Answers are grounded in this paper and cite their sources.
Structured review reveals median error reductions of 20.3% across financial markets, indicating experimental gains lack universal real-world superiority.
Forecasting financial markets remains difficult because price series are noisy, nonlinear, nonstationary, and subject to structural breaks. This structured review synthesizes empirical machine learning (ML) and deep learning (DL) research on equities, indices, commodities, foreign exchange, and cryptocurrencies, covering classical ML, recurrent networks, CNN-based models, attention and Transformer architectures, multimodal and graph-based systems, and reinforcement-learning approaches. Because studies differ in task formulation, horizon, data frequency, target variable, feature set, validation design, and metric choice, the review combines qualitative synthesis with structured quantitative evidence mapping rather than a single pooled effect size. The paired-error aggregation identified 17 peer-reviewed studies and 47 same-dataset, same-horizon proposed-versus-baseline error comparisons. On a study-level median basis, this subset showed a median relative error reduction of 20.3%, with an interquartile range of 5.7%-50.7% and a full range of −0.8%-71.5%. These results indicate that ML/DL methods often improve reported performance in controlled experiments, but they do not establish universal real-world superiority. Comparability remains limited by validation design, transaction costs, slippage, market regimes, and inconsistent uncertainty reporting. Future work should prioritize leakage-safe validation, calibration, interpretability, uncertainty quantification, and realistic economic evaluation.
No takes yet. Share an insight, caveat, or question.
Khemka et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: