Paper: arXiv 2509.04812

Authors: Hongyi Liu

Abstract

We propose a new pseudo-Siamese Network for Asset Pricing (SNAP) model, based on deep learning approaches, for conditional asset pricing. Our model allows for the deep alpha, deep beta and deep factor risk premia conditional on high dimensional observable information of financial characteristics and macroeconomic states, while storing the long-term dependency of the informative features through long short-term memory network. We apply this method to monthly U.S. stock returns from 1970-2019 and find that our pseudo-SNAP model outperforms the benchmark approaches in terms of out-of-sample prediction and out-of-sample Sharpe ratio. In addition, we also apply our method to calculate deep mispricing errors which we use to construct an arbitrage portfolio K-Means clustering. We find that the arbitrage portfolio has significant alphas.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • 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 employs advanced deep learning architectures (pseudo-Siamese networks with LSTM) and derives conditional asset pricing models with nonlinear stochastic discount factors, indicating high mathematical complexity. It demonstrates empirical rigor through out-of-sample backtesting on a long historical dataset (1970-2019), reporting performance metrics like Sharpe ratios and significant alphas on constructed portfolios.

Research Flowchart

  flowchart TD
  Goal["Research Goal: Develop a deep learning conditional asset pricing model & quantify mispricing"]
  Data["Data: U.S. Stocks (1970-2019)<br>Financial & Macro Characteristics"]
  Method["Methodology: Pseudo-Siamese Network for Asset Pricing (SNAP)<br>Deep Alpha/Beta + LSTM"]
  Proc["Computational Process:<br>1. Deep Factor Estimation<br>2. Mispricing Error Calculation<br>3. K-Means Clustering"]
  Out["Key Findings: High Out-of-Sample Sharpe Ratio<br>Significant Alphas in Arbitrage Portfolio"]

Cited References (138)

