---
title: Machine Learning Predictability: Factors Affecting Stock Return Forecasts
url: https://www.ml-quant.com/papers/ssrn/4552395/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 4552395
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4552395
featured: 2023-08-30
citations: unknown
topic: Econometrics & Forecasting
---


# Machine Learning Predictability: Factors Affecting Stock Return Forecasts

Factors Affecting Stock Return Forecasts: The study shows that while machine learning strategies can predict short-term returns for small firms and early historical data, they have not provided significant economic gains for most of the U.S. market in the past 20 years.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4552395
- Identifier: SSRN 4552395
- Released: 2023-08-18
- First featured: Quant Letter No. 14 (2023-08-30): https://www.ml-quant.com/issues/2023-08-30/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Econometrics & Forecasting

## Related

- [TKAN: Temporal Kolmogorov-Arnold Networks](https://www.ml-quant.com/papers/ssrn/4825654/): The article presents Temporal Kolomogorov-Arnold Networks (TKANs), a new neural network design that merges the benefits of Recurrent Neural Networks and Long Short-Term Memory for improved multistep time series forecasting.
- [Robust agents learn causal world models](https://www.ml-quant.com/papers/arxiv/2402.10877/): The research suggests that intelligent agents must learn an approximate causal model to generalize to new domains, impacting fields like transfer learning and causal inference.
- [Retail Demand Forecasting: A Comparative Study for Multivariate Time Series](https://www.ml-quant.com/papers/arxiv/2308.11939/): The study creates improved retail demand prediction models using macroeconomic factors and past sales data.
- [Augmented HAR](https://www.ml-quant.com/papers/ssrn/4516177/): The Augmented HAR algorithm, combined with artificial neural networks, enhances forecast accuracy for stocks with less than seven years of data.
- [Automated regime classification in multidimensional time series data using sliced Wasserstein k-means clustering](https://www.ml-quant.com/papers/arxiv/2310.01285/): The paper explores the use of Wasserstein k-means clustering on multidimensional time series data for automated regime detection, proving its effectiveness in identifying different market regimes in real financial time series.
- [Supervised Dynamic PCA: Linear Dynamic Forecasting with Many Predictors](https://www.ml-quant.com/papers/arxiv/2307.07689/): A new dynamic forecasting method using supervised Principal Component Analysis (PCA) has been introduced, which is more effective in predicting U.S. macroeconomic variables.