Allocation of attention within venture capital firms

Abstract

In this article we use a simple queuing network to model the process through which entrepreneurs receive venture capital funding. Our model focuses in particular on the allocation of venture capitalists' attention between pre- and post-investment activities, and on the degree of selectivity in deciding which ventures to fund. Based upon this model we develop expressions for the financial performance of the venture capital process, both overall and also from the perspectives of the entrepreneurs, investors, managers, and venture capitalists involved. For these financial measures we derive the optimal allocation of attention between pre- and post-investment activities, and the optimal proportion of venture proposals to accept. Further analysis shows how these different financial measures and optimal values respond to changes in the business climate. More interestingly, our analysis also shows where and to what extent the different parties could be expected to agree or disagree on how best to manage the process.

Introduction

Cooperative relationships between venture capitalists (VCs) and entrepreneurs are necessary for the success of VC backed companies (Sapienza and Korsgaard, 1996; Steier and Greenwood, 1995; Cable and Shane, 1997). Timmons and Bygrave (1986) propose that the cooperative relationship between a VC and entrepreneur is more important to the success of the business than the capital itself. While it appears that co-operation provides significant benefits, it does not characterise all VC–entrepreneur relationships (Sahlman, 1990). There is a degree of uncertainty that the VC will pursue his or her own self interests rather than comply with the requirements of the contract for venture capital, i.e., VCs face agency risk (adapted from Fiet (1995)). This agency perspective offers a description of the risks a principal (the VC) faces when dealing with an agent (the entrepreneur) and actions that the VC can take to “force” the entrepreneur to co-operate. For example, the VC can introduce penalties for defection including a dilution of the entrepreneur's equity in the venture (Sahlman, 1990), replacement of the entrepreneur with professional management (Barney et al., 1996; Hoffman and Blakely, 1987) and through compensation structures (Sahlman, 1990).

There has been less research, however, about the incentives for VCs to pursue opportunistic behaviour at the expense of co-operation (and at the expense of the entrepreneur). In this case the VC can be considered the agent and the entrepreneur the principal (Cable and Shane, 1997). For example, VCs may act opportunistically to harvest a venture (and obtain their profits) rather than reinvest in the venture's future products and development (Sahlman, 1990), they may pressure the entrepreneur to seek short-term profits (Gomez-Mejia et al., 1990) even to the detriment of long-term profitability and/or the VCs may underinvest in the venture (Cable and Shane, 1997; Gorman and Sahlman, 1989; Bygrave and Timmons, 1991). For another example, VCs often face situations where they are flooded with entrepreneurs' requests for financing (in the form of lengthy business plans) and simultaneously need to assist current portfolio ventures in improving performance. The more active VCs spend 35 hours per month per venture assisting the management team, while less active firms spend less than 7 hours per month per investment (Elango et al., 1995).

Therefore, it appears that VCs may opportunistically manage the venture capital process in a way that advantages them and disadvantages the entrepreneur. Furthermore, entrepreneurs are not the only stakeholders of the venture capital process and the way that it is managed likely impacts the desirability of the outcome for these other stakeholder groups. For example, scholars have relatively ignored the relationship between VC firms and their funds providers (Robbie et al., 1997).

The important question is how should time, effort and/or resources within the VC firm be allocated? The optimal allocation likely depends on the perspective taken. What is optimal for the VC might not be optimal for the entrepreneur or the limited partners supplying the fund's capital to the VC. Therefore, rather than focus on the control mechanisms that a VC can use to ensure the entrepreneur acts in a co-operative manner, this paper investigates the way the VC process is managed and its impact on the desirability of the outcome, as perceived by the different stakeholder groups.

The above discussion leads to this paper's two key research questions. What is the allocation of a VC firm's attention (time, talent, and treasure) between pre-investment activities and post-investment activities that is considered optimal across all stakeholders? How and under what conditions is the allocation of attention that is optimal for the process overall (i.e. is best for the common good) different from what might be preferred by the various stakeholders (entrepreneurs, investors, managers and employees)? We investigate these important questions by using a queuing network model. A queuing network is a system of queues (waiting lines) where the departures from one queue become the arrivals at another. Queuing networks are useful models of systems that involve some form of customer flow (e.g., of venture proposals) amongst various locations (e.g., stages in the venture capital process) where the inherent variability of the flow can lead to line-ups and delays at a given location (e.g., an overflowing stack of proposals to review). Despite the usefulness of this approach to the investigation of the venture capital process it has yet to be used for this purpose, though it is often used to evaluate systems of machines in a factory (e.g., Papadopoulos et al., 1993) or the operations of health care service providers (e.g., Bretthauer and Cote, 1998).

By addressing these important issues facing VC firms, the present article makes a number of contributions to the literature. First, a VC firm (as all firms) has a number of stakeholders each with different expectations and aspirations for the firm. This article provides a deeper understanding of the trade-offs amongst these stakeholder groups. Second, research on the VC process has made a significant contribution to the literature, in particular Tyebjee and Bruno (1984), although such process research is limited in number. This article builds on these primarily descriptive studies, to provide a decision model with prescriptive benefits. Such benefits are likely to have important practical implications for the managers of VC firms and may also benefit other entrepreneurial financing firms (e.g., the Small Business Development Corporation).

Third, the allocation of attention is an important issue as individual VCs have limited time and energy to allocate among the various tasks of the venture capital process. To VC firms, VCs represent a costly resource to acquire and maintain (i.e., a substantial fixed cost for the VC firm) and represent the basis of a VC firm's competitive advantage. Thus, VCs' time is valuable resource that needs to be effectively managed. This paper provides prescriptions on how the process can be managed to provide a higher profit for the VC firm. But the support of stakeholders is also important and this paper provides some insight into the possible reactions of different stakeholder groups depending on how the process has been managed.

The remainder of the paper proceeds as follows. We begin by reviewing the existing research on the VC process. We then describe our assumptions about the venture capital process in terms of operations and financing. Based upon these assumptions, we formulate models of financial performance for the VC process from the different viewpoints of the investors, managers, and VC employees. We analyse these different measures of performance to investigate how they would vary in response to changing business conditions, and more interestingly how the different entities involved would likely agree or disagree regarding the management of the process. We conclude with a discussion of the limitations of the model and suggest possible avenues for future research. Appendix A contains the details of our mathematical analysis.