Strategic Management Journal Strat. Mgmt. J., 31: 990–1017 (2010) Published online EarlyView in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/smj.851 Received 20 August 2007; Final revision received 18 February 2010 ENTREPRENEURIAL FINANCE MEETS ORGANIZATIONAL REALITY: COMPARING INVESTMENT PRACTICES AND PERFORMANCE OF CORPORATE AND INDEPENDENT VENTURE CAPITALISTS GARY DUSHNITSKY1* and ZUR SHAPIRA2 1 London Business School, Regent’s Park, London, U.K. 2 Stern School of Business, New York University, New York, U.S.A. This paper investigates the effect of compensation of corporate personnel on their investment in new technologies. We focus on a specific corporate activity, namely corporate venture capital (CVC), describing minority equity investment by established-firms in entrepreneurial ventures. The setting offers an opportunity to compare corporate investors to investment experts, the independent venture capitalists (IVCs). On average, we observe a performance gap between corporate investors and their independent counterparts. Interestingly, the performance gap is sensitive to CVCs’ compensation scheme: it is the largest when CVC personnel are awarded performance pay. Not only do we study the association between incentives and performance but we also document a direct relationship between incentives and the actions managers undertake. For example, we observe disparity between the number of participants in venture capital syndicates that involve a corporate investor, and those that consist solely of IVCs. The disparity shrinks substantially, however, for a subset of CVCs that compensate their personnel using performance pay. We find a parallel pattern when analyzing the relationship between compensation and another investment practice, staging of investment. To conclude, the paper investigates the three elements of the principal-agent framework, thus providing direct evidence that compensation schemes (incentives) shape investment practices (managerial action), and ultimately investors’ outcome (performance). Copyright 2010 John Wiley & Sons, Ltd. INTRODUCTION reality in which corporate research and develop- ment (R&D) investment is undertaken, however, Corporate personnel are often required to pursue often hinders these efforts. In a seminal paper, investment in innovative yet risky projects. The Jensen (1993) notes that the annual R&D spend- ing of either IBM or General Motors exceeds the aggregate annual disbursement in the venture capi- Keywords: principal-agent framework; incentives; tal industry, yet the economic successes of venture- compensation; corporate entrepreneurship; venture capital backed firms have been profound. He ascribes this ∗ Correspondence to: Gary Dushnitsky, London Business School, Regent’s Park, London NW1 4SA, UK. to unfavorable incentives within corporate research E-mail: [email protected] facilities. In doing so, Jensen underscores the need Copyright 2010 John Wiley & Sons, Ltd. Entrepreneurial Finance Meets Organizational Reality 991 to understand the effects of compensation schemes Importantly, this setting affords empirical inves- on investment decisions and ensuing outcomes. tigation that goes beyond correlating incentives We heed the call and study the impact of cor- and performance. We open the black box and porate incentives on one such corporate activity: document a direct relationship between incentives corporate venture capital (CVC). CVC is the and the investment practices managers engage in. practice of minority equity investment by estab- The paper therefore addresses a lacuna in the lished firms in entrepreneurial ventures, that is, principal-agent literature. Moreover, our research innovative companies that seek capital to con- setting affords intuitive interpretation of man- tinue operations. CVC investment opens a window agerial action and corporate performance in the onto new markets and novel technologies, thus spirit of Jensen (1993) and Hamel (1999)—to offering established firms an opportunity to ad- the extent that IVCs are expert investors, they vance their innovation efforts. Per Jensen’s (1993) serve as a benchmark against which CVCs can be astute comparison, we investigate the effect of assessed. incentives awarded to corporate and independent To that end, the paper conducts extensive anal- venture capitalists (IVC): How do compensation yses of venture capital investors during the 1990s. schemes influence the investment practices ven- We study the direct relationship between investors’ ture capitalists pursue? What are the performance compensation schemes and investment practices implications? using a sample of 13,096 investment rounds by Research motivated by these questions can ben- corporate and independent venture capitalist. This efit from, as well