Angels and Venture Capitalists: Substitutes or Complements?

Start-up firms need multiple rounds of funding as they grow from mere ideas into substantial companies. Google and Facebook are well-known examples. Early stage funding may be provided by a variety of parties, especially angel investors (wealthy individuals investing their own money) and venture capitalists (professional investors investing institutional money). In this paper we ask how these two types of investors interact with each other, whether they are part of a single track, or a multi-track dynamic funding ecosystem

The dominant view of how angels and VCs interact is that they are complementary players of a unified ecosystem. Successful ventures start with angel funding and later seek out VC funding as their needs for capital grows larger. According to this dominant view companies start their funding history with angel funding, and the successful ones then “graduate” from angel funding to VC funding.

One may wonder if this dominant view derives from what Silicon Valley looks like, or, indeed, what it used to look like. In this paper we consider the ecosystem of British Columbia (BC), Canada, an emerging ecosystem that has burgeoning angel and venture capital markets. While systematic angel data is usually hard to come by, we benefit from access to administrative data of a BC government program that offers a 30% tax credit to private individuals who invest in eligible start-up companies. We can observe the financial histories with detailed investor compositions of 469 BC-based, primarily high-tech start-up companies that were funded over the period 1995-2009.

We develop and test several alternative hypotheses about the nature of interactions between investors in start-up companies. Specifically, we ask how the prior presence of investor types relates to subsequent investor choices. Our hypotheses recognize the possibility that different types of investors can be complements or substitutes to each other.

We find strong evidence for dynamic persistence within investor types. This means that a company that currently obtains funding from a particular investor-type is likely to raise more funding from investors of the same type. We also find significant negative dynamic effects between angel and VC financing. Companies that currently obtain angel funding are less likely to subsequently obtain VC funding; and vice versa. Together these findings suggest angels and VCs are dynamic substitutes. This finding challenges the dominant view that start-up ecosystems consist of just a single track along which companies progress. Instead, there seem to be separate angel and VC tracks that do not often cross each other often. There are of course migrations from angel to VC funding, and vice versa, yet our results show that they are the exception, not the rule. Importantly, our findings are robust to alternative model specifications, and continue to apply to both successful and unsuccessful companies.

We also ask whether our dynamic substitutes result reflects financing choices that are investor-led or company-led. In case of the former there is a treatment effect: investors affect the future funding choices of companies. In the latter case there is a selection effect: company characteristics determine the evolution of investor choices. We find evidence consistent with the presence of a selection, rather than a treatment effect. This suggests that company characteristics drive dynamic investor choices, rather than the other way around. It appears that some companies are better served by a governance regime involving angels, while others are better suited for venture capital.

Our findings have important policy implications. Many governments all over the world support venture capital financing, but few support angel financing. Our analysis suggests that this approach may be fundamentally unbalanced or incomplete, because it fails to reach entrepreneurial companies that are better served by angels.