Paper: arXiv 2604.19476

Authors: Yikuan Huang, Zheqi Fan, Kaiqi Hu, Yifan Ye

Abstract

Text-based financial networks are increasingly used to study cross-stock return predictability. A common approach constructs links from similarities in firms’ disclosure embeddings, but such networks often contain spurious edges because textual proximity does not necessarily imply economic connection. We propose a two-stage framework that first builds a sparse candidate graph from 10-K embeddings and then uses a large language model to classify and filter candidate edges according to their economic relations. The refined graph is used to aggregate pair-level mean-reversion signals into stock-level trading signals with relation-aware and distance-based weights. In a backtest on S&P 500 constituents from 2011 to 2019, LLM-based edge filtering improves the long-short Sharpe ratio from 0.742 to 0.820 and reduces maximum drawdown from $-$10.47% to $-$7.85%. These results suggest that LLM-based reasoning can improve the economic fidelity of text-derived financial networks and strengthen cross-stock predictability.

Complexity vs Empirical Score

  • Math Complexity: 4.5/10
  • Empirical Rigor: 6.0/10 (capped from 8.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper relies on standard empirical finance backtesting techniques (Sharpe ratio, drawdown, factor regression) rather than complex stochastic calculus or novel mathematical proofs. Its primary innovation is implementing a practical, data-heavy machine learning workflow for signal processing, making it a quintessentially ‘Street Trader’ approach.

Research Flowchart

  flowchart TD
    A[<b>Research Goal:</b> Improve cross-stock predictability via filtered semantic networks] --> B[<b>Data Input:</b> 10-K Filings]
    B --> C[<b>Stage 1:</b> Construct sparse candidate graph via embedding similarity]
    C --> D[<b>Stage 2:</b> LLM-based reasoning to filter for economic relations]
    D --> E[<b>Process:</b> Weight signals & aggregate pair-level mean-reversion]
    E --> F[<b>Outcomes:</b> Sharpe ratio 0.742 → 0.820; Max Drawdown -10.47% → -7.85%]

Cited References (182)

