Early-stage fundraising of university spin-offs: a study through demand-site perspectives

Abstract

University spin-offs have increasingly received attention from academia, governments and policy-makers. However, there are only a limited number of studies within the university spin-off context which fully understand the contribution made by the founding team to fundraising, specifically how they use their social networks and capabilities. Employing resource-based theory and social networks approach, this paper examines whether a founding team exploits its social networks and capabilities to signal the value of a university spin-off. Capabilities are analysed through a set of constructs – technology, strategy, human capital, organizational viability and commercial resource – that have been derived from previous literature. The contribution made by social networks is evaluated using three dimensions – structure, governance and content – which form the construct of relationships within a network. Based on data from 181 university spin-offs in Spain, this paper empirically demonstrates that by exploiting social networks a founding team can improve its capabilities which, in turn, enhance its fundraising ability.

Keywords:

• University spin-offs
• fundraising
• founding teams
• resource-based view
• social networks

Introduction

The financing of university spin-offs has received increasing attention from academia, governments and policy-makers. Recent studies have addressed government financing policies (Wonglimpiyarat 2013), venture capital investment decision-making (Aouni, Colapinto, and La Torre 2013), the roles of venture capitalists in the development of new ventures (Paik and Woo 2013) and the contributions of an entrepreneur’s social capital to the fundraising activities (Ozmel, Robinson, and Stuart 2013). However, it is recognized that early-stage fundraising continues to be a major challenge for university spin-offs to develop their inventions and knowledge into practical applications (Lindstrom and Olofsson 2001; Widding, Mathisen, and Madsen 2009). The imperfections of capital markets that arise from the uncertainty of investment returns, the asymmetric information between entrepreneurs and potential investors, and the lack of collateral create financial constraints and funding gaps for university spin-offs (Baldock 2015; Carpenter and Petersen 2002; Lehner, Grabmann, and Ennsgraber 2015). Studying the financing activities of the university spin-offs therefore requires more attention from both research and policy-makers.

In their theoretical review, Rasmussen and Sørheim (2012) found that previous studies emphasize the roles of investors in funding in both the early and later stages of development of new ventures. However, most of current research studying the early-stage fundraising of new ventures (including spin-offs) has been oriented towards the supply side (the investors) (Lindstrom and Olofsson 2001) despite the fact that the issues from the demand side (Murray 1999) significantly impact business development and so ultimately determines investment returns. Thus, the first distinction between this study and others is that it takes a demand-side perspective, looking specifically at the founding teams of university of spin-offs. It has emerged that new ventures are more likely to be created by founders plural, rather than singular (Gartner and Vesper 1994), and that entrepreneurial teams are at the heart of any new venture (Cooper and Daily 1997). Founding teams have become more popular and important modes of new business development (Chandler, Honig, and Wiklund 2005; Lasch, Le Roy, and Yami 2007); their importance also is reflected in the insights from investors who constantly consider the quality of teams as an important funding criterion (Zacharakis and Meyer 1998).

This study is also distinct from others by focusing on the early-stage fundraising of university spin-offs which face a fundamentally different set of challenges on account of the context in which they were created (Vohora, Wright, and Lockett 2004a). A university spin-off is characterized by highly innovative products or services that are new and unique to the market (Heirman and Clarysse 2004). However, performance of these spin-offs is comparatively poor compared to other new ventures because founding teams have to deal with complex tasks in unfamiliar and uncertain business environments (Shane 2004) which are exacerbated by their limited industrial experience and/or access to non-technical networks (Cooper and Daily 1997). This generates a certain amount of scepticism from investors about the likely success of spin-offs (Clarysse and Moray 2004). To offset these limitations, universities will often provide incubation space to staff with commercially potential ideas (Clarysse and Moray 2004). This creates an artificial time lag between idea generation and company formation that creates an opportunity to fine-tune the idea and change the structure and composition of the founding team before incorporation (Vanaelst et al. 2006). Changes in personnel are often necessary because the technological founders may exhibit less commitment to the commercialization of the idea, have lower growth aspirations (Clarysse and Moray 2004; Vanaelst et al. 2006) and view themselves more as part-time entrepreneurs (Müller 2010). The time lag therefore allows an opportunity to re-balance the founding team of the university spin-off with the introduction of individuals with more commercial experience, particularly in the market segments targeted by the spin-off (Visintin and Pittino 2014; Vohora, Wright, and Lockett 2004b).

