Venture capital (VC) firms are notoriously secretive. In a recent paper, we examined the causes and consequences of that preference for secrecy using late-2002 court rulings that forced some of the largest public limited partners (LPs) who invest in VC funds to disclose formerly confidential return information. We show that in response to these rulings, the most successful VC firms dropped the afflicted public LPs and replaced them with private and foreign LPs not subject to the disclosure requirements. The dropped public LPs reallocated their capital to less successful and younger VC firms and engaged in contractual innovation to avoid disclosure. The rest of this post explains the initial shock, the follow-on impact, and how our findings inform discussion about the financing of innovation.
A Shock to the Required Disclosure of Public Investors
In October 2002, the University of Texas Investment Management Company (UTIMCO) responded to pressure from journalists by making its historical fund-level returns available to the public. This event triggered a chain of lawsuits across different states, calling for the disclosure of not just fund-level returns, but also fees and even portfolio company financials.
An important feature of these disclosure lawsuits was uncertainty regarding the scope of information that would need to be disclosed. The lawsuits were typically brought by journalists under state-level open information acts, or Freedom of Information Acts (FOIAs) enacted by each state over the previous decades (Hurdle, 2005). The plaintiffs sought comprehensive disclosure covering not just historical fund returns, but also identities of other investors in the funds and detailed portfolio company information ranging from names and equity stakes to valuations.
We use Preqin data to study how these rulings affected VC-LP relationships. We construct a sample of connections between VC funds and their LPs based on reported capital commitments between 1999 through to 2005. This window covers the dot-com boom and the enactment of state-level amendments limiting the scope of disclosure by public institutions. Our final sample includes both public and nonpublic LPs that made investments in VC partnerships located in the US. We split limited partners into treated public investors, consisting of public pension funds and public university endowments, and non-treated investors, a diverse group ranging from private pension funds and endowments to funds of funds and foreign investors. We rank VC firms based on the number of IPOs among their portfolio companies and label the top twenty VC firms as top-tier.
We use a difference-in-differences framework to test how these funding relationships responded to the court rulings. Specifically, we focus on how the probability to invest with top-tier VC firms evolved for the treated public investors vis-à-vis the control group of non-treated LPs. Our analysis shows that public LPs saw a large decrease in allocations to top-tier VC funds, despite little change in overall investment in VC. We also confirm our results using foreign LPs, another group that is not affected by FOIAs enacted in the United States. The recurring feature of capital commitments to venture capital funds allows us to exploit limited partner fixed effects, that is, any unobservable time-invariant characteristics of institutional investors cannot drive the results. Using this specification, we find that public LPs are almost 10% less likely to invest in top-tier VCs relative to the control group after the disclosure shock.
Our empirical design also allows us to alleviate concerns regarding the state-specific time-variant factors by the inclusion of the granular LP state × time fixed effects. While FOIAs are relatively uniform across states, they did have pre-existing differences that could lead to potentially different interpretations of chapters governing the disclosure of private information. In addition, the timing and the scope of relevant amendments to FOIAs varied from one state to another. We show that the results are not driven by these types of time-variant state-level characteristics.
The empirical evidence and anecdotal accounts indicate that top-tier VC firms responded differently to the disclosure requirements. Many top VC firms, such as Sequoia Capital, openly evicted their public LPs and publicly protested any form of disclosure. Other VC firms, often younger and less successful ones, took advantage of the situation. One such example is MPM Capital LP which closed its largest-ever fund in 2002.
The greater responsiveness of top-tier VC firms to this shock points to the importance of capital supply being a driver behind these results. Public LPs often account for the largest stakes in VC funds and act as the primary suppliers of capital; 29% of the total number of commitments in our sample are from public investors. Because of that, many VC firms would find it hard to exclude public LPs from the pool of their capital providers. As the most successful VC firms have already scaled close to their maximum size, they have less dependency on public LPs. They can substitute away from this class of investors as they typically have oversubscribed funds with many potential LPs wishing to participate in any new fund. Therefore, prominent VC firms are less reliant on any particular LPs and consequently more likely to forgo commitments from public LPs during the period of uncertainty surrounding the movement for increased disclosure.
