Algorithms and venture investment decisions: better, fairer or biased?

Matthew Levy, Eric Liguori

Research output: Contribution to journalReview articlepeer-review

Abstract

Purpose: This paper is a rejoinder to the work of Blohm, Antretter, and colleagues recently published in both Entrepreneurship Theory and Practice and Harvard Business Review titled “It's a Peoples Game, Isn't It?! A Comparison Between the Investment Returns of Business Angels and Machine Learning Algorithms” and “Do Algorithms Make Better – and Fairer – Investments than Angel Investors?”, respectively. Design/methodology/approach: While we agree with authors of prior scholarship on the importance of counteracting human biases, honing expert intuition and optimizing the odds of success in investment decision-making contexts, in the spirit of open academic discourse, this paper respectfully challenges some of the underlying assumptions concerning algorithmic bias on which prior work is based. Findings: Investing remains part art and part science, and while algorithms may begin to play a more significant role in investment decision-making, human intuition remains hard to imitate. In both people and in algorithms, sources of bias remain both implicit and explicit and often have systemic roots, so more research continues to be needed to fully understand why algorithms produce potentially biased outcomes across a wide array of contexts. Originality/value: This paper contributes to our collective understanding on the use of algorithms in making investment decisions, highlighting the fact that bias exists in humans and algorithms alike, even when the best of intentions are present.

Original languageEnglish (US)
JournalJournal of Small Business and Enterprise Development
DOIs
StateAccepted/In press - 2021

All Science Journal Classification (ASJC) codes

  • Business, Management and Accounting (miscellaneous)
  • Strategy and Management

Fingerprint

Dive into the research topics of 'Algorithms and venture investment decisions: better, fairer or biased?'. Together they form a unique fingerprint.

Cite this