Using machine learning to identify and measure the family influence on companies

Mario Daniele Amore (), Valentino D’angelo (), Isabelle Le Breton-Miller (), Danny Miller, Valerio Pelucco () and Marc van Essen ()
Additional contact information
Mario Daniele Amore: HEC Paris - Ecole des Hautes Etudes Commerciales
Valentino D’angelo: Bocconi University (Italy, Milan)
Isabelle Le Breton-Miller: HEC Montréal - HEC Montréal
Danny Miller: HEC Montréal - HEC Montréal
Valerio Pelucco: Libera Università Internazionale degli Studi Sociali Guido Carli (Italy, Rome) - LUISS
Marc van Essen: University of South Carolina [Columbia]

Post-Print from HAL

Abstract: Many studies have focused on family firms. Yet, grasping the nature of these organizations remains challenging because firms' familiness can take many forms, which are hard to trace with traditional data. We use ChatGPT – an application of machine learning – to try to unravel the complex and intangible influence of families on firms in large datasets. Whereas it often classifies family firms consistently with equity criteria, ChatGPT appears able to gauge families' legacy and values. Hence, it detects more family firms in countries where families have a strong influence on firms even without large equity stakes. Also, ChatGPT often treats lone-founder firms as non-family firms, whereas it assigns a higher family score to firms that are eponymous, heir-led, and with multiple family directors. Finally, classifying family firms using ChatGPT provides financially relevant information to investors.

Keywords: ChatGPT; Family Firms; Family ownership; Legacy; Management; Values (search for similar items in EconPapers)
Date: 2024-12-01
References: Add references at CitEc
Citations:

Published in Journal of Family Business Strategy, 2024, 15 (4), ⟨10.1016/j.jfbs.2024.100644⟩

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-05656773

DOI: 10.1016/j.jfbs.2024.100644

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().