A critical dive into Diag Davenport’s 2022 Paper ‘Predictably Bad Investments: Evidence from Venture Capitalists'
This is an exploration of s a 2022 study by Dr. Davenport from Princeton on the topic of "Predictably Bad Investments" in the venture capital world, highlighting the potential of machine learning in predicting investment outcomes. Dr. Davenport asserts that up to half of VC investments are predictably bad, drawing attention to the gap between current investment practices and the possibilities offered by advanced analytics. The discussion covers the methodology, data analysis, and key findings from the study, including the significant predictors of startup success or failure, such as the educational background of founders and their prior investment history. Additionally, it addresses the biases in decision-making processes within the VC industry and contrasts the investment strategies of top-performing firms with those performing poorly. The video also delves into the limitations of current predictive models and the role of machine learning in refining investment decisions, suggesting a shift towards more data-driven approaches in the VC space.
References
P A Gompers , W Gornall , S N Kaplan , I A Strebulaev, How do venture capitalists make decisions, Journal of Financial Economics , volume 135 , issue 1 , p. 169 - 190 Posted: 2020
Kleinberg, J., H. Lakkaraju, J. Leskovec, J. Ludwig, and S. Mullainathan (2018a). Human decisions and machine predictions. The Quarterly Journal of Economics 133 (1), 237–293.
Lyonnet, V. and L. H. Stern (2022). Venture capital (mis) allocation in the age of ai. Fisher College of Business Working Paper (2022-03)
Listen to all episodes on audio:
#venturecapital #startup #innovation
Share this post