as contribute to, the principal- analysis focuses on investment practices that CVCs agent literature. By linking agents’ (e.g., man- and IVCs commonly use to manage investment agers) pay to performance, the theory conjec- uncertainty (Gompers and Lerner, 2001): staging tures, principals (e.g., shareholders) motivate them (i.e., targeting distinct stages of a venture’s devel- to invest in profitable yet risky projects that opment; Gompers, 1995) and syndication (i.e., the agents may otherwise forego. Empirically, a coinvesting a round by two or more investors; large body of work investigates the association Lerner, 1994). For instance, analysis of investment between firms’ compensation scheme and ultimate stage indicates that CVC investors target ventures performance. The evidence, however, is incon- at later stages of development. That is, the invest- clusive (Core, Holthausen, and Larcker, 1999; ment practices of CVCs and IVCs differ. Interest- Dalton et al., 2003; Murphy, 1999; Tosi et al., ingly, the magnitude of the difference is affected 2000; Core, Guay, and Larcker, 2003). Notably, by the nature of corporate venture capitalists’ com- the insight that incentives affect performance by pensation. It is large when CVC personnel receive shaping managers’ behavior remains implicit to little or no performance pay, yet shrinks signifi- these studies. A more conclusive support for the cantly when they are awarded performance pay. principal-agent framework requires studying not Analysis of investment syndicates yields similar only the indirect incentive-performance associa- results. tion but also the theoretical mechanism that medi- We also compare investors’ ultimate perfor- ates the two—namely, managerial behavior (Mur- mance as a function of their compensation phy, 1999). Unfortunately, there is scant empirical schemes. The analysis covers 2,830 corporate work on the latter topic. Using the market for and independent investors who participated in entrepreneurial finance as an empirical setting, this the aforementioned rounds. A similar pattern em- paper presents a comprehensive analysis of these erges: CVCs experience successful portfolio exits issues. at a rate that differs from IVCs, and the mag- Specifically, CVC provides an attractive set- nitude of the performance gap is sensitive to ting to study all three elements of the principal- CVCs’ incentives. It is large when CVC person- agent framework. Theoretically, the framework nel are awarded performance pay, and diminishes analyzes how incentives shape managerial in- when they receive little or no performance pay. vestment in profitable yet uncertain projects. It We present mediation models that further illu- thus parallels our setting: CVC personnel invest strate that the compensation-performance associa- in new ventures whose technological viability tion is mediated by managers’ investment and commercial prospects are highly uncertain. practices. Copyright 2010 John Wiley & Sons, Ltd. Strat. Mgmt. J., 31: 990–1017 (2010) DOI: 10.1002/smj 992 G. Dushnitsky and Z. Shapira We review the principal-agent literature below. between executive compensation and firm perfor- The following section develops the hypotheses. mance. The evidence, however, is inconclusive. Data, methods, and results are discussed thereafter. Gomez-Mejia (1994: 199), for instance, states that The last section concludes. ‘it is amazing how little we know about executive pay in spite of the massive volume of empirical work... even more discouraging, when taken as THEORETICAL BACKGROUND a whole, results are conflicting and disappoint- ing.’ A literature overview by a leading compen- The principal-agent framework analyzes the rela- sation economist, Murphy (1999: 2539), also notes tionship between incentives and risk-taking behav- ‘...there is surprisingly little direct evidence that ior. A key upshot is that compensation schemes higher pay-performance sensitivities lead to higher may be used to guide a manager toward perfor- stock-price performance.’ Management scholars mance maximizing choices by affecting his or her echo that observation: ‘Researchers express shock risk preferences. This section reviews major theo- when they find pay/performance sensitivities are retical and empirical findings and points at areas low and the results inconsistent with their theory.’ for further investigation. (Tosi et al., 2000: 331). Recently, Dalton et al. Theoretically, an agency relationship is said to (2003 : 14) conducted a meta-analysis of 229 stud- exist between a firm’s shareholders (i.e., prin- ies in economics, finance, and management, and cipal) and its managers (i.e., agents). Principal- concluded that ‘...the empirical evidence provides agent models assume shareholders are risk neu- no consensus.’