  1. Closing the GARCH gap: Continuous time GARCH modeling (1996)
  2. Cross-stock momentum and factor momentum (2023)
  3. How Much of Cross-Stock Momentum Reflects Underreaction? (2025)
  4. Can ChatGPT forecast stock price movements? Return predictability and large language models (2023)
  5. Expected returns and large language models (2023)
  6. BloombergGPT: A large language model for finance (2023) — link
  7. From Text to Alpha: Can LLM s Track Evolving Signals in Corporate Disclosures? (2025) — link
  8. Finben: A holistic financial benchmark for large language models (2024)
and 174 more
  1. Structured Beliefs and Fund Performance: An LLM -Based Approach (2024)
  2. Pixiu: A large language model, instruction data and evaluation benchmark for finance (2023) — link
  3. A Financial Brain Scan of the LLM — → 2508.21285
  4. Dual Peer Effects and Cross-Stock Predictability (2026)
  5. On cross-stock predictability of peer return gaps in China (2026)
  6. Economic links and predictable returns (2008)
  7. Do industries explain momentum? (1999)
  8. Geographic lead-lag effects (2020)
  9. Shared analyst coverage: Unifying momentum spillover effects (2020)
  10. Empirical investigation of an equity pairs trading strategy (2019)
  11. Principal portfolios (2023)
  12. Text-based industry momentum (2018)
  13. Market segmentation and cross-predictability of returns (2010)
  14. Jumps and stochastic volatility: Exchange rate processes implicit in deutsche mark options (1996)
  15. Pricing and calibration in the 4-factor path-dependent volatility model — → 2406.02319
  16. The risk premia embedded in index options (2015)
  17. Model specification and risk premia: Evidence from futures options (2007)
  18. Forecasting with option-implied information (2013)
  19. Measuring tail risk (2024)
  20. Information in derivatives markets: Forecasting prices with prices (2025)
  21. Merton (1976) implied jump (2025)
  22. Modeling the Implied Volatility Smirk in China : Do Non-Affine Two-Factor Stochastic Volatility Models Work? (2025)
  23. Empirical option pricing models (2022)
  24. Levy jump risk: Evidence from options and returns (2014)
  25. Post-‘87 crash fears in the S&P 500 futures option market (2000)
  26. Transform analysis and asset pricing for affine jump-diffusions (2000)
  27. Fast filtering with large option panels: Implications for asset pricing (2024)
  28. A tale of two option markets: Pricing kernels and volatility risk (2016)
  29. Specification analysis of option pricing models based on time-changed L 'e vy processes (2004)
  30. Smiles & smirks: Volatility and leverage by jumps (2022)
  31. Assessing models of individual equity option prices (2021)
  32. A closed-form solution for options with stochastic volatility with applications to bond and currency options (1993)
  33. Option valuation using the fast Fourier transform (1999)
  34. A novel pricing method for European options based on Fourier -cosine series expansions (2009)
  35. Efficient options pricing using the fast Fourier transform (2011)
  36. The dynamics of stochastic volatility: Evidence from underlying and options markets (2003)
  37. Alternative models for stock price dynamics (2003)
  38. Volatility dynamics for the S&P 500 : Further evidence from non-affine, multi-factor jump diffusions (2012)
  39. Empirical analysis of affine versus nonaffine variance specifications in jump-diffusion models for equity indices (2015)
  40. Jump and volatility dynamics for the S&P 500: Evidence for infinite-activity jumps with non-affine volatility dynamics from stock and option markets (2017)
  41. Model Complexity and Out-of-Sample Performance: Evidence from S&P 500 Index Returns (2018)
  42. Option profit and loss attribution and pricing: A new framework (2020)
  43. Interpretability in deep learning for finance: A case study for the Heston model (2026)
  44. Chronologically consistent large language models — → 2502.21206
  45. Machine learning and the stock market (2023)
  46. Is technical analysis in the foreign exchange market profitable? A genetic programming approach (1997)
  47. Technical analysis as a sentiment barometer and the cross-section of stock returns (2023)
  48. AI Agent Misinformation when Assisting Financial Decision-Making: Early Evidence From Stock Recommendations (2025)
  49. Deep surrogate for non-affine stochastic volatility option valuation models (2026)
  50. Trading Volume Alpha (2024)
  51. On Options-Driven Realized Volatility Forecasting: Information Gains via Rough Volatility Model (2025)
  52. Large language models for financial and investment management: Applications and benchmarks (2024)
  53. Large Language Models for Financial and Investment Management: Models, Opportunities, and Challenges. (2024)
  54. Expected returns and large language models (2022)
