Note: This is an edited excerpt from a research proposal I wrote for a course on network theory. For the purposes of this post, I did not include any of the sections on the data I would gather or the statistical analysis I would perform. If you have questions or suggestions about data or analysis, please feel free to contact me via the contact form on my site. Alternately, if you want to collaborate on executing this study, please also reach out through the contact form.
I consider this proposal a “working draft” of the first section of a longer paper. In order to complete that longer paper, I need data. So, if you have a huge database of VC deals, please let me know.
The objective of this paper was to propose a method for identifying the most influential venture capital firms (by using betweenness centrality as a proxy for influence) and determining whether there existed a correlation between the “promiscuity” of lead VCs and those VCs which were most effective at increasing valuations of the underlying ventures between the A and B rounds, rather than focusing on IPO valuations like other research has done.
I did include the works cited and my annotations to said citations at the bottom of this post.
In 2007, Hochberg et al. published a paper titled “Whom You Know Matters: Venture Capital Performance and Investment Networks” in which they establish that more well-connected VCs perform better at the individual company and portfolio level over the course of the venture lifecycle. At the earliest stages of the venture lifecycle, however, entrepreneurs understand the importance of partnering with a high-quality venture capitalist. Just as picking the right business partners can set up a project for success, picking the right investor should therefore augment performance.
However, it’s not enough to simply raise the first round from just any investors; a new venture must raise from the “right” ones. The “best” early-stage investors are capable of “adding value” to an upstart company. This value can come in the form of connections, industry expertise, and other factors besides the ability to write a big check. For the entrepreneur, when it comes to deciding which firm (or more rarely, an individual) to partner with as the lead investor on the startup’s first round of outside capital investment, preference skews heavily in favor of those investors were capable of adding the most value to the firm besides the capital. When the most important aspects of this value is the ability to connect the entrepreneur and her company with future investors that also carry domain expertise and industry experience.
The idea for this research project came from a casual glance at a database of “unicorn” companies (companies with $1 billion plus valuations founded after January, 2003) in which we saw commonalities in the funds which originate and lead those companies first rounds. In the blog post which introduced the concept of unicorn companies, Cowboy Ventures’s Aileen Lee explains that VCs care about “billion-dollar exits” because most venture capital funds’ returns depend on the runaway growth of just a few portfolio companies. Owning twenty percent of just two different billion-dollar companies returns the initial capital of a $400 million fund.
The stakes are high for venture capital firms, and that they have an incentive to maximize the valuations of their investments from round to round as quickly as possible. The NVCA says that over 16,000 internet-related companies were funded since 2003; Mattermark says there were 12,291 software and internet companies funded between January, 2011 and January, 2013. By Lee’s estimate, some 60,000 software and internet companies were funded in the past decade (Jan. 2003 through Jan. 2013). If that is true, then the 39 unicorns on her list in November 2013 represent just 0.07%, or roughly 1 in 1,538 funded internet and software startups. Since Lee’s article was published in TechCrunch in November, 2013, about forty other companies joined the “unicorn club”, but the previous year and a half has been prodigious for the technology investment business. However much the odds have changed in favor of extreme success is infinitesimal.
Because it is virtually impossible to predict which companies will become billion-dollar success stories, we assume that VCs distribute their value-adding efforts approximately equally across their portfolio companies, especially at the riskiest early stages of the startup lifecycle. However, as we observed, a disproportionate share of the most valuable tech companies were originally funded by just a handful of venture capital firms.
It might be possible to predict success at a much more quotidian scale. Why are some VCs so much more likely than others to give rise to runaway successes? Or, to reframe the question in a manner which can result in a more generalized set of answers: What makes these firms so special? All else being equal, what do they possess that other firms do not?
In this research proposal we hope to further develop the concept of power as it relates to the contemporary venture capital industry. First we will explore a few old models of understanding power in the venture capital industry, then we will introduce a pair of network-theoretic hypotheses informed by work on structural holes and several papers about deal syndication patterns. Finally, we propose a strategy for acquiring, processing and parsing the data to confirm or disprove our hypotheses.
The Old Model: A Virtuous Cycle of Value Creation?
Megginson and Weiss (1991) argue that the power of a venture capital firm comes from its reputation. Rather than implement a more contemporary understanding of reputation as connective power, Megginson and Weiss claim that the reputation of a venture capital firm is wielded most effectively when portfolio companies are transitioning from private to public markets. (Megginson and Weiss, 882) To wit, they claim that VCs act as certifying bodies, and brokers of information between the executives of the underlying company and corporate underwriters in an initial public offering. This certifying role of the venture capitalist, by their logic, is ideal for situations of great information asymmetry between executives in the underlying company and external markets, and in situations of great uncertainty, which is exactly where “young, rapidly growing, research and development-intensive companies” (e.g. startups) find themselves.
