Dushnitsky_Lenox (SMJ 2005)

Embed Doc
Copy Link
Readcast
Collections

CommentGo Back

Download

Strategic Management Journal

Strat. Mgmt. J.

,

26

: 947–965 (2005)Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/smj.488

WHENDOFIRMSUNDERTAKER&DBYINVESTINGINNEWVENTURES?

GARY DUSHNITSKY

1

* and MICHAEL J. LENOX

2

1

TheWhartonSchool,UniversityofPennsylvania,Philadelphia,Pennsylvania,U.S.A.

2

Fuqua School of Business, Duke University, Durham, North Carolina, U.S.A.

We explore the conditionsunder which firms are likely to pursueequity investment in new venturesas a way to source innovative ideas. We find that firms invest more in new ventures—commonlyreferred to as ‘corporate venture capital’—in industries with weak intellectual property protec-tion and, to some extent, in industries with high technological ferment and where complementarydistribution capability is important. Furthermore, we find that the greater a firm’s cash flow and absorptive capacity, the more likely it is to invest. Our results suggest that in Schumpeterianenvironments incumbents may supplement their innovative efforts by tapping into the knowledgegenerated by new ventures.

Copyright



2005 John Wiley & Sons, Ltd.

Scholars have long been interested in the com-ponents and form of the ‘knowledge productionfunction’—the process by which innovative inputsare transformed into innovative outputs. Histor-ically, the innovation literature has focused onthe role of internal research and development onﬁrm innovation (e.g., Griliches, 1979). However,internal R&D expenditures play only a partialrole in ﬁrm innovation rates. Increasingly, schol-ars recognize that the ability to exploit

external

knowledge is critical to firm innovation (Cohenand Levinthal, 1990; Henderson and Cockburn,1994; Teece, Pisano, and Shuen, 1997). Indeed,in the past decade attention has shifted to therole of innovative inputs that reside outside thefirm’s boundaries. Among others, researchers havelooked at how firms access knowledge in aca-demic and government labs through professional

Keywords: innovation; external R&D; corporate venturecapital

∗

Correspondence to: Gary Dushnitsky, The Wharton School,University of Pennsylvania, 2031 Steinberg Hall–Dietrich Hall,Philadelphia, PA 19104, U.S.A.E-mail: gdushnit@wharton.upenn.edu

networks (Cohen, Nelson, and Walsh, 2002), inestablished competitors through alliances (Hage-doorn and Schakenraad, 1994; Gulati, 1995; Pow-ell, Koput, and Smith-Doerr, 1996), and in newventures through equity investment (Dushnitskyand Lenox, 2005).For the most part, researchers have studied thepotential for various external sources to provideinnovative knowledge. The alliance literature hasfound that innovative alliance partners may pro-vide important learning benefits to firms (Hage-doorn and Schakenraad, 1994; Dussauge, Gar-rette, and Mitchell, 2000; Stuart, 2000; Rothaer-mal, 2001). Others have found that maintaininglinks with universities and professional networksis important for innovating. However, the resultsof these studies are conditional on the firms suc-cessfully establishing linkages. Less studied are thefactors affecting the initial selection of these exter-nal sources especially with respect to alternativeinvestments such as internal research and develop-ment.A handful of scholars have begun to address thisissue in the alliance literature by examining the

Copyright



2005 John Wiley & Sons, Ltd.

Received 16 June 2003Final revision received 11 April 2005

948

G. Dushnitsky and M. J. Lenox

decision to form an R&D alliance (Gulati, 1995;Stuart, 1998; Ahuja, 2000). However, these studiesare limited because they do not observe the costto the ﬁrm of participating. As a result, they areunable to discern the elasticity of external invest-ment with respect to various industry and ﬁrm fac-tors including investment in internal R&D. Whilethe decision to commit resources towards

inter-nal

innovative inputs (i.e., R&D expenditures)has received much scrutiny (Hall, 1992; Himmel-berg and Petersen, 1994), there remains a need tostudy ﬁrms’ decisions to commit resources towards