  55. A simple, positive semi-definite, heteroskedasticity and autocorrelationconsistent covariance matrix (1987)
  56. Technical indicators and the cross-section of corporate bond returns in a machine learning era (2025)
  57. Presidential address: The scientific outlook in financial economics (2017)
  58. ChatGPT and DeepSeek : Can they predict the stock market and macroeconomy? (2025) — link
  59. Generative AI for Finance: A New Framework (2025)
  60. Option valuation under no-arbitrage constraints with neural networks (2021)
  61. A nonparametric approach to pricing and hedging derivative securities via learning networks (1994)
  62. Deep Learning (2016)
  63. Forecasting value-at-risk using deep neural network quantile regression (2024)
  64. Inferring volatility dynamics and risk premia from the S&P 500 and VIX markets (2019)
  65. Explainable machine learning models of consumer credit risk (2023)
  66. Increasing the transparency of pricing dynamics in the US commercial real estate market with interpretable machine learning algorithms (2023)
  67. Explainable AI in credit risk management (2021) — link
  68. On deep calibration of (rough) stochastic volatility models (2025)
  69. Deep Calibration with Artificial Neural Network: A Performance Comparison on Option-Pricing Models (2023)
  70. A new closed-form discrete-time option pricing model with stochastic volatility (2023)
  71. Pricing and calibration in the 4-factor path-dependent volatility model — → 2406.02319
  72. Forecasting gold price with the XGBoost algorithm and SHAP interaction values (2021)
  73. Bankruptcy prediction using machine learning and Shapley additive explanations (2023)
  74. XAI –explainable artificial intelligence (2019)
  75. Approximation rates for deep calibration of (rough) stochastic volatility models (2024)
  76. Machine Learning and Data Sciences for Financial Markets: A Guide to Contemporary Practices (2023)
  77. Statistical machine learning for quantitative finance (2023)
  78. Explainable Artificial Intelligence (XAI) : Concepts, taxonomies, opportunities and challenges toward responsible AI (2020)
  79. Deep calibration of the quadratic rough Heston model (2022)
  80. Machine learning methods in finance: Recent applications and prospects (2023)
  81. ( R e-) Imag (in) ing price trends (2023)
  82. Narrative asset pricing: Interpretable systematic risk factors from news text (2023)
  83. Large Language Models and Futures Price Factors in China (2026)
  84. A taxonomy of anomalies and their trading costs (2016)
  85. Giving content to investor sentiment: The role of media in the stock market (2007)
  86. Foundations of technical analysis: Computational algorithms, statistical inference, and empirical implementation (2000)
  87. When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks (2011)
  88. Can ChatGPT forecast stock price movements? return predictability and large language models — → 2304.07619
  89. Characteristics are covariances: A unified model of risk and return (2019)
  90. Returns to buying winners and selling losers: Implications for stock market efficiency (1993)
  91. On persistence in mutual fund performance (1997)
  92. Empirical asset pricing via machine learning (2020)
  93. A five-factor asset pricing model (2015)
  94. Short-Term Moving Average Distance and the Cross-Section of Stock Returns (2025)
  95. False (and missed) discoveries in financial economics (2020)
  96. Lucky factors (2021)
  97. Advances in financial machine learning (2018)
  98. Implementing AI Foundation Models in Asset Management: A Practical Guide. (2025)
  99. Systematic investment strategies (2017)
  100. Chain-of-thought prompting elicits reasoning in large language models (2022)
  101. Choices matter when training machine learning models for return prediction (2024)
  102. Beyond Fama-French factors: Alpha from short-term signals (2023)
  103. Reflexion: Language agents with verbal reinforcement learning (2023)
  104. ReAct : Synergizing Reasoning and Acting in Language Models (2023)
  105. Industry return predictability: A machine learning approach (2019)
  106. Alpha go everywhere: Machine learning and international stock returns (2025)
  107. Common risk factors in the returns on stocks and bonds (1993)
  108. Simple technical trading rules and the stochastic properties of stock returns (1992)
  109. Persuading investors: A video-based study (forthcoming)
  110. GPT ’s idea of stock factors (2024)
  111. Autonomous Market Intelligence: Agentic AI Nowcasting Predicts Stock Returns (2026)
  112. Capital asset prices: A theory of market equilibrium under conditions of risk (1964)
  113. Neural network-based automatic factor construction (2020)
  114. Selecting Computational Models for Asset Management: Financial Econometrics versus Machine Learning-Is There a Conflict? (2020)
