«John R. Graham, Co-Supervisor David T. Robinson, Co-Supervisor Manuel Adelino Alon P. Brav Manju Puri Aaron K. Chatterji Dissertation submitted in ...»
6.2 Innovation Improvements and CVC Termination What determines the termination and hibernation of CVCs? To echo Table 4.2, which shows that innovation deterioration motivates CVC initiations, I conclude my analysis of the CVC life cycle by examining corporate innovation at termination.
Table 6.2 Panel A performs simple statistical tests that compare innovation levels at the initiation and termination of the CVC life cycle.
The analysis is performed on all CVCs that can be assigned a termination date (upper panel) and on the subgroup that stayed in business for at least ﬁve years. When examining the industry-year adjusted innovation measures, we observe statistically signiﬁcant improvements at the CVC termination point compared to the initiation stage.
I exploit a hazard model to statistically relate innovation improvements and the decision to terminate a CVC. A CVC parent ﬁrm enters the sample in the year of CVC initiation. The key variable of interest is ∆Innovation, which measures the diﬀerence between innovation level in year t and that of the initiation year. The result is shown in columns (1) and (2) of Table 6.2 Panel B. The positive and signiﬁcant coeﬃcients mean that larger improvements of innovation from the initiation year motivate parent ﬁrms to terminate CVC investment.
To capture how innovation improvements aﬀect the decision to put CVC into 66 Table 6.2: This table studies the decision to terminate Corporate Venture Capital.
Panel A examines average innovation improvement through the CVC life cycle by comparing innovation performance at CVC initiation and CVC termination (deﬁnition as in Table 6.1). Innovation performance is measured using innovation quantity and quality, and both are adjusted using the industry (3-digit SIC level) peers in the same year. I also report the t-statistics for the diﬀerences in means between the two time points. *, **, *** denote statistical signiﬁcance at the 10%, 5%, and 1% levels, respectively. The analysis is performed on all CVCs with a disclosed or deﬁned termination date and the subgroup that lasts for at least ﬁve years.
Panel B studies the eﬀect of innovation improvement on CVC termination and investment decisions. The regressions are performed on the panel of CVC sample in their active years. The key variable ∆Innovationi,t is deﬁned as the diﬀerence of innovation between year t and the year of initiation. In columns (1) and (2), the dependent variable is a CVC termination dummy, and the speciﬁcation is estimated using a Hazard model. In columns (3) and (4), the dependent variable is the annual number of investments in portfolio companies, and the model is estimated using Ordinary Least Squares (OLS). Firm-level control variables include ROA, size (logarithm of total assets), leverage, and R&D ratio (R&D expenditures scaled by total assets). *, **, *** denote statistical signiﬁcance at the 10%, 5%, and 1% levels, respectively.
Panel A: Innovation at CVC Initiation and Termination
67 hibernation, I investigate the intensive margin of CVC investment—the number of portfolio companies a CVC invests in each year and the key variables of interests, ∆Innovation, are deﬁned as above. Columns (3) and (4) present the results, and the ﬁndings are consistent with columns (1) and (2)—innovation improvements are associated with a lower level of CVC activities.
Overall, Table 6.2 matches the ﬁnding at the initiation stage, and is consistent with the information acquisition hypothesis, which predicts that when ﬁrms regain their upward trajectory in corporate innovation, the marginal informational beneﬁt of CVC shrinks, which in turn leads to the termination or hibernation of the CVC unit.
68 7 Conclusion and Literature Revisited
How do corporations ﬁnance and manage their innovation process in the pursuit of long-term growth? This paper sheds light on this fundamental question by studying an emerging economic phenomenon, Corporate Venture Capital (CVC). Armed with an identiﬁcation strategy that allows me to isolate ﬁrm-speciﬁc innovation shocks, I ﬁnd that ﬁrms launch CVC programs following a deterioration in innovation, and their main motivation is to acquire information and innovation knowledge from the entrepreneurial sector. This information acquisition rationale leads me to characterize the life-cycle dynamics of CVC—evolving through the initiation, operation, and termination stages—in which CVC parent ﬁrms strategically select information sources (portfolio companies), actively integrate newly acquired information into corporate decisions, and terminate CVCs when the informational beneﬁt diminishes.
