Paper: arXiv 2407.13685

Abstract

Trend following and momentum investing are common strategies employed by asset managers. Even though they can be helpful in the proper situations, they are limited in the sense that they work just by looking at past, as if we were driving with our focus on the rearview mirror. In this paper, we advocate for the use of Artificial Intelligence and Machine Learning techniques to predict future market trends. These predictions, when done properly, can improve the performance of asset managers by increasing returns and reducing drawdowns.

Complexity vs Empirical Score

  • Math Complexity: 4.0/10
  • Empirical Rigor: 3.0/10
  • Quadrant: Philosophers — conceptual discussion, limited math and data

Why this score: The paper advocates for deep learning over linear models and discusses concepts like universal approximators, but the excerpt contains no mathematical derivations or formulas. Empirically, it lacks code, backtests, or statistical metrics, relying instead on conceptual discussions and a single chart example.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Predict Market Trends using AI"] --> B["Data: Multi-asset Historical Data"]
  B --> C["Methodology: Deep Learning Models<br>e.g., CNN, LSTM, Transformers"]
  C --> D["Computational Process:<br>Training & Hyperparameter Tuning"]
  D --> E["Computational Process:<br>Validation on Out-of-Sample Data"]
  E --> F["Key Findings:<br>AI beats Trend Following"]
  F --> G["Outcomes:<br>Higher Returns & Reduced Drawdowns"]

Cited References (120)

