Original title: Quantum Weighted Moving Average for Predicting Limit Order Book Trends
Matthias Kamm, Dinh-Long Vu, Patrick Rebentrost · September 2026
Predicting future stock price trends using market order records, called limit order book data, is a difficult financial task. Researchers wanted to see if quantum computers could help forecast this information. To find out, they designed a hybrid approach called the quantum weighted moving average model. It works in two main steps. First, normal computers clean up and scale the financial numbers across time and features. Second, a special quantum layer processes that data, focusing on the most useful parts of the timeline. The team tested their method against standard computers on two stock market datasets. They did not find a quantum advantage, but their combined classical-quantum design performed nearly as well as top traditional models.
Imagine packing for a trip where you have far too many items to fit in your suitcase. You need to decide what's essential based on the weather and your plans, but overpacking wastes space and underpacking leaves you unprepared. In this paper, the researchers face a similar challenge with a flood of stock price data: their method learns to highlight only the most predictive bits to forecast future trends.
This work matters for financial analysts and quantum researchers exploring practical applications. It shows that quantum-inspired models can be competitive in real-world financial forecasting, even without quantum speedup, suggesting a path for future hybrid approaches.
The abstract explicitly states that no evidence of quantum advantage was found. The model's performance is close to classical methods but not better. Evaluations are limited to specific datasets, and the methods have acknowledged limitations without further details.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Quantum Weighted Moving Average for Predicting Limit Order Book Trends — Matthias Kamm, Dinh-Long Vu, Patrick Rebentrost
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.