  115. Agentic AI and the Future of Institutional Asset Management. (2025)
  116. … and the cross-section of expected returns (2016)
  117. A trend factor: Any economic gains from using information over investment horizons? (2016)
  118. A new anomaly: The cross-sectional profitability of technical analysis (2013)
  119. The virtue of complexity in return prediction (2024)
  120. Financial machine learning (2023)
  121. Machine learning in energy economics and finance: A review (2019)
  122. Machine learning in accounting and finance research: A literature review (2024)
  123. RESHAPE: Explaining accounting anomalies in financial statement audits by enhancing SHapley additive explanations (2022)
  124. Machine learning for quantitative finance: Fast derivative pricing, hedging and fitting (2018)
  125. Deep Learning for Exotic Option Valuation (2022)
  126. A neural network approach to understanding implied volatility movements (2020)
  127. Wavelet-based option pricing: An empirical study (2019)
  128. Deep learning of transition probability densities for stochastic asset models with applications in option pricing (2025)
  129. Stripping the Discount Curve–a Robust Machine Learning Approach (2026)
  130. Deep learning in asset pricing (2024)
  131. Deep learning statistical arbitrage (2026)
  132. Deep neural networks, gradient-boosted trees, random forests: Statistical arbitrage on the S&P 500 (2017)
  133. Industry return prediction via interpretable deep learning (2025)
  134. Taming the curse of dimensionality: quantitative economics with deep learning (2024)
  135. Short-term market risks implied by weekly options (2017)
  136. Parametric inference and dynamic state recovery from option panels (2015)
  137. How and when are high-frequency stock returns predictable? (2026)
  138. Machine learning vs. economic restrictions: Evidence from stock return predictability (2023)
  139. Intraday market return predictability culled from the factor zoo (2025)
  140. A Value for $n$-person Games (1953) — link
  141. A unified approach to interpreting model predictions (2017)
  142. Mathematical methods for financial markets (2009)
  143. On the nature of (jump) skewness risk premia (2024)
  144. The jump-risk premia implicit in options: Evidence from an integrated time-series study (2002)
  145. Volatility Models in Practice: Rough, Path-Dependent, or Markovian ? (2025)
  146. Joint SPX & VIX calibration with Gaussian polynomial volatility models: Deep pricing with quantization hints (2025)
  147. The price of the smile and variance risk premia (2021)
  148. Riding on the smiles (2011)
  149. Persistent and transient variance components in option pricing models with variance-dependent Kernel (2024)
  150. A First Option Calibration of the GARCH Diffusion Model by a PDE Method (2018) — link
  151. Switching to nonaffine stochastic volatility: A closed-form expansion for the inverse gamma model (2016)
  152. Time-series and cross-sectional stock return forecasting: New machine learning methods (2020)
  153. Artificial Intelligence–Powered (Finance) Scholarship (2026)
  154. Pairs trading: Performance of a relative-value arbitrage rule (2006)
  155. Network momentum across asset classes — → 2308.11294
  156. Beyond prompting: An autonomous framework for systematic factor investing via agentic AI — → 2603.14288
  157. Production complementarity and information transmission across industries (2024)
  158. Technological links and predictable returns (2019)
  159. The effect of innovation similarity on asset prices: Evidence from patents’ big data (2023)
  160. Competition links and stock returns (2022)
  161. The Memorization Problem: Can We Trust LLM s’ Economic Forecasts? — → 2504.14765
  162. One News, Two Markets: LLM -Derived Sentiment and Trading Volume (2025)
  163. Volatility forecasting with machine learning and intraday commonality (2024)
  164. Factor models, machine learning, and asset pricing (2022)
  165. Presidential address: Discount rates (2011)
  166. Automated volatility forecasting (2025)
  167. Empirical Dynamic Asset Pricing: Model Specification and Econometric Assessment (2006)
  168. Empirical performance of alternative option pricing models (1997)
  169. A New Model for the Joint Valuation of S&P 500 and VIX Options: Specification Analysis (2025)
  170. VIX derivatives, hedging and vol-of-vol risk (2020)
  171. The quadratic rough Heston model and the joint S&P 500/VIX smile calibration problem (2020)
  172. 4/2 rough and smooth (2025)
  173. The volatility surface: A Practitioner’s Guide (2011)
  174. Model complexity and out-of-sample performance: Evidence from S&P 500 index returns (2018)

8 of 182 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.