Long-Term Impact on Contractual Terms
The fraction of public LPs’ commitments to top-tier VC firms to some extent recovers in two years after the initial lawsuits. We argue there are three reasons for this. First, legal uncertainty regarding the scope of required disclosure was substantially resolved. The initial FOIA lawsuits sought a comprehensive disclosure of information, including VC portfolio company financials. This possibility was gradually ruled out as courts consistently set a precedent that limited the scope of the disclosure to information on past returns and clarified that the disclosure will not include the sensitive portfolio company-level data.
Second, the opposition from prominent VC firms such as Sequoia Capital, as well as the disquiet of the excluded public LP managers, led states to adjust their FOIAs to exclude these data, most importantly the information regarding the portfolio companies. For example, the state of Michigan amended its FOIA act in 2004 allowing the University of Michigan to keep its records private. This development also led some other states to reconsider their attitude toward the scope of the state-level open information acts.
Third, VCs and public LPs coordinated to create a limited reporting approach, where VC firms report only selective information to the limited partners upon mutual consent stated in the limited partnership agreement. While not every LP has accepted this approach, it has allowed others to maintain or regain their access to top VC funds. For example, years later, the University of California was accepted to Sequoia Capital’s new fund raised in 2010, though the University subsequently reported only the amount of capital commitment to the fund. According to the 2018-2019 study of partnership agreements by “Buyouts Insider,” approximately 50% of limited partnership agreements contain a clause allowing general partners (GPs) to withhold certain information from their public LPs.
We corroborate the last channel using Preqin’s return data. By exploring the availability of the return data, we establish that the performance information of top-tier VC firms is more likely to be missing in Preqin after the lawsuits. These results confirm that the most successful VC firms avoided disclosure of their data by altering LPAs and reshuffling investors. This bias in the return information in the post-treatment period constrains our ability to reliably measure if public LPs incurred economic losses following the court-mandated disclosure. At the same time, pre-2002 returns do not exhibit bias and are significantly higher for top-tier VC firms. This observation mitigates the possibility that the results are driven by VCs’ unwillingness to disclose poor dotcom bubble returns that could be associated with a lack of skill and hurt their future fundraising prospects. While our ability to quantify the economic consequences of the disclosure on public LPs is limited, future academic research might be able to address this question, as return data become more widely available to scholars.
Our results showcase the unique characteristics of the VC-LP relationships. One of the defining features of these relationships is the large amount of bargaining power possessed by top VC firms. We link that back to a key feature of the VC model: an inability to scale. VC firms cannot double their investment rate by simply doubling the size of their investments, like mutual funds could, as the companies they invest in might not be able to absorb the financing. Similarly, VC firms cannot simply target companies that are twice as large, like buyout firms could, as that would entail a fundamental change in the nature of their business. This size limit means that, even after outstanding performance, most VC firms see relatively small increases in assets under management, as documented by Metrick and Yasuda (2010).
This inability to scale has given successful VC firms remarkable bargaining power (Mulcahy et al., 2012). While even the best hedge funds and private equity firms typically charge 2% management fees and 20% carried interest, the best VC funds can charge significantly larger amounts (Kaplan, 1999). However, this higher level of compensation does not completely capture the value generated by high-quality VC firms, with several papers arguing that top-tier funds consistently outperform other VCs. This structure has led to a rare power dynamic where, in the same way as an IPO underwriter, VCs pick and choose their limited partners – and they are notoriously picky. We show how this power dynamic manifests in a preference for secrecy, an unexplored but important dimension of this relationship. Our results demonstrate that matching between LPs and VC firms is of two-sided nature and the regulatory environment surrounding institutional investors is a relevant dimension influencing the choice of LPs by VC partnerships.
Rustam Abuzov is an Assistant Professor of Business Administration at the Darden School of Business of the University of Virginia.
Will Gornall is an Assistant Professor of Finance at the UBC Sauder School of Business.
Ilya A. Strebulaev is the David S. Lobel Professor of Private Equity at the Stanford Graduate School of Business.
This post was adapted from their paper, “The Value of Privacy and the Choice of Limited Partners by Venture Capitalists,” available on SSRN.