Stock Recommendations for Individual Investors: A Temporal Graph Network Approach with Mean-Variance Efficient Sampling

Paper: arXiv 2404.07223

Abstract

Recommender systems can be helpful for individuals to make well-informed decisions in complex financial markets. While many studies have focused on predicting stock prices, even advanced models fall short of accurately forecasting them. Additionally, previous studies indicate that individual investors often disregard established investment theories, favoring their personal preferences instead. This presents a challenge for stock recommendation systems, which must not only provide strong investment performance but also respect these individual preferences. To create effective stock recommender systems, three critical elements must be incorporated: 1) individual preferences, 2) portfolio diversification, and 3) the temporal dynamics of the first two. In response, we propose a new model, Portfolio Temporal Graph Network Recommender PfoTGNRec, which can handle time-varying collaborative signals and incorporates diversification-enhancing sampling. On real-world individual trading data, our approach demonstrates superior performance compared to state-of-the-art baselines, including cutting-edge dynamic embedding models and existing stock recommendation models. Indeed, we show that PfoTGNRec is an effective solution that can balance customer preferences with the need to suggest portfolios with high Return-on-Investment. The source code and data are available at https://github.com/youngandbin/PfoTGNRec.

Complexity vs Empirical Score

• Math Complexity: 6.5/10
• Empirical Rigor: 6.0/10 (capped from 7.0: transaction costs not modelled; see penalty flags)
• Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper uses advanced graph neural networks (temporal graph networks) and optimization theory (mean-variance efficient sampling), indicating substantial math complexity. It is evaluated on real-world individual trading data with source code and data available, demonstrating strong empirical backing and implementation readiness.

Research Flowchart

  flowchart TD
  A["Research Goal: <br>Stock Recommendations for Individual Investors<br>1. Individual Preferences 2. Portfolio Diversification 3. Temporal Dynamics"] --> B
  
  subgraph B ["Methodology: PfoTGNRec Model"]
      B1["Temporal Graph Network"] --> B2["Dynamic Embedding"] --> B3["Diversification-Enhanced Sampling"]
  end
  
  C["Input Data: <br>Real-world Individual Trading Data"] --> B
  
  B --> D{"Computational Process: <br>Time-Varying Collaborative Signals<br>+<br>Mean-Variance Efficient Sampling"}
  
  D --> E
  
  subgraph E ["Key Findings/Outcomes"]
      E1["Superior Performance<br>vs. State-of-the-Art Baselines"]
      E2["Balanced Trade-off:<br>Preferences + High ROI"]
      E3["Effective Solution<br>for Dynamic Stock Recommendations"]
  end

Cited References (93)

1. Changes in US Family Finances from 2019 to 2022 (2023) — link
2. Behavior Pattern of Individual Investors in Stock Market (2014) — link
3. Tracking S&P 500 index funds (2001)
4. DALBAR Products and Services: QAIB (2023) — link
5. Next-Generation Personalized Investment Recommendations (2021) — link
6. Deep learning in finance and banking: A literature review and classification (2020)
7. Financial applications of machine learning: A literature review (2023)
8. The Wisdom of the Robinhood Crowd (2022) — link

1 of 93 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.