  1. Estimation and Inference of Impulse Responses by Local Projections (2005) — link
  2. Macrofinancial History and the New Business Cycle Facts
  3. Barro-ursua macroeconomic data (2010) — [link](https://scholar. harvard. edu/barro/publications/barro-ursua-macroeconomic-data)
  4. The Cross - Section of Volatility and Expected Returns (2006) — link
  5. Scaling Up Spike -and- Slab Models for Unsupervised Feature Learning (2013) — link
  6. Does Academic Research Destroy Stock Return Predictability ?: Does Academic Research Destroy Stock Return Predictability ? (2016) — link
  7. A skeptical appraisal of asset pricing tests (2010) — link
  8. A five-factor asset pricing model (2015) — link
and 130 more
  1. International tests of a five-factor asset pricing model (2017) — link
  2. The common factor in idiosyncratic volatility: Quantitative asset pricing implications (2016) — link
  3. Comparing Asset Pricing Models (2018) — link
  4. Estimating latent asset-pricing factors (2020) — link
  5. Asset Pricing with Fading Memory (2019) — link
  6. Valuation Risk and Asset Pricing : Valuation Risk and Asset Pricing (2016) — link
  7. Deep Learning in Asset Pricing (2020) — link
  8. Factors That Fit the Time Series and Cross - Section of Stock Returns (2019) — link
  9. Parameter Learning in General Equilibrium : The Asset Pricing Implications (2016) — link
  10. Equilibrium Asset Prices and No - Arbitrage with Portfolio Constraints (1997) — link
  11. Deep Learning for Predicting Asset Returns (2018) — link
  12. The Cross - Section of Expected Stock Returns (1992) — link
  13. Risk, Return , and Equilibrium : Empirical Tests (1973) — link
  14. Understanding deep learning requires rethinking generalization (2017) — link
  15. A Neural Representation of Sketch Drawings (2017) — link
  16. Shrinking the cross-section (2020) — link
  17. Dissecting Characteristics Nonparametrically (2020) — link
  18. No Arbitrage and Arbitrage Pricing : A New Approach (1993) — link
  19. Evidence on the Characteristics of Cross Sectional Variation in Stock Returns (1997) — link
  20. Interpreting Factor Models (2018) — link
  21. The Role of Conditioning Information in Deducing Testable Restrictions Implied by Dynamic Asset Pricing Models (1987) — link
  22. Characteristics are covariances: A unified model of risk and return (2019) — link
  23. Betting against beta (2014) — link
  24. Overcoming limits of arbitrage: Theory and evidence (2014) — link
  25. Efficient Semiparametric Estimation of the Fama – French Model and Extensions (2012) — link
  26. Characteristics- Based Factors (2020) — link
  27. The Cross - Section of Risk and Return
  28. A Dynamic Model of Characteristic - Based Return Predictability (2019) — link
  29. Conditioning Variables and the Cross Section of Stock Returns (1999) — link
  30. On Stable Factor Structures in the Pricing of Risk : Do Time - Varying Betas Help or Hurt ? (1998) — link
  31. Hedging Risk Factors (2019) — link
  32. Digesting Anomalies : An Investment Approach (2015) — link
  33. Aggregation of Information About the Cross Section of Stock Returns : A Latent Variable Approach (2017) — link
  34. Portfolio formation with preselection using deep learning from long-term financial data (2020) — link
  35. Economic Forces and the Stock Market (1986) — link
  36. Betting Against Betting Against Beta (2018) — link
  37. The Characteristics that Provide Independent Information about Average U . S . Monthly Stock Returns (2017) — link
  38. Factor Investing : A Bayesian Hierarchical Approach (2020) — link
  39. Implications of Security Market Data for Models of Dynamic Economies (1991) — link
  40. Machine Learning versus Economic Restrictions : Evidence from Stock Return Predictability (2020) — link
  41. FRED - MD : A Monthly Database for Macroeconomic Research (2016) — link
  42. Mean- Variance Theory in Complete Markets (1982) — link
  43. Resurrecting the ( C ) CAPM : A Cross ‐ Sectional Test When Risk Premia Are Time ‐ Varying (2001) — link
  44. Empirical Asset Pricing via Machine Learning (2020) — link
  45. Long- Run Stockholder Consumption Risk and Asset Returns (2009) — link
  46. Extra- Market Components of Covariance in Security Returns (1974) — link
  47. Risk and return in an equilibrium APT : Application of a new test methodology (1988) — link
  48. Semiparametric estimation of a characteristic-based factor model of common stock returns (2007) — link
  49. Predictive regressions with time-varying coefficients (2012) — link
  50. Nonparametric regression using deep neural networks with ReLU activation function (2020) — link
  51. Asset Pricing : Revised Edition (2009)
  52. Investor Sentiment and the Cross - Section of Stock Returns (2006) — link
  53. Convolutional RNN : an Enhanced Model for Extracting Features from Sequential Data (2017) — link
  54. Multivariate time series forecasting via attention-based encoder–decoder framework (2020) — link
  55. Presidential Address : Discount Rates (2011) — link
  56. Soft biometric thermal face recognition using FWT and LDA feature extraction method with RBM DBN and FFNN classifier algorithms (2017) — link
  57. Macroeconomic Indicator Forecasting with Deep Neural Networks (2017) — link
  58. A Simple and Effective Model - Based Variable Importance Measure (2018) — link
  59. Use of some sensitivity criteria for choosing networks with good generalization ability (1995) — link
  60. Interpreting neural-network connection weights (1991)
  61. Siamese Neural Networks for One -shot Image Recognition
