Paper: arXiv 2303.09406

Authors: Chang Liu

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.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper presents a novel architecture (FS-GCLSTM) combining graph convolutions across all LSTM components with corporate supply chain data. It demonstrates strong empirical rigor through portfolio-based backtests on major indices (S&P 500 and Eurostoxx 600), showing clear practical utility despite standard point-forecast error metrics.

Research Flowchart

  flowchart TD
    A[Research Goal: Predict Stock Returns using Value-Chain Interdependencies] --> B[Data Inputs: S&P 500 & Eurostoxx 600 Price Data + LSEG Value-Chain Networks]
    B --> C[Graph Construction: Nodes = Companies, Edges = Supplier-Customer Links]
    C --> D[FS-GCLSTM Model: Apply Graph Convolutions to Input, Hidden, and Cell States]
    D --> E[Evaluation: Time Series Forecasting & Portfolio Strategy Backtesting]
    E --> F[Outcomes: Superior Annualized Returns, Sharpe & Sortino Ratios]