external

innovative inputs.In this paper, we focus on one strategy avail-able for firms to source external knowledge. Weexplore the conditions under which establishedfirms source innovative ideas through investmentin external entrepreneurial ventures (Roberts andBerry, 1985). Commonly referred to as ‘corporateventure capital’ (CVC), these investments consistof minority equity stakes in relatively new, notpublicly traded companies that are seeking capitalto continue operation. High-tech companies (e.g.,Intel, Sony, and Motorola), pharmaceutical giants(e.g., J&J), and media concerns (e.g., News Corp.)have invested millions in start-ups. In the year2000 alone, nearly $16 billion was invested byover 300 corporations—representing 15 percent of the entire venture capital market. Despite the eco-nomic downturn and subsequent reduction in CVCinvestment, numerous companies have maintaineda strident commitment to their venturing programs(Chesbrough, 2002).Corporate venture capital investment is anappealing setting for the study of firms’ decisionsto pursue external innovative inputs. Previouswork has found that CVC investment may bean effective way for firms to increase theirinnovative output (Dushnitsky and Lenox, 2005).Unlike other inter-organizational arrangements,CVC investment is a capital expenditure that iseasily observed and measured. The deployment of other external innovative inputs is often difficultto observe and it is even more difficult todetermine their cost. For example, what price doesone place on maintaining personal ties with starscientists (Cohen

et al

., 2002)? Data on the costof maintaining R&D alliances are typically notavailable to researchers and may not even becalculated by alliance members. The ability tomeasure the dollar amount of corporate venturecapital investments enables us to better captureits elasticity with respect to various industry andfirm factors. More importantly, these investmentsare observed irrespective of their success orcontribution to firm innovation rates.We propose a number of hypotheses concerningthe decision to invest corporate venture capital.The driving logic behind our hypotheses is thata profit-seeking firm chooses to invest corporateventure capital when CVC’s marginal innovativeoutput is expected to be higher than that of inter-nal R&D. An empirical test of these hypothesesis presented based on a sample of over 1000 U.S.public firms during the time period 1990–99. Pri-mary data were gathered from Venture Economics’VentureXpert database of the venture capital indus-try. These data were augmented with data fromStandard & Poor’s Compustat dataset, the NBERversion of the U.S. Patent database (Hall, Jaffe, andTratjenberg, 2001), and the Carnegie Mellon Sur-vey (CMS) of Research and Development (Cohen,Nelson, and Walsh, 2001).We find that firms invest more in new ven-tures in industries with high technological ferment,weak intellectual property protection, and wherecomplementary distribution capability is impor-tant. Furthermore, we find that the greater a firm’scash flow and absorptive capacity, the more likelyit is to invest. Interestingly, we present evidencethat internal R&D and CVC investment are per-haps complements rather than substitutes vyingfor research dollars. These results have impor-tant implications for the organization of R&D ingeneral and the use of CVC in particular. Ourresults suggest that in Schumpeterian environmentsincumbent firms may choose to tap into the knowl-edge generated by new ventures as a way toincrease their own innovation rates.

THEORY AND HYPOTHESES

A number of scholars have advanced the idea thatentrepreneurial ventures are likely to be the sourceof highly valuable and innovative ideas (Tushmanand Anderson, 1986; Kortum and Lerner, 2000;Shane, 2001a). At the heart of this argument isa consideration of the ability of ﬁrms to employstar scientists in internal labs. Amit, Muller, andCockburn (1995) propose that the decision to starta new venture is undertaken when the value of self-employment is higher than the opportunity cost(i.e., lost salary from incumbent). In the later half

Copyright



2005 John Wiley & Sons, Ltd.

Strat. Mgmt. J.

