Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha

A common issue faced by investors who use technical analysis is the reconciliation of conflicting trading signals, especially when these signals are highly correlated, such as those generated by multiple moving averages. This study expands on a model-free algorithm inspired by reinforcement learning...

Full description

Saved in:
Bibliographic Details
Main Authors: Javier H. Ospina-Holguín, Ana M. Padilla-Ospina
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Cogent Economics & Finance
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/23322039.2025.2490818
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850125090286469120
author Javier H. Ospina-Holguín
Ana M. Padilla-Ospina
author_facet Javier H. Ospina-Holguín
Ana M. Padilla-Ospina
author_sort Javier H. Ospina-Holguín
collection DOAJ
description A common issue faced by investors who use technical analysis is the reconciliation of conflicting trading signals, especially when these signals are highly correlated, such as those generated by multiple moving averages. This study expands on a model-free algorithm inspired by reinforcement learning to address the challenge of reconciling trading signals while taking transaction costs into account. The algorithm is trained to optimize alpha, a widely used measure of risk-adjusted return. Principal component analysis is utilized to reduce the dimensionality of a modified version of moving average signals, which are then used to define the input state. A policy network, represented by a feedforward neural network, is trained using historical data to convert states into trading actions. An evaluation network calculates and optimizes alpha by adjusting the policy network’s parameters. The algorithm utilizes a zero-arbitrage portfolio to accurately isolate alpha from the underlying asset’s return. By combining 199 simple moving average signals in a systematic manner, the algorithm was able to maximize the [Formula: see text] asset pricing model alpha using a high-volatility United States stock portfolio as a risky asset. The algorithm demonstrates superior performance compared to both individual moving average signals and existing combination algorithms.
format Article
id doaj-art-592cea17c9f54a5f9c43231e2320cdd0
institution OA Journals
issn 2332-2039
language English
publishDate 2025-12-01
publisher Taylor & Francis Group
record_format Article
series Cogent Economics & Finance
spelling doaj-art-592cea17c9f54a5f9c43231e2320cdd02025-08-20T02:34:10ZengTaylor & Francis GroupCogent Economics & Finance2332-20392025-12-0113110.1080/23322039.2025.2490818Reinforcement learning meets technical analysis: combining moving average rules for optimal alphaJavier H. Ospina-Holguín0Ana M. Padilla-Ospina1Department of Accounting and Finance, Universidad del Valle, Cali, ColombiaDepartment of Administration and Organizations, Universidad del Valle, Cali, ColombiaA common issue faced by investors who use technical analysis is the reconciliation of conflicting trading signals, especially when these signals are highly correlated, such as those generated by multiple moving averages. This study expands on a model-free algorithm inspired by reinforcement learning to address the challenge of reconciling trading signals while taking transaction costs into account. The algorithm is trained to optimize alpha, a widely used measure of risk-adjusted return. Principal component analysis is utilized to reduce the dimensionality of a modified version of moving average signals, which are then used to define the input state. A policy network, represented by a feedforward neural network, is trained using historical data to convert states into trading actions. An evaluation network calculates and optimizes alpha by adjusting the policy network’s parameters. The algorithm utilizes a zero-arbitrage portfolio to accurately isolate alpha from the underlying asset’s return. By combining 199 simple moving average signals in a systematic manner, the algorithm was able to maximize the [Formula: see text] asset pricing model alpha using a high-volatility United States stock portfolio as a risky asset. The algorithm demonstrates superior performance compared to both individual moving average signals and existing combination algorithms.https://www.tandfonline.com/doi/10.1080/23322039.2025.2490818Signal combinationtechnical analysismoving average trading rulesasset pricingalpha optimizationreinforcement learning
spellingShingle Javier H. Ospina-Holguín
Ana M. Padilla-Ospina
Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
Cogent Economics & Finance
Signal combination
technical analysis
moving average trading rules
asset pricing
alpha optimization
reinforcement learning
title Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
title_full Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
title_fullStr Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
title_full_unstemmed Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
title_short Reinforcement learning meets technical analysis: combining moving average rules for optimal alpha
title_sort reinforcement learning meets technical analysis combining moving average rules for optimal alpha
topic Signal combination
technical analysis
moving average trading rules
asset pricing
alpha optimization
reinforcement learning
url https://www.tandfonline.com/doi/10.1080/23322039.2025.2490818
work_keys_str_mv AT javierhospinaholguin reinforcementlearningmeetstechnicalanalysiscombiningmovingaveragerulesforoptimalalpha
AT anampadillaospina reinforcementlearningmeetstechnicalanalysiscombiningmovingaveragerulesforoptimalalpha