  1. On the robustness of interpretability methods (2018) — link
  2. Visualizing the effects of predictor variables in black box supervised learning models (2020)
  3. Learning Deep Architectures for AI (2009) — link
  4. Algorithms for Hyper-Parameter Optimization (2011) — link
  5. Random Search for Hyper-Parameter Optimization (2012) — link
  6. Making a Science of Model Search: Hyperparameter Optimization in Hundreds of Dimensions for Vision Architectures (2013) — link
  7. Redes Neuronales & Deep Learning - Volumen I: Entrenamiento de redes neuronales artificiales [Neural Networks and Deep Learning - Volume I: Training Artificial Neural Networks, in Spanish] (2019) — link
  8. Redes Neuronales & Deep Learning - Volumen II: Regularización, optimización & arquitecturas especializadas [Neural Networks and Deep Learning - Volume II: Regularization, Optimization, and Specialized Architectures, in Spanish] (2019) — link
and 112 more
  1. Explanation and justification in machine learning: A survey (2017)
  2. A Racing Algorithm for Configuring Metaheuristics (2002)
  3. Training with Noise is Equivalent to Tikhonov Regularization (1995) — link
  4. Kaggle forecasting competitions: An overlooked learning opportunity (2021) — link
  5. Random Forests (2001) — link
  6. A Survey on the Explainability of Supervised Machine Learning (2021) — link
  7. Regularization and Feedforward artificial neural network training with noise (2003) — link
  8. A LSTM-based method for stock returns prediction: A case study of China stock market (2015) — link
  9. Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation (2014)
  10. Stocks on the Move: Beating the Market with Hedge Fund Momentum Strategies (2015)
  11. Introduction to Derivative-Free Optimization (2009)
  12. Support-Vector Networks (1995) — link
  13. Approximation by superpositions of a sigmoidal function (1989) — link
  14. A fast and elitist multiobjective genetic algorithm: NSGA-II (2002)
  15. An effective algorithm for hyperparameter optimization of neural networks (2017)
  16. Fast Monte Carlo Algorithms for Matrices III: Computing a Compressed Approximate Matrix Decomposition (2006) — link
  17. Intelligence Emerging: Adaptivity and Search in Evolving Neural Systems (2015)
  18. Support vector regression machines (1996)
  19. Finding Structure in Time (1990) — link
  20. Desarrollo de un sistema de trading algorítmico basado en Deep Learning [Develoment of an Algorithmic Trading System based on Deep Learning, in Spanish] (2022)
  21. Efficient and Robust Automated Machine Learning (2015) — link
  22. Auto-sklearn: Efficient and Robust Automated Machine Learning (2019) — link
  23. Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learning (2022) — link
  24. All Models are Wrong, but Many are Useful: Learning a Variable’s Importance by Studying an Entire Class of Prediction Models Simultaneously (2019) — link
  25. Greedy Function Approximation: A Gradient Boosting Machine (2001) — link
  26. PREDICTIVE LEARNING VIA RULE ENSEMBLES (2008)
  27. Neural network model for a mechanism of pattern recognition unaffected by shift in position - Neocognitron - (1979)
  28. Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position (1980) — link
  29. On the approximate realization of continuous mappings by neural networks (1989) — link
  30. Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning (2016) — link
  31. Peeking Inside the Black Box: Visualizing Statistical Learning With Plots of Individual Conditional Expectation (2015) — link
  32. Deep Learning (2016) — link
  33. A simple and effective model-based variable importance measure (2018) — link
  34. Weight decay backpropagation for noisy data (1998) — link
  35. Hyperparameter optimization: A spectral approach (2017) — link
  36. Stochastic Neighbor Embedding (2002) — link
  37. Visualizing data using t-SNE. (2008)
  38. Simplifying neural nets by discovering flat minima (1995) — link
  39. Long Short-Term Memory (1997) — link
  40. Ridge Regression: Biased Estimation for Nonorthogonal Problems (1970)
  41. Discovering additive structure in black box functions (2004)
  42. Multilayer feedforward networks are universal approximators (1989) — link
  43. CNNpred: CNN-based stock market prediction using a diverse set of variables (2019) — link
  44. Analysis of a Complex of Statistical Variables into Principal Components (1933)
  45. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift (2015) — link
  46. Learning of modular structured networks (1995) — link
  47. Population based training of neural networks (2017) — link
  48. Non-stochastic best arm identification and hyperparameter optimization (2016)
  49. Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency (1993) — link
  50. An analysis of noise in recurrent neural networks: convergence and generalization (1996) — link
  51. 101 Formulaic Alphas (2016)
  52. Gaussian Process Bandit Optimisation with Multi-fidelity Evaluations (2016) — link
  53. Examples are not enough, learn to criticize! Criticism for Interpretability (2016) — link
  54. The effects of scheduling, workload type and consolidation scenarios on virtual machine performance and their prediction through optimized artificial neural networks (2011)
  55. On the accuracy of mapping by neural networks trained by backpropagation with forgetting (1996) — link
  56. Nonlinear Principal Component Analysis Using Autoassociative Neural Networks (1991)
  57. ImageNet Classification with Deep Convolutional Neural Networks (2012) — link
  58. The development of the time-delay neural network architecture for speech recognition (1988)
  59. A time-delay neural network architecture for isolated word recognition (1990) — link
  60. Phoneme recognition using time-delay neural networks (1989) — link
  61. Relative Strength as a Criterion for Investment Selection (1967)
  62. Hyperband: A novel bandit-based approach to hyperparameter optimization (2018)
  63. A generalized framework for population based training (2019)
  64. A system for massively parallel hyperparameter tuning (2020)
  65. The irace package: Iterated racing for automatic algorithm configuration (2016)
  66. A unified approach to interpreting model predictions (2017)
  67. What makes an optimization problem hard? (1996) — link
  68. Bayesian Interpolation (1992) — link
  69. Gradient-based Hyperparameter Optimization through Reversible Learning (2015) — link
  70. Tensor-CUR Decompositions for Tensor-based Data (2006) — link
  71. Tensor-CUR Decompositions for Tensor-Based Data (2008) — link
  72. K-sparse autoencoders (2013) — link
  73. Global optimization of Lipschitz functions (2017)
  74. Connectionist Models (2009)
  75. Umap: Uniform manifold approximation and projection for dimension reduction (2018) — link
  76. Evolving deep neural networks (2024)
  77. Explanation in artificial intelligence: Insights from the social sciences (2019) — link
  78. Interpretable machine learning: Interpretable Machine Learning: A Guide for Making Black Box Models Interpretable (2021) — link
  79. Universal Approximation Depth and Errors of Narrow Belief Networks with Discrete Units (2014) — link
  80. A Focused Backpropagation Algorithm for Temporal Pattern Recognition (1989) — link
  81. Adding Gradient Noise Improves Learning for Very Deep Networks (2015) — link
  82. Stock market’s price movement prediction with LSTM neural networks (2017) — link
  83. On Lines and Planes of Closest Fit to Systems of Points in Space (1901) — link
  84. Analyzing noise in autoencoders and deep networks (2014) — link
  85. Why should i trust you? Explaining the predictions of any classifier (2016)
  86. Anchors: High-precision model-agnostic explanations (2018)
  87. Contractive auto-encoders: Explicit invariance during feature extraction (2011)
  88. How Does Batch Normalization Help Optimization? (2018) — link
  89. Nonlinear Component Analysis as a Kernel Eigenvalue Problem (1998) — link
  90. Dynamic Mode Decomposition of Numerical and Experimental Data (2008)
  91. Dynamic Mode Decomposition of Numerical and Experimental Data (2010)
  92. Traffic sign recognition with multi-scale Convolutional Networks (2011) — link
  93. A VALUE FOR n-PERSON GAMES (1953)
  94. Creating artificial neural networks that generalize (1991) — link
  95. Fooling LIME and SHAP: Adversarial attacks on post hoc explanation methods (2020)
  96. Practical Bayesian Optimization of Machine Learning Algorithms (2012) — link
  97. Scalable Bayesian Optimization Using Deep Neural Networks (2015) — link
  98. Gaussian process optimization in the bandit setting: no regret and experimental design (2010)
  99. Dropout: A Simple Way to Prevent Neural Networks from Overfitting (2014) — link
  100. Explaining prediction models and individual predictions with feature contributions (2014)
  101. Freeze-Thaw Bayesian Optimization (2014) — link
  102. Regression Shrinkage and Selection via the Lasso (1996) — link
  103. Solutions of Ill-Posed Problems (1977)
  104. Attention is All you Need (2017) — link
  105. Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion. (2010)
  106. An empirical analysis of dropout in piecewise linear networks (2014) — link
  107. Generalization by Weight-Elimination with Application to Forecasting (1990) — link
  108. Does Trend Following Work on Stocks? (2005)
  109. The Lack of a Priori Distinctions Between Learning Algorithms (1996) — link
  110. No free lunch theorems for optimization
  111. Causal Interpretations of Black-Box Models (2021) — link
  112. Regularization and Variable Selection via the Elastic Net (2005) — link

0 of 120 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.