Beyond establishing the CVC life cycle and the information acquisition rationale behind these activities, the paper is a stepping stone toward understanding several broad economic questions.
Organizing Innovation. This paper joins the endeavor to understand the architecture of innovation and contributes to this literature by suggesting three areas 69 for future work. First, more research should be done to achieve a better understanding of details in CVC operations. Second, the information acquisition motive behind organizing innovation, which I highlight in this paper, has been largely overlooked in the literature (Tirole, 2010) but is worth further exploration. Third, the interaction between CVC investment and alternative organizational forms considered here calls for future studies that could consider the system of organizing innovation as a whole by seriously incorporating the interactions among diﬀerent organizational structures and a dynamic intertemporal scope.
Information Economics. Information is important in all areas of ﬁnance, yet information choices have been hard to study both in asset pricing and in corporate ﬁnance, either theoretically or empirically (Van Nieuwerburgh and Veldkamp, 2010).
Empirical work on corporate decisions regarding information management is especially limited by the unobservability of related behaviors. By examining the CVC life cycle, this paper obtains several results regarding information acquisition and use that are hard to show under alternative settings. Future work could explore the CVC setting to answer more questions at the intersection of information economics and corporate ﬁnance.
Creative Destruction. In broader terms, this paper provides new evidence concerning the co-movement of entrepreneurship, creative destruction, and economic growth. Entrepreneurial companies and incumbent ﬁrms diﬀer in their ability to develop radical and disruptive innovation and to capture new investment opportunities (Hall, 1993; Henderson, 1993; Jensen, 1993; Adelino, Ma, and Robinson, 2016;
Acemoglu and Cao, 2015), and this diﬀerence generates the creative destruction momentum. By highlighting CVC as an eﬀective incumbent-entrepreneur bridge, this paper essentially suggests that the two seemingly disentangled sectors could be closely intertwined, which in turn aﬀects both micro-level corporate behaviors and the aggregate process of creative destruction.
Background Reading on Corporate Venture Capital A.1 What Is (not) Corporate Venture Capital?
Corporate Venture Capital (CVC) is the investment of corporate funds directly in external startup companies (Chesbrough, 2002), usually in the form of minority equity investment (Gompers and Lerner, 2000) made through organized and systematic eﬀort (Ivanov and Xie, 2010). In practice, a CVC parent initiates CVC business by launching a unit (formally or informally), which manages CVC investment for the ﬁrm.
In addition to the deﬁnition above, it is helpful to describe CVC by explaining what it is not. When a corporation funds an external fund managed by a third party and eﬀectively works as a limited partner, it is usually not considered as a CVC.
CVC is also distinctive from strategic alliances or joint ventures in which two or multiple parties reach an agreement to pursue a set of agreed-upon objectives; in contrast, CVC is an investment-based relationship without speciﬁc objectives, and the investment amount is signiﬁcantly smaller. In essence, it is best to think of CVC as a subset of Venture Capital whereby an established ﬁrm is systematically investing, 71 without using a third-party investment ﬁrm, in an external, early-stage startup that it does not own or control.
CVC is typically organized as a normal corporate division rather than a traditional VC fund. First, capital sourced to a CVC is almost 100% from its corporate parent (either the headquarters or related business units); more than 60% do not have a ﬁxed amount and fund is allocated on deal-by-deal basis (like capital budgeting).
Second, most CVCs have a corporate mission, and more than two-thirds of them report directly to the corporate headquarter. Third, CVC executives are typically compensated through traditional corporate schemes such as salary or bonus, and rarely via carried interests or schemes similar to carried interests.