A final source of distinctiveness of this paper is its focus on the unobservable elements – the capabilities and social networks of founding teams – that signal the value of university spin-offs. In particular, the paper, inspired by the idea of focusing on demand-side perspectives, examines some fundamental questions which will contribute to the theory-based understanding of the early-stage fundraising of university spin-offs. Can a founding team exploit its social networks to signal the value of a university spin-off to improve the chance of obtaining early-stage funds? Can a founding team use its capabilities to signal the value of a university spin-off to improve the chance of obtaining early-stage funds? To address these questions and strengthen the theoretical and empirical foundation of university spin-off studies, the paper adopts a resource-based view to measure the capabilities of founding teams in terms of entrepreneurial technology, strategy, human capital, organizational viability and commercial resources, and uses social capital theory to analyse the networks of founding teams through three dimensions: structure, governance and content. These characteristics will be analysed in terms of their influence on raising new investment by a university spin-offs. Additionally, signalling theory will be employed to develop and test a theoretical framework linking the early-stage fundraising of university spin-offs to both the capabilities and social networks of the founding teams. The results presented are based upon a sample of 181 Spanish university spin-offs based in 35 universities across all regions of Spain. Each spin-off was created and developed by a founding team and responses were obtained from the members of founding teams. The findings indicate that the capabilities of founding teams affect the early-stage fundraising of university spin-offs, but fail to demonstrate the relationships between early-stage fundraising and the social networks of founding teams. The findings from this research make a valuable contribution to the theoretical base for the rationale behind the creation of ecosystem based on social networks to promote the creations and growth of university spin-offs.

Early-stage fundraising and financial market imperfections

The early-stage financial needs of university spin-offs develop through three phases: seed, start-up and early-growth (Lindstrom and Olofsson 2001). In the university spin-off process models of Shane (2004) and Vohora, Wright, and Lockett (2004a), seed capital is typically provided by the host institution or public funding sources to support the research activities and develop the initial business concept. Start-up finance is required for early organizing efforts, including business registration to create a legal entity. Early-growth finance is needed for the initial product development and market entry. This paper solely considers the early-stage fundraising activities of spin-offs. It is suggested that effective fundraising helps entrepreneurs in the commercialization process (Powers and McDougall 2005), whereas undercapitalization is a consistent cause of failures not only in the foundation stage but also in the growth period of new ventures (Rosman and O’Neill 1993). Thus, founding teams must choose to explore suitable financial sources within the capital market depending upon their growth goals, the nature of ownership, firm size and sector of the spin-offs (Riding, Orser, and Chamberlin 2012). Lindstrom and Olofsson (2001) have suggested that university spin-offs experience more difficulties in early-stage financing than start-ups from other origins. Geographical context also matters. Spain has been ranked as the second worst country on barriers to entrepreneurship in 29 nations. Spanish university spin-offs face particularly severe financial constraints compared to the rest of Europe because of the lack of private financial institutions and venture capitalists to financially support the creation of a high-risk venture (Berbegal-Mirabent, Lafuente, and Solé 2013).

In this study, early-stage financial supporters are classified into existing investors (i.e. follow-on investors), who provide seed capital to create university spin-offs, and potential investors (i.e. new investors), who may invest in spin-offs in the future (Harrison and Mason 2000; Shane 2004). Lindstrom and Olofsson (2001) suggested that finding access to these financial sources has become a key challenge for early-stage spin-offs because of the effects of capital market imperfections. Carpenter and Petersen (2002) suggested three reasons for these effects. First, the low probability of financial success and the high failure rate of university spin-offs generate uncertainty regarding investment returns, which affects the investment decisions of the investors. Second, the university spin-offs have limited assets to use for collateral and thus possess little salvage value in the event of failure. Third, it is difficult for financial providers to evaluate new knowledge because of information asymmetry between university spin-offs and potential investors. Founding teams and investors have unequal access to information on the spin-off (Certo 2003). That is, founding teams possess inside information on the true intentions, planned activities and value of the firms, whereas outside investors do not (Prasad, Bruton, and Vozikis 2000). This information asymmetry can result in the rejection of good investment opportunities or underinvestment in these opportunities (Myers and Majluf 1984). Information asymmetry influences the ability of a new venture to access financial sources and determines the firm’s capital structure (Fama and French 2005). Because their ultimate purpose is to maximize returns, investments are likely to be undertaken when financial providers perceive the value of university spin-offs and can mitigate the risks (Cumming and Johan 2008). Thus, to surmount the effects of capital market imperfections and attract more financial supporters, this paper proposes that founding teams could provide information that signals that university spin-offs have wealth-creating potential.