In their analysis Megginson and Weiss find that venture capital fund reputation is a function of three factors, two of which are relevant here. First, they assert that the reputation and effectiveness of VCs is measured by the degree to which they reduce incidence of underpricing in initial public offerings. (Megginson and Weiss, 893) In other words, the amount of money “left on the table” is reduced by effective VCs, allowing underwriters to capture as much value as possible. Thus, underwriters who have worked with VCs rationally opt to continue working with them, and VCs gain reputation.
The second finding relevant to us is that the shares a venture capital firm holds in a given venture post-IPO acts as a “bonding mechanism” between the VC, the now-public company, and the underwriters. Instead of harvesting gains in their entirety, most VCs sell only a small portion of their holdings, usually giving up voting control (>51% stake) in the company. Many sell no shares at the IPO. The large post-offering holdings, according to Megginson and Weiss, can be used as an additional sign of credibility at the time of offer since VCs forego the opportunity to profit directly from maliciously misleading underwriters. (Megginson and Weiss, 900) This allows VCs to further develop trust with underwriters, and can lead to increased capital investment from future limited partners.
To summarize Megginson and Weiss’s findings somewhat crudely, we can say that venture capital will continue to make a lot of money because it has made a lot of money (for underwriters and its limited partners) in the past. Reputation pays dividends in the form of more money from limited partners on future funds and more lucrative deals with startups during the initial deal making process. (Hsu 2004)
In another study of reputation and the price entrepreneurs pay for it, David Hsu (2004) finds that entrepreneurs entertaining competing offers from many venture capital firms will not only opt for the offer from the firm which promises to deliver the most additional value (by way of their “reputation”), but they will also forego offers at higher valuations (thus giving away more equity) in exchange for affiliation for a “reputable” investor. (Hsu, 1809) In other words, the most reputable VCs in the business act as the arbiter of valuation when it comes time to go public, benefit from the shares they hold in now-public companies, and they also get to originate those initial deals relatively cheaply.
Approaching a Networked Understanding of Venture Capital
Does access to better deal flow simply beget better deal flow, which leads to better returns, and thus access to better deal flow? Is success in the venture business a closed loop, or, is it possible for a venture capital firm to “hack” their way to the top without putting in decades of work into building a track record of good return on investment for LPs by waiting for early investments to go public or be acquired? Conversely, can startup companies set themselves up for higher performance by working with their lead venture capitalist to secure more syndicate partners in order to access information, follow-on investors in future rounds and market opportunities?
Venture capital is a business built on relationships, even at the origination stage of the deal-making process. At the origination stage, VCs rely on their personal and professional networks to source deals for origination. (Fried and Hisrich 1994, 32) And once a deal is finalized and the company joins the venture capital portfolio, venture capital firms tend to act as brokers between portfolio companies and future capital partners and market gatekeepers within the venture capital fund’s network. (Hsu 2004, 1809)
Although a long explanation of the various cultural mores of technology startups would beyond the scope of any final research, and far beyond the scope of this proposal, suffice it to say that early stage startups exhibit many of the characteristics Boltanski and Chiapello (2005) discuss in the context of projects. Startups, especially those which haven’t received funding yet, are more like loose associations of individuals, each with different skills they bring to bear on the task at hand: building something people want.
If a startup company is a project for its cofounders and early employees, it is just as much a project for investors to collaborate on. Although instead of building a product or service, investors seek to augment the efforts of entrepreneurs. They bring their own resources to bear on the task of increasing the valuation of the companies they work with as fast and as efficiently as possible. Here too we can think of venture capital firms in networked terms. In fact, many firms use their networked characteristics as a selling point. (Hsu, 1810)
Whereas people join startup projects by collaborating, and entrepreneurs recruit new collaborators based on the complementary skill sets new hires bring to the effort, lead investors bring other VCs in to participate on the round through a process called “syndication”. The conventional wisdom about syndication is that it is a way for VCs to hedge against risk. (Megginson and Weiss, 899) However, we assert that in a networked understanding of venture capital deal-making, opening up a deal to quality syndicate partners does more than merely spread risk around. A quality team of investors better position the startup for success.
We already find some evidence in previous work that more experienced VCs tend to have more syndicate partners. (However, interestingly, Hochberg et al find that a networked understanding of VC performance negates many of the beneficial effects of experience on performance.) (Hochberg et al, 254) Hopp (2010) finds that VCs with more experience tend to rely more on syndication, and that the longer the VC has been performing transactions, they are more likely to bring other investors into the deal. (Hopp, 418) Additionally, Hopp makes the point that bringing new syndicate partners into future fundraising rounds helps the VC escape “competency traps”. (Hopp, 420) Although Hopp doesn’t say so explicitly, these new investors fill structural holes in the network, giving VCs and their portfolio companies access to more information and connections.
We also find strong evidence that investors which cast a wide net in their search for syndicate partners add the most value over the life of the firm. Hochberg et al’s (2007) paper establishes that large syndicate networks add more value to upstart companies by measuring valuations at acquisition or IPO. They find that VC firms that occupy more central, or influential, positions in the VC network enjoy better investment performance, both at the fund and portfolio company levels, from origination through exit. (Hochberg et al, 293) Although their paper was among the first to do a serious network analysis of venture capital deal networks, and their findings are significant and interesting for late-stage companies, we are interested in exploring the impact of lead investors’ ability to syndicate deals on the growth of early stage companies.