A.2 Economic Motivation and Operation: Survey Evidence
In Macmillan et al. (2008), a survey jointly conducted by the National Venture Capital Association, the Wharton School, and the MIT Sloan School, CVC practitioners provide rich evidence that can be seen as supporting the information acquisition hypothesis behind CVC investment.1 In response to a question regarding the important strategic aims of CVC investment, 100% of the respondents agreed on the importance of “CVC provides window on new technology,” 54% considering it extremely important. Ninety-sic percent agreed on CVC’s advantages in “providing window on new markets.” When asked about the speciﬁc use of new information accessed through CVC investment, Corporate Venture Capitalists responded: “seek new directions” (92%), “develop new products” (90%), and “supporting existing businesses” (94%).
To fulﬁll the role of acquiring technological and market knowledge from entrepreneurial companies, CVCs are organized and operated to bridge their corporate parents with their portfolio startups. On the one hand, CVCs actively interact with 1 I urge interested readers to refer to the original report.
72 portfolio companies by taking board seats. More than 75% of CVCs frequently take board seats in their portfolio companies, and two-thirds always take a seat. These board seats are as often non-voting board observers rather than as voting board members. On the other hand, CVCs communicate with business units within their parent ﬁrm that are relevant to their CVC portfolio companies—98% CVCs involve R&D or business development personnel in their operation, and 64% intensely do so.
A.3 Acquiring Portfolio Companies
Information acquisition through CVCs seldom evolves into asset consolidation (i.e., acquisition of the portfolio companies). Recent studies examine acquisition cases when CVC investors acquire portfolio companies in which they invested (Benson and Ziedonis, 2010; Dimitrova, 2013). In general, acquiring portfolio companies is rare—fewer than one-ﬁfth of CVC investors acquired their portfolio companies.
CVCs that did conduct such acquisitions acquired fewer than 5% of their portfolio companies (that is, one out of 20 investments). In another tabulation, acquiring portfolio companies constitute about 16% of all the acquisitions conducted by those CVC investors.
Interestingly, Benson and Ziedonis (2010) show that acquiring portfolio companies destroys shareholder value, because the market usually reacts negatively to such deals.
Dimitrova (2013) shows that CVC parents acquire portfolio companies when their innovation productivity is low and that those acquisitions on average signal poor future innovation prospects. In broader terms, these ﬁndings are consistent with the information acquisition rationale proposed in this paper—the main beneﬁt of CVC investment is from acquiring information from not speciﬁc assets of, portfolio companies.
73A.4 Macromedia: An Illustrative Case
As an illustration of Corporate Venture Capital, consider Macromedia Ventures, a CVC unit launched in early 2000 by Macromedia, Inc., a California-based graphics and web development software company. In the late 1990s, the ﬁrm experienced a downturn in its traditional area of media-design authoring software. It therefore laid out a blueprint to explore new technologies on the market such as Internet-based tools and services, and the CVC unit was an integral part of the blueprint. Macromedia Ventures was organized as an internal division of Macromedia, Inc., and invested in entrepreneurial companies on behalf of its corporate parent. Within two years of CVC initiation, Macromedia Ventures had invested in nine companies (12 rounds in total), covering Internet-based technologies such as online distribution and mobile solutions.
The reason behind the CVC investment, declared by Rob Burgess, then Chairman and CEO of Macromedia, “Macromedia already delivers products that deﬁne the cutting edge of the Internet, but it is in our best interest to tap the entrepreneurial innovation going on around us, given the pace of the Internet.” The short period for conducting CVC investment was important for Macromedia, because it enabled the ﬁrm to recover from its deteriorated condition and ﬁrmly establish itself as a successful web ﬁrm. The selected portfolio companies guided and facilitated the redevelopment of Macromedia’s product lines and future business models. For instance, Macromedia Ventures invested in Alterego Networks, a developer of enterprise web services that is an innovative extension of Macromedia’s technology, and AtomFilms, an innovative online short-ﬁlm distributer that facilitated Macromedia’s extension to online entertainment services. Some other portfolio companies were also successful.
This appendix provides more discussions on the instrument Obsolescence, which is used in the paper to shock corporate innovation.
B.1 The Conceptual Idea and Its Roots The proposition that knowledge obsolescence aﬀects innovation has its roots in four basic observations.