Economics of Entrepreneurship,
Equity Crowdfunding and Venture Capital
This research stream focuses on how the networks enabled by crowdfunding platforms, such as Kickstarter and AngelList, shape the selection and funding of early-stage entrepreneurial projects. I also study the selection performed offline by angels and venture capitalists, and how it can be integrated with digital platforms to expand the types of startups that are funded.
In 2017, this body of work was recognized with the Kauffman Junior Faculty Fellowship in Entrepreneurship Research (KJFF).
RESEARCH ON CROWDFUNDING
In the first phase of the evolution of crowdfunding platforms, experts were not involved in the selection of entrepreneurial projects. Started in the creative industries, crowdfunding was a response to the challenging economics of digital content in an environment where piracy was rampant. Artists turned to the web to sell the only type of music that could not be stolen: music that had not been produced yet. Building on the simple game theory of the provision point mechanism - whereby funds would only be allocated if a sufficient number of investors committed to support a project - crowdfunding offered a new source of early-stage capital. By shifting market power from traditional intermediaries to the crowd, it also promised to expand access to funding to ideas and entrepreneurs turned down by experts. Except, crowdfunding carried its own idiosyncratic set of trade-offs and challenges...
Crowdfunding: Geography, Social Networks, and the Timing of Investment Decisions
"The online mechanisms do not (yet) eliminate frictions related to preexisting social networks"
"Crowdfunding: Geography, Social Networks and the Timing of Investment Decisions" - JEMS, 2015
With Ajay Agrawal and Avi Goldfarb
Journal of Economics and Management Strategy
Vol. 24, Number 2, pp. 253-274.
In the paper, we explore the frictions crowdfunding removes in the matching process between entrepreneurs and investors, and uncover how the social connections between the entrepreneurs and their family and friends end up influencing overall fundraising dynamics on crowdfunding platforms. While crowdfunding removes some of the distance-sensitive costs associated with evaluating and funding projects, it does not remove all of them. In particular, information that travels through offline networks leads socially connected funders to form their consideration set in a different way. By contributing early, this group generates the signal of quality (accumulated capital) everyone else relies on to make their investment decisions. Therefore, while crowdfunding platforms may appear global in nature, because of search frictions, asymmetric information and path dependency they are heavily constrained by pre-existing social ties. This substantially limits their ability to democratize access to capital. The paper was awarded a NET Institute Grant.
Can Capital Defy the Law of Gravity?
Investor Networks and Startup Investment
With Xiang Hui
Early crowdfunding platforms were based on a premise of complete disintermediation from experts: The crowd would directly fund projects based on the information shared online by the entrepreneurs, bypassing in the process traditional gatekeepers. This approach becomes problematic when equity is involved, since the degree of asymmetric information and the risk of moral hazard are higher than in reward-based crowdfunding. Platforms have therefore experimented with market design solutions targeted at counterbalancing these risks.
"The introduction of syndication
increased capital flows to non-hub regions by 25%"
"Can Capital Defy the Law of Gravity? Investor Networks and Startup Investment" - WP, 2017
In the paper, we study how online syndication by professional investors changed the allocation of capital on the leading US equity crowdfunding platform. Using novel data on investments and startup valuations, we find that the introduction of intermediaries increased capital flows to non-hub regions by 25%, a result that relies on syndicate leads having pre-existing professional ties in these areas. The early-stage deals closed through an intermediary in these new regions, moreover, tend to be associated with better returns, suggesting that experts can play a key role in arbitraging opportunities and expanding access to capital across regions.
