Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Fanshawe, Jack, Masih, Rumi, Cameron, Alexander
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2601.04602
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917189865439232
author Fanshawe, Jack
Masih, Rumi
Cameron, Alexander
author_facet Fanshawe, Jack
Masih, Rumi
Cameron, Alexander
contents This paper studies forward-looking stock-stock correlation forecasting for S\&P 500 constituents and evaluates whether learned correlation forecasts can improve graph-based clustering used in basket trading strategies. We cast 10-day ahead correlation prediction in Fisher-z space and train a Temporal-Heterogeneous Graph Neural Network (THGNN) to predict residual deviations from a rolling historical baseline. The architecture combines a Transformer-based temporal encoder, which captures non-stationary, complex, temporal dependencies, with an edge-aware graph attention network that propagates cross-asset information over the equity network. Inputs span daily returns, technicals, sector structure, previous correlations, and macro signals, enabling regime-aware forecasts and attention-based feature and neighbor importance to provide interpretability. Out-of-sample results from 2019-2024 show that the proposed model meaningfully reduces correlation forecasting error relative to rolling-window estimates. When integrated into a graph-based clustering framework, forward-looking correlations produce adaptable and economically meaningfully baskets, particularly during periods of market stress. These findings suggest that improvements in correlation forecasts translate into meaningful gains during portfolio construction tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2601_04602
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Forecasting Equity Correlations with Hybrid Transformer Graph Neural Network
Fanshawe, Jack
Masih, Rumi
Cameron, Alexander
Computational Finance
Trading and Market Microstructure
This paper studies forward-looking stock-stock correlation forecasting for S\&P 500 constituents and evaluates whether learned correlation forecasts can improve graph-based clustering used in basket trading strategies. We cast 10-day ahead correlation prediction in Fisher-z space and train a Temporal-Heterogeneous Graph Neural Network (THGNN) to predict residual deviations from a rolling historical baseline. The architecture combines a Transformer-based temporal encoder, which captures non-stationary, complex, temporal dependencies, with an edge-aware graph attention network that propagates cross-asset information over the equity network. Inputs span daily returns, technicals, sector structure, previous correlations, and macro signals, enabling regime-aware forecasts and attention-based feature and neighbor importance to provide interpretability. Out-of-sample results from 2019-2024 show that the proposed model meaningfully reduces correlation forecasting error relative to rolling-window estimates. When integrated into a graph-based clustering framework, forward-looking correlations produce adaptable and economically meaningfully baskets, particularly during periods of market stress. These findings suggest that improvements in correlation forecasts translate into meaningful gains during portfolio construction tasks.
title Forecasting Equity Correlations with Hybrid Transformer Graph Neural Network
topic Computational Finance
Trading and Market Microstructure
url https://arxiv.org/abs/2601.04602