,

26

: 947–965 (2005)

Corporate Ventures and Research

949

of the 20th century, highly skilled human capital(labor) has become more important in generatinginnovation than physical capital (Zingales, 2000).In this new setting, skilled researchers will likelydisassociate themselves from his or her corporatelaboratory and form independent firms (Aghionand Tirole, 1994).Following this line of reasoning, researchers willopt away from fixed salary (i.e., remaining a cor-porate employee) and toward profit sharing (i.e.,founding their own new venture) only when theythink the idea is highly lucrative (Dix and Gandel-man, 2000). Thus, we expect to observe the forma-tion of new ventures only when entrepreneurs havehighly innovative ideas (Aghion and Tirole, 1994).Consistent with this prediction, Kortum and Lerner(2000) observe that entrepreneurial, human-capitalintensive ventures generate higher levels of patent-ing output than established firms. Shane (2001a)provides empirical evidence that the decision toform a new venture is associated with underlyingentrepreneurial inventions that have high economicvalue.This line of reasoning suggests that the marginalR&D productivity of new ventures is likely to behigher than established firms. While this may sug-gest that established firms will thus favor CVCinvestment over internal R&D, we must recog-nize that there are potential costs to incumbentsusing CVC investment as a vehicle to tap theseinnovative ventures. First, the presence of signif-icant information asymmetries between new ven-tures and their corporate investors opens incum-bent firms to potential adverse selection—a prob-lem that is likely far less pronounced in internallaboratories. Second, the independent entrepreneurhas greater leverage to hold up the investing firm.Thus,

ex ante

adverse selection and

ex post

hold-up may negate the learning beneﬁts of investingin innovative new ventures. In considering CVCinvestment, a focal ﬁrm regards the marginal R&Dproductivity of new ventures

net

of potential lossesdue to the inherent adverse selection and hold-up problems. Consequently, we expect to observeCVC investment only in technological domainswhere CVC’s

net

marginal innovative output isexpected to be higher than that of internal R&D.In the sections that follow, we develop a set of hypotheses concerning the conditions under whichthis balance will favor CVC investment and firmswill consequently seek knowledge through equityinvestment in new ventures.While our hypotheses assume that firms investin new ventures to acquire knowledge, we mustrecognize that a firm may pursue CVC investmentsimply to generate a high return on investment(Block and MacMillan, 1993; Chesbrough, 2002;Siegel, Siegel, and MacMillan, 1988). During thestock market bubble of the late 1990s, some firmsviewed CVC investment as a way to capitalize onthe inflated values of technology ventures. Firmsgained a return on investment primarily by sellingshares in a venture after an initial public offering(Gompers and Lerner, 2001). During the latterhalf of the 1990s, the price of many venturesdoubled on the first day of trading (Ritter, 2001).Such lucrative exits were highly dependent onmarket conditions and had a strong periodicitycorresponding to the stock market.While it is important to consider financial driversof CVC investment, we propose that firms mainlypursue such investments for strategic reasons.Previous research suggests that most firms viewCVC investment as a window on technology. Thedeclared goal of Nokia Ventures, the CVC programof Nokia, is to ‘fuel future growth and to boostnew product and long-term business development’(

Business Wire

, 1998). Surveys support this obser-vation (Block and MacMillan, 1993; Chesbrough,2002; Ernst & Young, 2002; Winters and Murﬁn,1988). Yost and Devlin (1993) report that 93 per-cent of corporate venture capitalists in their sam-ple view strategic objectives as one of their mainobjectives. Siegel

et al

. (1988) report that corpora-tions rank ‘exposure to new technologies and mar-kets’ as the leading objective for engaging in cor-porate venture capital programs. Similar results arereported by Block and MacMillan (1993) and Win-ters and Murﬁn (1988) and more recently in a sur-vey of more than 40 corporations (Ernst & Young,2002). Further support is provided by recent empir-ical work examining the relationship between CVCinvestment and ﬁrm innovation (Dushnitsky andLenox, 2005).

Industry drivers

First, we examine the degree to which generalindustry and technology characteristics may drivethe decision to invest in new ventures within aparticular sector. We propose that ﬁrms will mostlikely invest in sectors with rich technologicalopportunities, weak intellectual property protec-tion (in particular, patent protection), and where

Copyright



2005 John Wiley & Sons, Ltd.

Strat. Mgmt. J.

,

26

: 947–965 (2005)

of 00

Leave a Comment