evaluation framework emphasizing practical forecasting accuracy.

Contribution

It presents LOBFrame, an open-source tool for large-scale Limit Order Book data processing, and introduces an operational framework for assessing forecast quality based on transaction prediction accuracy.

Findings

Stock microstructure affects deep learning forecast efficacy.

High forecast accuracy does not always lead to actionable trading signals.

Traditional metrics may not adequately evaluate Limit Order Book predictions.

Abstract

We exploit cutting-edge deep learning methodologies to explore the predictability of high-frequency Limit Order Book mid-price changes for a heterogeneous set of stocks traded on the NASDAQ exchange. In so doing, we release `LOBFrame', an open-source code base to efficiently process large-scale Limit Order Book data and quantitatively assess state-of-the-art deep learning models' forecasting capabilities. Our results are twofold. We demonstrate that the stocks' microstructural characteristics influence the efficacy of deep learning methods and that their high forecasting power does not necessarily correspond to actionable trading signals. We argue that traditional machine learning metrics fail to adequately assess the quality of forecasts in the Limit Order Book context. As an alternative, we propose an innovative operational framework that evaluates predictions' practicality by focusing on the probability of accurately forecasting complete transactions. This work offers academics and practitioners an avenue to make informed and robust decisions on the application of deep learning techniques, their scope and limitations, effectively exploiting emergent statistical properties of the Limit Order Book.