Machine learning applied to accounting variables yields the risk-return metrics of private company portfolios*
Elias Cavalcante-Filho () and
Rodrigo De Losso Flavio Abdenur
Authors registered in the RePEc Author Service: Rodrigo De-Losso
No 2018_23, Working Papers, Department of Economics from University of São Paulo (FEA-USP)
Constructing optimal Markowitz Mean-Variance portfolios of publicly-traded stock is a straighforward and well-known task. Doing the same for portfolios of privately-owned firms, given the lack of historical price data, is a challenge. We apply machine learning models to historical accounting variable data to estimate risk-return metrics â€“ specifically, expected excess returns, price volatility and (pairwise) price correlation â€“ of private companies, which should allow the construction of Mean-Variance optimized portfolios consisting of private companies. We attain out-of-sample ð ‘…2 s around 45%, while linear regressions yield ð ‘…2 s of only about 10%. This short paper is the result of a real-world consulting project on behalf of Votorantim S.A (â€œVSAâ€ ), a multinational holding company. To the authorsâ€™ best knowledge this is a novel application of machine learning in the finance literature.
Keywords: assent pricing; Machine Learning; Portfolio Theory (search for similar items in EconPapers)
JEL-codes: G12 G17 (search for similar items in EconPapers)
New Economics Papers: this item is included in nep-big, nep-cmp and nep-rmg
References: View references in EconPapers View complete reference list from CitEc
Citations: Track citations by RSS feed
Downloads: (external link)
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
Persistent link: https://EconPapers.repec.org/RePEc:spa:wpaper:2018wpecon23
Ordering information: This working paper can be ordered from
Access Statistics for this paper
More papers in Working Papers, Department of Economics from University of São Paulo (FEA-USP) Contact information at EDIRC.
Bibliographic data for series maintained by Pedro Garcia Duarte ().