Learning Temporal Patterns in Financial Time Series: A Comparative Study of Quantum LSTM and Quantum Reservoir Computing

TL;DR

This paper compares quantum and classical models for financial time series forecasting, demonstrating that quantum architectures can match or slightly outperform classical methods with proper encoding and lag selection.

Contribution

It introduces quantum hybrid architectures for financial forecasting, evaluating their performance against classical models using amplitude encoding and lag structures.

Findings

Quantum models match classical baselines in univariate forecasting.

Quantum models can outperform classical models in multivariate, correlated input scenarios.

Proper lag selection and amplitude encoding are crucial for quantum model performance.

Abstract

This study explores quantum and classical hybrid architectures for financial time-series fore casting, focusing on Quantum Long Short-Term Memory (QLSTM) networks and Quantum Reservoir Computing (QRC), using univariate and multivariate lag structures on real financial data. We assess how lag embeddings affect predictive accuracy and robustness. Data are en coded into quantum states via amplitude encoding, enabling efficient representation of normalized lagged observations under realistic qubit constraints. The recurrent dynamics of QLSTM and the reservoir of QRC are implemented as parameterized quantum circuits, while classical optimizers train the readout and, where applicable, variational circuit parameters. We benchmark quantum models against classical LSTM and reservoir computing using common error like metrics. Our results show that, with suitable lag selection and amplitude encoding, quantum-enhanced archi tectures match classical baselines in univariate settings and can modestly outperform them in multivariate regimes with correlated inputs, where expressive encodings are most beneficial.