  62. … and the Cross - Section of Expected Returns (2016) — link
  63. The arbitrage theory of capital asset pricing (1976) — link
  64. The Delisting Bias in CRSP ’s Nasdaq Data and Its Implications for the Size Effect (1999) — link
  65. Cross-sectional Tests of Conditional Asset Pricing Models : Evidence from the German Stock Market (2007) — link
  66. Empirical Cross - Sectional Asset Pricing (2013) — link
  67. The Variation of Economic Risk Premiums (1991) — link
  68. Time-varying conditional covariances in tests of asset pricing models (1989) — link
  69. The Conditional CAPM and the Cross - Section of Expected Returns (1996) — link
  70. Tests of the Relations Among Marketwide Factors , Firm - Specific Variables , and Stock Returns Using a Conditional Asset Pricing Model (1996) — link
  71. Estimation and Evaluation of Conditional Asset Pricing Models (2011) — link
  72. Firm Characteristics , Cross - Sectional Regression Estimates , and Asset Pricing Tests (2020) — link
  73. Signature Verification using a " Siamese " Time Delay Neural Network (1994) — link
  74. Deep Fundamental Factor Models (2020) — link
  75. A Dynamic Network of Arbitrage Characteristics (2020) — link
  76. Taming the Factor Zoo : A Test of New Factors (2020) — link
  77. Characteristic-based mean-variance portfolio choice (2012) — link
  78. Arbitrage Portfolios (2020) — link
  79. Models of Financial Stability and Their Application in Stress Tests (2018) — link
  80. Power Enhancement in High - Dimensional Cross - Sectional Tests (2015) — link
  81. Signature Verification using a " Siamese " Time Delay Neural Network (1994) — link
  82. On Persistence in Mutual Fund Performance (1997) — link
  83. Common risk factors in the returns on stocks and bonds (1993) — link
  84. Multifactor Explanations of Asset Pricing Anomalies (1996) — link
  85. Dissecting Anomalies (2008) — link
  86. Multilayer feedforward networks are universal approximators (1989) — link
  87. Long Short - Term Memory (1997) — link
  88. Deep Learning : mathematics and neuroscience (2016)
  89. Autoencoder asset pricing models (2020) — link
  90. Generating Sequences With Recurrent Neural Networks (2014) — link
  91. Recurrent Neural Network Regularization (2015) — link
  92. Learning a similarity metric discriminatively, with application to face verification (2005) — link
  93. Deep learning (2016) — link
  94. The Elements of Statistical Learning : Data Mining , Inference , and Prediction , Second Edition (2009)
  95. Evaluating the specification errors of asset pricing models (2001) — link
  96. A Cross - Sectional Test of an Investment - Based Asset Pricing Model (1996) — link
  97. Dropout: a simple way to prevent neural networks from overfitting (2014)
  98. Adam: A Method for Stochastic Optimization (2017) — link
  99. Forecasting Economics and Financial Time Series : ARIMA vs. LSTM (2018) — link
  100. Stock Market Prediction Using LSTM Recurrent Neural Network (2020) — link
  101. Does Finance Benefit Society ? A Language Embedding Approach (2020) — link
  102. Credit Booms Gone Bust : Monetary Policy , Leverage Cycles , and Financial Crises , 1870-2008 (2012) — link
  103. The Rate of Return on Everything , 1870–2015 (2019) — link
  104. The Memory Advantage of Long Short - Term Memory Networks for Bond Yield Forecasting (2019) — link
  105. Learning long-term dependencies with gradient descent is difficult (1994) — link
  106. Note on Grouping (1957) — link
  107. An accurate comparison of methods for quantifying variable importance in artificial neural networks using simulated data (2004) — link
  108. De- Biased Machine Learning of Global and Local Parameters Using Regularized Riesz Representers (2020) — link
  109. Double/debiased machine learning for treatment and structural parameters (2018) — link
  110. Double/debiased machine learning for treatment and structural parameters (2018) — link
  111. LSTM : A Search Space Odyssey (2017) — link
  112. Chapter 10 Forecasting with Many Predictors (2006) — link
  113. The Poor Predictive Performance of Asset Pricing Models (2008) — link
  114. Capital Asset Prices : A Theory of Market Equilibrium Under Conditions of Risk * (1964) — link
  115. The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets (1965) — link
  116. Forecasting in dynamic factor models subject to structural instability (2009)
  117. The intertemporal capital asset pricing model with dynamic conditional correlations (2010) — link
  118. Regression-based estimation of dynamic asset pricing models (2015) — link
  119. Alpha Go Everywhere : Machine Learning and International Stock Returns (2020) — link
  120. Publication Bias and the Cross - Section of Stock Returns (2020) — link
  121. The Limits of p- Hacking : Some Thought Experiments (2020) — link
  122. Chapter 15 Anomalies and market efficiency (2003) — link
  123. Replicating Anomalies (2020) — link
  124. The 1918 “ Spanish Flu ” in Spain (2008) — link
  125. Deep Learning for Conditional Asset Pricing Models — → 2509.04812
  126. A Survey of Deep Learning for Scientific Discovery (2020) — link
  127. AlphaPortfolio : Direct Construction through Reinforcement Learning and Economically Interpretable AI (2020) — link
  128. Deep Learning of Representations for Unsupervised and Transfer Learning
  129. International Stock Return Predictability : What Is the Role of the United States ? (2013) — link
  130. A Tale of Two Types : Generalists vs. Specialists in Asset Management (2020) — link

1 of 138 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.