7–11 Dec 2026
The University of Sydney
Australia/Sydney timezone
AIP Congress 2026

Gaussian Boson Sampling on quantum microprocessor for Equity Network Analysis

Not scheduled
20m
Belinda Hutchinson Building (The University of Sydney )

Belinda Hutchinson Building

The University of Sydney

Abercrombie St & Codrington St NSW 2008
Contributed Oral ANZOS | Quantum Computing and Quantum Information (ANZCOP QCQI)

Description

In equity markets, many risks do not arise from individual stocks in isolation, but from shared price movements, high return correlations, similar sector exposures, and comparable sensitivities to common risk factors. Stock clustering can help reveal latent groups of equities that may react similarly to market shocks, enabling investors to detect hidden risk concentrations, assess whether a portfolio is overly exposed to a single market structure, and support more disciplined allocation decisions. Conventional methods such as k-means and DBSCAN remain useful benchmarks; however, their performance may depend on assumptions that are not always consistent with equity-market data. In particular, k-means can be less reliable under non-convex structures or ambiguous boundaries, while density-based methods may be affected by neighbourhood-parameter selection and uneven density distributions.

In this work, a Gaussian Boson Sampling (GBS)-based clustering algorithm is proposed to uncover latent correlation structures in the stock market. The study constructs a weighted graph using stock return correlations, volatility features, and sector attributes, where nodes represent individual equities and edge weights quantify market similarity between assets. After the graph is embedded into a GBS photonic chip, the resulting photon sampling distribution naturally tends to select dense subgraphs corresponding to highly connected asset groups. Through repeated sampling, density filtering, and classical post-processing, candidate asset groups are extracted, leading to economically interpretable stock clusters with similar co-movement patterns and risk exposures.

The results show that the proposed method achieves higher silhouette and weighted density scores than DBSCAN, while its intra–inter cluster cohesion is higher than that of k-means. This approach may be applied to risk dispersion, dynamic asset allocation, sector rotation analysis, and systemic risk monitoring, offering a feasible technical pathway for deploying integrated photonic quantum technologies in quantitative finance.

I am the presenting author Yes

Author

Tian CHEN (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong; 3 School of Electronic Information, Central South University, Changsha, China)

Co-authors

Aiqun Liu (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;) Hong CAI (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;) Jinjing SHI (School of Electronic Information, Central South University, Changsha, China) Lap-Pui CHAU (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;) Lipket CHIN (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;) Wei Wang (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;) Zan Tang (1 Research Institute for Quantum Technology (RIQT), The Hong Kong Polytechnic University, Hong Kong; 2 Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong;)

Presentation materials

There are no materials yet.