SLACK TIME AND INNOVATION
With Ajay Agrawal, Avi Goldfarb and Hong Luo
The question of how crowdfunding changes the types of entrepreneurs and ideas that are funded is also the focus of "Slack Time and Innovation". Using large-scale data on Kickstarter projects, their founding teams and skills, we investigate how this new source of capital allows individuals who are not full-time innovators to de-risk entrepreneurial ideas. We focus on a demographic - college students - that often has the right human capital but not the access to capital necessary to commercialize ideas. This gives us regional variation in terms of when students are more versus less likely to have time for experimentation (college breaks). The paper starts from a key insight from the literature - the fact that slack can be both a source of breakthroughs, but also of resource misallocation - to develop and test a simple theoretical model for the role low-opportunity cost time plays in innovation. Using a difference-in-differences approach, we confirm the main predictions of the model: Slack time induces more projects on both tails of the outcome distribution, with the right tail being driven by complex projects developed by teams with diverse skills. Our results highlight that a sufficient amount of contiguous time and "overlapping slack" are critical for facilitating team coordination and implementing high-potential projects. Stricter screening mechanisms are also effective in conjunction with slack to reduce the number of marginal projects being developed.
SOME SIMPLE ECONOMICS OF CROWDFUNDING
IPE, University of Chicago Press (2014)
Vol. 14.1: 63-97
With Ajay Agrawal and Avi Goldfarb
In the paper, we discuss how crowdfunding, while sometimes an effective substitute for traditional sources of capital, also introduces new distortions. The lower cost of capital, ability to cheaply assess market demand, receive feedback, and build a community of supporters, all need to be weighted against increased expropriation risk due to early disclosure, costs of managing the crowd, and absence of incentives for mentorship from professional investors. Similarly, while crowdfunding opens new investment opportunities for funders, it also exposes them to inexperienced entrepreneurs and to an environment with low incentives for due diligence. In the text we also review market design mechanisms that can mitigate these issues.
PREDICTION AND MACHINE LEARNING
Hidden in Plain Sight:
Venture Growth With or Without Venture Capital
With Jorge Guzman and Scott Stern. MIT Working paper, 2017
While the study of high-growth firms has focused predominantly on VC-backed startups, the majority of IPOs and acquisitions are achieved without venture capital. This motivated us to use population-level data, combined with a predictive analytics approach, to shed light on these `missing' growth firms and on the process of venture capital selection. We find that the observables that are associated at birth with selection into VC are also predictive of growth within the non-VC sample, suggesting that firms with growth potential - irrespective of funding path - are similar to each other. The approach also allows us to estimate an upper bound on the returns to VC: While a naive comparison would lead to a drastic over-estimate, our results show that VC-funded firms are 2.4 to 6 times more likely to achieve an exit. Overall, our findings suggest the presence of alternative paths to growth (with or without venture capital), but a single profile for high potential firms.
Soft Information versus Bias in New Venture Finance:
Machine Intelligence
versus Human Judgment
With Chris Foster and Ramana Nanda. Work in progress, 2017
In the paper, we explore the types of errors domain experts make in screening startups. When outcomes are extremely skewed even skilled evaluators may struggle. Score aggregation - often the default within angel groups, accelerators and VC firms - only exacerbates the issue by cancelling useful variance. Over time, pattern matching may also lead experts to favor idiosyncratic types of ventures. Our analysis of a large set of thousands of applications to a top global accelerator shows little correlation between expert scores and startup outcomes. When we then train a machine learning model to replicate human decision making, the algorithm achieves similar performance to the humans by heavily weighting keywords related to the founding team. Once unconstrained from imitating human judgment, the algorithm selects a broader set of dimensions, and delivers dramatically higher returns. Our results highlight that while the team might be important for venture success, early-stage investors may be overemphasizing it. Moreover, the higher performance achieved by the machine learning algorithm hints at the existence of additional traits of high potential startups that are currently understudied.
POLICY PAPERS
ARE SYNDICATES THE KILLER APP
OF EQUITY CROWDFUNDING?
California Management Review
Volume 58, Issue 2, Winter, 2016.
With Ajay Agrawal and Avi Goldfarb
Can Equity Crowdfunding Democratize Access to Capital and Investment Opportunities?
MIT Innovation Initiative
Policy Report, 2016
With Catherine Fazio and Fiona Murray