The search for time-series predictability-based anomalies

This paper introduces a new algorithm for exploiting time-series predictability-based patterns to obtain an abnormal return, or alpha, with respect to a given benchmark asset pricing model. The algorithm proposes a deterministic daily market timing strategy that decides between being fully invested...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Javier Humberto Ospina-Holguín, Ana Milena Padilla-Ospina
Formato:	article
Lenguaje:	EN
Publicado:	Vilnius Gediminas Technical University 2021
Materias:	stock market investment algorithm trading rules alpha maximization market timing artificial intelligence Business HF5001-6182
Acceso en línea:	https://doaj.org/article/db9d8a0a547b43dbbb152a59426a965c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:db9d8a0a547b43dbbb152a59426a965c
record_format	dspace
spelling	oai:doaj.org-article:db9d8a0a547b43dbbb152a59426a965c2021-11-29T16:03:52ZThe search for time-series predictability-based anomalies10.3846/jbem.2021.156501611-16992029-4433https://doaj.org/article/db9d8a0a547b43dbbb152a59426a965c2021-11-01T00:00:00Zhttps://www.aviation.vgtu.lt/index.php/JBEM/article/view/15650https://doaj.org/toc/1611-1699https://doaj.org/toc/2029-4433 This paper introduces a new algorithm for exploiting time-series predictability-based patterns to obtain an abnormal return, or alpha, with respect to a given benchmark asset pricing model. The algorithm proposes a deterministic daily market timing strategy that decides between being fully invested in a risky asset or in a risk-free asset, with the trading rule represented by a parametric perceptron. The optimal parameters are sought in-sample via differential evolution to directly maximize the alpha. Successively using two modern asset pricing models and two different portfolio weighting schemes, the algorithm was able to discover an undocumented anomaly in the United States stock market cross-section, both out-of-sample and using small transaction costs. The new algorithm represents a simple and flexible alternative to technical analysis and forecast-based trading rules, neither of which necessarily maximizes the alpha. This new algorithm was inspired by recent insights into representing reinforcement learning as evolutionary computation. First published online 29 November 2021 Javier Humberto Ospina-HolguínAna Milena Padilla-OspinaVilnius Gediminas Technical Universityarticlestock marketinvestment algorithmtrading rulesalpha maximizationmarket timingartificial intelligenceBusinessHF5001-6182ENJournal of Business Economics and Management (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	stock market investment algorithm trading rules alpha maximization market timing artificial intelligence Business HF5001-6182
spellingShingle	stock market investment algorithm trading rules alpha maximization market timing artificial intelligence Business HF5001-6182 Javier Humberto Ospina-Holguín Ana Milena Padilla-Ospina The search for time-series predictability-based anomalies
description	This paper introduces a new algorithm for exploiting time-series predictability-based patterns to obtain an abnormal return, or alpha, with respect to a given benchmark asset pricing model. The algorithm proposes a deterministic daily market timing strategy that decides between being fully invested in a risky asset or in a risk-free asset, with the trading rule represented by a parametric perceptron. The optimal parameters are sought in-sample via differential evolution to directly maximize the alpha. Successively using two modern asset pricing models and two different portfolio weighting schemes, the algorithm was able to discover an undocumented anomaly in the United States stock market cross-section, both out-of-sample and using small transaction costs. The new algorithm represents a simple and flexible alternative to technical analysis and forecast-based trading rules, neither of which necessarily maximizes the alpha. This new algorithm was inspired by recent insights into representing reinforcement learning as evolutionary computation. First published online 29 November 2021
format	article
author	Javier Humberto Ospina-Holguín Ana Milena Padilla-Ospina
author_facet	Javier Humberto Ospina-Holguín Ana Milena Padilla-Ospina
author_sort	Javier Humberto Ospina-Holguín
title	The search for time-series predictability-based anomalies
title_short	The search for time-series predictability-based anomalies
title_full	The search for time-series predictability-based anomalies
title_fullStr	The search for time-series predictability-based anomalies
title_full_unstemmed	The search for time-series predictability-based anomalies
title_sort	search for time-series predictability-based anomalies
publisher	Vilnius Gediminas Technical University
publishDate	2021
url	https://doaj.org/article/db9d8a0a547b43dbbb152a59426a965c
work_keys_str_mv	AT javierhumbertoospinaholguin thesearchfortimeseriespredictabilitybasedanomalies AT anamilenapadillaospina thesearchfortimeseriespredictabilitybasedanomalies AT javierhumbertoospinaholguin searchfortimeseriespredictabilitybasedanomalies AT anamilenapadillaospina searchfortimeseriespredictabilitybasedanomalies
_version_	1718407255771054080

The search for time-series predictability-based anomalies

Ejemplares similares