Social networks of founding teams

The quality of a founding team’s social networks, which is an external resource, is an important element in the fundraising process of a university spin-off (Shane 2004; Vohora, Wright, and Lockett 2004a). A social network includes single nodes (actors) and linkages between these nodes (dyads) and is “a sum of actual and potential resources embedded within, available through, and derived from the networks of relationships possessed by individual social units” (Nahapiet and Ghoshal 1998). Here, the analysis divides a social network into three components: structure, governance and content, as suggested by both Amit and Zott (2001) and Hoang and Antoncic (2003). Network structure refers to the properties of the connections and the personal configurations of the relationships among actors (Burt 1987). The presence or absence of network ties, network configurations and network morphology are the most important facets of the structural dimension (Tichy, Tushman, and Fombrun 1979). These facets describe the pattern of relationships as density, connectivity and hierarchy (Amit and Zott 2001). Network governance is defined as the mechanisms that govern the relationships among actors, the legal forms of actors, and the incentives for participation within networks. These mechanisms based on power, influence, reputation, relationship reciprocity and trust support network sustainability more than legal enforcement (Amit and Zott 2001). Content within a network refers to exchanging resources (Amit and Zott 2001). Such resources include ideas, information and advice (Smeltzer, Van Hook, and Hutt 1991) and the less tangible emotional support for entrepreneurs who are willing to take risks and increase their persistence to remain in business (Bruderl and Preisendorfer 1998).

Founding teams have the capacity to exploit links with industrial sectors to support the development of commercial (Dubini and Aldrich 1991), management and leadership expertise (Kitagawa and Robertson 2012). They can also use co-operative links with university staff to access the latest knowledge and technology, which reduces development costs (Markman et al. 2005) in the creation of innovative products (Lockett and Wright 2005). The greater that the density of these links is (the level of interconnectedness), the greater the opportunity a founding team will have to access the resource available in the network (Newbert and Tornikoski 2013). Interconnectedness is often a function of an actor’s position in the network, and founding teams that occupy a central position in a network would expect to have more opportunities to investigate and access the resources more efficiently and effectively (Stam and Elfring 2008).

Antecedent activity often results in reciprocal arrangements within networks that enable the founding team to access critical resources through cooperative arrangements that have been established over time (Witt 2004). As these relationships develop, trust is enhanced between the founding teams and their networks. This enables the teams to bypass expensive search activity using the network to reduce risk and limit the need for expensive due diligence when accessing key resources (Jones, Hesterly, and Borgatti 1997). Reciprocity and trust increase the reputation of a founding team over time, and this characteristic creates greater breadth and depth of interaction with the network. In essence, when combined, the mechanisms that govern networks enhance the competitive advantages that a founding team can acquire from its networks (Witt 2004).

The process of mobilizing resources from external sources is a vital task in the entrepreneurial process (Aldrich and Martinez 2001). It has been suggested that founding teams may access critical resources at below-market cost thanks to their relationships with resource gatekeepers (Newbert and Tornikoski 2013). The types and quality of such resources characterize the content of networks (Amit and Zott 2001). Resource types can be tangible or intangible and include ideas, strategic advice (Yli-Renko, Autio, and Sapienza 2001), access to financial providers (Kitagawa and Robertson 2012), technology (Lockett and Wright 2005), appropriate staff (Tolstoy and Agndal 2010) and emotional support (Bruderl and Preisendorfer 1998). In the case of spin-offs, the social capital of a university can often confer security and scientific credibility that enables access to resource gatekeepers (Newbert and Tornikoski 2013). In addition, where university incubators are employed, spin-offs can take advantage of previously developed and fostered internal and external networks that can provide access to important information and resource (Kitagawa and Robertson 2012; Patton and Marlow 2011). The value of networks to a spin-off depends on the collective activities of the founding team and university support mechanisms to identify, acquire and exploit appropriate relationships (Chandler and Lyon 2009). For the identified reasons, this paper proposes that the social networks of founding teams, developed in conjunction with university support, can contribute importantly to the resource and knowledge acquisition of founding teams.

H1: The founding team of a university spin-off can improve its capabilities by exploiting social networks.

