The micro-determinants of portfolio gyrations in mutual funds: evidence from machine learning models

Fabrizio Ferriani () and Sabina Marchetti
Additional contact information
Fabrizio Ferriani: Bank of Italy

No 913, Questioni di Economia e Finanza (Occasional Papers) from Bank of Italy, Economic Research and International Relations Area

Abstract: We investigate the micro-determinants of portfolio gyrations in equity mutual funds that invest in emerging markets. Our analysis focuses on portfolio holding variations driven by asset managers' decisions, rather than by price revaluation, and matches this information with a comprehensive set of 54 stock-level characteristics. Using gradient boosting models (GBMs), we explore the non-linear relationships between stock characteristics and portfolio adjustments. Our findings show that firms' size and investment-related features, alongside equity stock attributes (e.g. market capitalization, traded volume, beta), are the most influential in explaining portfolio turnovers. Additionally, we provide evidence on how the relative importance of these characteristics shifts, based on sample partitions determined by market conditions (downturn vs recovery), investor type (institutional vs retail) and investment strategies (active vs passive).

Keywords: mutual funds; emerging markets; machine learning (search for similar items in EconPapers)
JEL-codes: G01 G11 G14 G23 (search for similar items in EconPapers)
Date: 2025-03
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.bancaditalia.it/pubblicazioni/qef/2025-0913/QEF_913_25.pdf (application/pdf)

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:bdi:opques:qef_913_25

Access Statistics for this paper

More papers in Questioni di Economia e Finanza (Occasional Papers) from Bank of Italy, Economic Research and International Relations Area Contact information at EDIRC.
Bibliographic data for series maintained by ().