Persistence in factor-based supervised learning models

In this paper, we document the importance of memory in machine learning (ML)-based models relying on firm characteristics for asset pricing. We find that predictive algorithms perform best when they are trained on long samples, with long-term returns as dependent variables. In addition, we report th...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autor principal:	Guillaume Coqueret
Formato:	article
Lenguaje:	EN
Publicado:	KeAi Communications Co., Ltd. 2022
Materias:	C45 C53 G11 G12 Electronic computers. Computer science QA75.5-76.95 Finance HG1-9999
Acceso en línea:	https://doaj.org/article/3d705e58b42b4cf7a6d9cbe210af6116
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:3d705e58b42b4cf7a6d9cbe210af6116
record_format	dspace
spelling	oai:doaj.org-article:3d705e58b42b4cf7a6d9cbe210af61162021-11-22T04:27:57ZPersistence in factor-based supervised learning models2405-918810.1016/j.jfds.2021.10.002https://doaj.org/article/3d705e58b42b4cf7a6d9cbe210af61162022-11-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S2405918821000143https://doaj.org/toc/2405-9188In this paper, we document the importance of memory in machine learning (ML)-based models relying on firm characteristics for asset pricing. We find that predictive algorithms perform best when they are trained on long samples, with long-term returns as dependent variables. In addition, we report that persistent features play a prominent role in these models. When applied to portfolio choice, we find that investors are always better off predicting annual returns, even when rebalancing at lower frequencies (monthly or quarterly). Our results remain robust to transaction costs and risk scaling, thus providing useful indications to quantitative asset managers.Guillaume CoqueretKeAi Communications Co., Ltd.articleC45C53G11G12Electronic computers. Computer scienceQA75.5-76.95FinanceHG1-9999ENJournal of Finance and Data Science, Vol 8, Iss , Pp 12-34 (2022)
institution	DOAJ
collection	DOAJ
language	EN
topic	C45 C53 G11 G12 Electronic computers. Computer science QA75.5-76.95 Finance HG1-9999
spellingShingle	C45 C53 G11 G12 Electronic computers. Computer science QA75.5-76.95 Finance HG1-9999 Guillaume Coqueret Persistence in factor-based supervised learning models
description	In this paper, we document the importance of memory in machine learning (ML)-based models relying on firm characteristics for asset pricing. We find that predictive algorithms perform best when they are trained on long samples, with long-term returns as dependent variables. In addition, we report that persistent features play a prominent role in these models. When applied to portfolio choice, we find that investors are always better off predicting annual returns, even when rebalancing at lower frequencies (monthly or quarterly). Our results remain robust to transaction costs and risk scaling, thus providing useful indications to quantitative asset managers.
format	article
author	Guillaume Coqueret
author_facet	Guillaume Coqueret
author_sort	Guillaume Coqueret
title	Persistence in factor-based supervised learning models
title_short	Persistence in factor-based supervised learning models
title_full	Persistence in factor-based supervised learning models
title_fullStr	Persistence in factor-based supervised learning models
title_full_unstemmed	Persistence in factor-based supervised learning models
title_sort	persistence in factor-based supervised learning models
publisher	KeAi Communications Co., Ltd.
publishDate	2022
url	https://doaj.org/article/3d705e58b42b4cf7a6d9cbe210af6116
work_keys_str_mv	AT guillaumecoqueret persistenceinfactorbasedsupervisedlearningmodels
_version_	1718418154904879104

Persistence in factor-based supervised learning models

Ejemplares similares