In the following sections, we will establish a set of hypotheses and a protocol for testing them.
Hypothesis one: Round leaders which are more effective in recruiting syndicate partners for the first round of outside money (Series A) add more value at the earliest stages. So, more partners per deal, all things being equal, result in a greater change in valuation from Series A to Series B.
Hypothesis two: VCs which have the most diverse set of syndicate partners across their portfolio of Series A deals on which they are lead investors add more value than firms with less diversity. This is to say, all things being equal, investment firms that are promiscuous with their syndicate partnerships will add more value in their comparatively less promiscuous counterparts.
For this proposed project, we will analyze the degree to which a given lead investor can augment the growth of a portfolio company based on the syndicate partners they bring on for that specific round, as well as the overall number of syndicate partners with whom the investor has engaged across their entire investing history.
We will observe a given firm’s level of betweenness centrality in the network. This differs from Hochberg et al’s (2007) approach of using degree centrality. Whereas their model uses directed ties between one type of node (VC firms) to establish degree centrality measures, our approach uses startups as an additional node type, and as the medium through which venture firms inter-relate.
[The remainder of the paper involves a lengthy and tedious discussion of data collection and analysis. I removed it for the purposes of publishing this post.]
Boltanski, Luc, Eve Chiapello, and Gregory Elliott. The New Spirit of Capitalism. London: Verso, 2005.
The chapters in Boltanski and Chiapello’s book provided a interesting framework for understanding the nature of projects as focal points for networks.
Fried, Vance H., and Robert D. Hisrich. “Toward a Model of Venture Capital Investment Decision Making.” Financial Management 23.3 (1994): 28-37. JSTOR. Wiley on Behalf of the Financial Management Association International. Web. 23 May 2015. <http://www.jstor.org/stable/3665619>
This article provided a great deal of information about the venture capital dealmaking process, and was used in this paper primarily for the narrative relayed about deal origination, and the relational nature thereof.
Gompers, Paul A., Josh Lerner, Margaret M. Blair, and Thomas Hellmann. “What Drives Venture Capital Fundraising?” Brookings Papers on Economic Activity. Microeconomics 1998 (1998): 149-204. JSTOR. Brookings Institution Press. Web. 24 May 2015. <http://www.jstor.org/stable/2534802>.
This article primarily concerned be economics, and other factors involved in the raising of a venture capital fund from limited partners. Although we did not cite this piece in our proposal, this paper is cited by many of the sources we did cite, and we reviewed this paper during the initial research for proposal.
Hochberg, Yael V., Alexander Ljungqvist, and Yang Lu. “Whom You Know Matters: Venture Capital Networks and Investment Performance.” The Journal of Finance 62.1 (2007): 251-301. JSTOR. Wiley for the American Finance Association. Web. 25 May 2015. <http://www.jstor.org/stable/4123462>.
Hochberg et al’s article is probably the most relevant, and important article we reviewed while writing this proposal. The paper established a network understanding of venture capital performance, and provided a quantitative methodology from which we borrowed.
Hopp, Christian. “When Do Venture Capitalists Collaborate? Evidence on the Driving Forces of Venture Capital Syndication.” Small Business Economics 35.4 (2010): 417-31. JSTOR. Springer. Web. 24 May 2015. <http://www.jstor.org/stable/40927528>.
This paper contains an excellent research into the factors which determine the likelihood of collaboration/syndication in venture capital deals. Much of what was said in this paper aligns with assumptions I have made going into writing this research proposal, and in it Hopp uses some very interesting methods for determining factors like round leadership, etc.
Hsu, David. “What Do Entrepreneurs Pay for Venture Capital Affiliation?”The Journal of Finance 59.4 (2004): 1805-844. JSTOR. Wiley for the American Finance Association. Web. 24 May 2015. <http://www.jstor.org/stable/3694879>
Hsu’s article deals with issues of venture capital reputation, and the price entrepreneurs pay to be affiliated with highly reputable venture capitalists. However, the article does not go so far as to establish the networked understanding of venture capital reputation, and makes use of Megginson and Weiss’s conception of the venture capitalist as a certification body and broker between late stage ventures and IPO underwriters.
Megginson, William L., and Kathleen A. Weiss. “Venture Capitalist Certification in Initial Public Offerings.” The Journal of Finance 46.3 (1991): 879-903. JSTOR. Wiley for the American Finance Association. Web. 29 May 2015. <http://www.jstor.org/stable/2328547>.
This article details what I argue isn’t old understanding of venture capital reputation. They basically frame the venture capitalist is an information broker between late stage ventures and IPO underwriters. They do not discuss the subject of syndication at all, and in so far as they discuss collaboration, it is only between the venture and the venture capitalist, in the venture capitalist and underwriters. Still though, it is an important paper which is given rise to other papers about the certification role of venture capitalists.