Selecting Private Equity Funds Using Machine Learning

Abstract

Prospective investors in Private Equity (PE) funds receive a large amount of non-standardized and qualitative information about fund manager investment strategies. Using a novel and proprietary sample of 380 Private Placement Memoranda, we combine for the first time Natural Language Processing techniques and Machine Learning algorithms to forecast PE fund success based on investment approach descriptions. Our findings suggest that these novel statistical techniques help select PE fund managers. Their increased usage should lead to more efficient private markets.

Authors

Reiner Braun, TUM School of Management
Borja Fernandez Tamayo, Université Côte d’Azur (UCA), SKEMA Business School
Florencio López-de-Silanes, Université Côte d’Azur (UCA), SKEMA Business School
Ludovic Phalippou, University of Oxford, Saïd Business School
Natalia Sigrist, Unigestion, SA