arXiv:2303.09406v2 Announce Type: replace-cross 
Abstract: Stock return prediction is fundamental to financial decision-making, yet traditional time series models fail to capture the complex interdependencies between companies in modern markets. We propose the Full-State Graph Convolutional LSTM (FS-GCLSTM), a novel temporal graph neural network that incorporates value-chain relationships to enhance stock return forecasting. Our approach features two key innovations: First, we represent inter-firm dependencies through value-chain networks, where nodes correspond to companies and edges capture supplier-customer relationships, enabling the model to leverage information beyond historical price data. Second, FS-GCLSTM applies graph convolutions to all LSTM components - current input features, previous hidden states, and cell states - ensuring that spatial information from the value-chain network influences every aspect of the temporal update mechanism. We evaluate FS-GCLSTM on Eurostoxx 600 and S&P 500 datasets using LSEG value-chain data. While not achieving the lowest traditional prediction errors, FS-GCLSTM consistently delivers superior portfolio performance, attaining the highest annualized returns, Sharpe ratios, and Sortino ratios across both markets. Performance gains are more pronounced in the denser Eurostoxx 600 network, and robustness tests confirm stability across different input sequence lengths, demonstrating the practical value of integrating value-chain data with temporal graph neural networks.

تم تقديم نهج جديد لتوقع عوائد الأسهم من خلال نموذج FS-GCLSTM (Full-State Graph Convolutional LSTM)، الذي يستخدم علاقات سلسلة القيمة لتعزيز دقة التوقعات. تتضمن هذه الطريقة تكامل الاعتماد المتبادل بين الشركات من خلال تمثيل الشركات كعقد وعلاقات المورد-العميل كحواف، مما يسمح بتحليل أكثر شمولاً يتجاوز بيانات الأسعار التاريخية.

Se ha introducido un nuevo enfoque para la predicción de rendimientos de acciones a través del Full-State Graph Convolutional LSTM (FS-GCLSTM), que utiliza relaciones de cadena de valor para mejorar la precisión de las previsiones. Este método integra las interdependencias entre empresas al representar a las compañías como nodos y las relaciones proveedor-cliente como aristas, permitiendo un análisis más completo más allá de los datos históricos de precios.

Une nouvelle approche de la prévision des rendements boursiers a été introduite avec le Full-State Graph Convolutional LSTM (FS-GCLSTM), qui utilise les relations de chaîne de valeur pour améliorer la précision des prévisions. Cette méthode intègre les interdépendances entre entreprises en représentant les sociétés comme des nœuds et les relations fournisseur-client comme des arêtes, permettant une analyse plus complète au-delà des données historiques de prix.

A novel approach to stock return prediction has been introduced through the Full-State Graph Convolutional LSTM (FS-GCLSTM), which utilizes value-chain relationships to enhance forecasting accuracy. This method integrates inter-firm dependencies by representing companies as nodes and supplier-customer relationships as edges, allowing for a more comprehensive analysis beyond historical price data.

Exploiting Supply Chain Interdependencies for Stock Return Prediction: A Full-State Graph Convolutional LSTM

Was this article worth reading? Share it

finlight.me

LucidQuery AI

Betron.io