Selecting Directors Using Machine Learning

Abstract

Can algorithms assist firms in their decisions on nominating corporate directors? Directors predicted by algorithms to perform poorly indeed do perform poorly compared to a realistic pool of candidates in out-of-sample tests. Predictably bad directors are more likely to be male, accumulate more directorships, and have larger networks than the directors the algorithm would recommend in their place. Companies with weaker governance structures are more likely to nominate them. Our results suggest that machine learning holds promise for understanding the process by which governance structures are chosen and has potential to help real-world firms improve their governance.

Predicting bitcoin price movements using sentiment analysis: a machine learning approach

“…Economic problems involving many variables (Erel et al, 2021); Complex interactions among variables resulting in high dimensionality in data (Easley et al, 2021); and Cases where prediction is more economically important than statistical inference (Li et al, 2020).…”

“…Various studies have been conducted to look at cryptos’ risk and return behavior using innovative techniques such as machine learning (ML) (Gogas and Papadimitriou, 2021, for a good overview). A hot topic remains the use of ML to forecast both the price and direction of financial assets, with the use of ML in financial forecasting benefiting from its ability to capture larger data sets and offers solutions to as follows: Economic problems involving many variables (Erel et al , 2021); Complex interactions among variables resulting in high dimensionality in data (Easley et al , 2021); and Cases where prediction is more economically important than statistical inference (Li et al , 2020). …”

Biased Auctioneers

“…We can therefore expect artworks associated with more aggressive estimates-relative to our automated valuations-to have lower future returns. 4 We find suggestive evidence in support of this hypothesis using data 3 These results contribute to the discussion on the relative strengths and weaknesses of men versus machines in financial-economic decision making (e.g., Abis (2020), Coleman, Merkley, and Pacelli (2020), Erel et al (2021), Fuster et al (2022) as well as the discussion on the implications of machine learning for job occupations (e.g., Autor (2015), Acemoglu and Restrepo (2018), Agrawal, Gans, and Goldfarb (2018), Brynjolfsson, Mitchell, and Rock (2018), Grennan and Michaely (2020)).…”

Social Networks Analysis in Accounting and Finance*

“…Furthermore, directors can facilitate business contacts that can be the source of useful business relationships, including new clients and/or suppliers, new capital or other economic benefits and resource exchange. However, these benefits may come at a cost; for example, board interlocks might propagate value decreasing practices, or well‐connected directors may be too busy to act as good monitors and advisors to the firm (Erel et al 2021). Studies in our review that examine the consequences of director networks document a positive association with firm performance (Larcker et al 2013), strategic advising (Brown et al 2019), financial reporting quality (Intintoli et al 2018), and management forecast accuracy (Ke et al 2020; Schabus 2022).…”