Nofsinger and Wang (2011) argued that in the venture’s early stages founding teams do not belong to professional networks in capital markets (e.g. networks for initial public offering and seasoned equity offering pricing and distribution, networks that offer co-underwriting, venture capitalist networks) and thus must rely on their social networks. Many studies have demonstrated that social ties provide a potential mechanism to reduce the information asymmetry between potential investors and founding teams (Freiburg and Grichnik 2012; Uzzi 1996). Financial providers can reduce the information asymmetry regarding the intentions and planned activities of the teams and the value of university spin-offs through contingency (i.e. incentive) contracts and monitors (Granovetter 2005). Asymmetric information can be alleviated via signals (Certo 2003) conveyed by the knowledgeable parties or/and through screening activity that seeks additional information from uninformed parties (Carpentier, L’Her, and Suret 2010). Social relationships enable potential investors to obtain private information on the talents and tendencies of members of founding teams (Nofsinger and Wang 2011) and to resolve moral-hazard issues (Shane and Cable 2002). By associating with well-regarded individuals and organizations, founding teams can increase their reputation based on the past performance of the members of founding teams in attracting and convincing investors to invest in their business projects (Podolny 1994). Social networks also leverage the trust between founding teams and financial providers (Kautonen et al. 2010), which may eventually positively influence the investment decision.

H2: The social networks of founding teams leverage the early-stage fundraising of university spin-offs.

Capabilities of founding teams

Vohora, Wright, and Lockett (2004a) characterized the capability construct as encompassing entrepreneurial technology, organizational viabilities, human capital, entrepreneurial strategy and commercial resources. This characterization is adopted here. For the purposes of this paper, a capability that supports entrepreneurial technology is defined as a research outcome with the potential to be commercialized due to its limited imitability (Gallini and Wright 1990) or its ability to create a significant scale, a range of application or value (McGrath 1997). Organizational viability refers to internal systems that create institutional routines (Nelson and Winter 1982) that originate in internal communication (Krueger Jr 2000). Formal control mechanisms are defined as institutionalized rules, missions and regulations that create desirable patterns of behaviour (Covin and Slevin 1991), and organizational support (Leonard-Barton 1992) refers to the provision of appropriate training and reward structures (Zahra 1993). The human capital construct is measured based on the levels of education and experience available within the management team (Alvarez and Busenitz 2001; McKelvie and Davidsson 2009). Measures of proactiveness, innovativeness, risk-taking and competitive aggressiveness (Lumpkin and Dess 1996) are employed to measure entrepreneurial strategy making. A firm’s commercial resources are represented by the quality of its bespoke relationships with customers (Nadherny 1998; Powell and DentMicallef 1997). To create and maintain these trusting and value-enhancing relationships, complex coordination and communication skills are required (Hall 1993).

Generally, potential investors tend to look for a signal of future success from university spin-offs when making investment decisions (Wiltbank et al. 2009). Each investor has different scales for, and ratings of, a spin-off’s abilities based on technology, the market, and the management stage (Douglas and Shepherd 2002) and the business type, risk/returns ratio and time to exit (Wiltbank et al. 2009). Additionally, other scholars of early-stage fundraising have found that investment decisions depend on the investor’s perception of management skills, the business model, the potential market, the growth perspective (Mason and Harrison 2004), the shortcut heuristic (Maxwell, Jeffrey, and Lévesque 2011) and the presentation of the founding teams (Clark 2008). In addition, investors require the presence of well-balanced teams with sufficient business capabilities as an important criterion of their funding decisions (Muzyka, Birley, and Leleux 1996). Taking the founding teams as the unit of analysis, this study proposes the stage of the team’s capabilities as an unobservable element that signals the value of a university spin-off. Although investors and founding teams have different perceptions of a university spin-off’s potential for success (Douglas and Shepherd 2002), this study proposes that the capabilities of founding teams, which represent a hidden value of university spin-offs, positively determine the early-stage fundraising ability of such spin-offs (Figure 1).

H3: The capabilities of founding teams influence the early-stage fundraising of university spin-offs.

Figure 1. Conceptual model.

Methods

Sample and data collection

Our sample was drawn from 69 Spanish universities located in 17 autonomous communities. Each university has an office for the transfer of research results (OTRI). The OTRIs were created by the universities, which are both public and private, as part of the first Spanish National R&D Plan of 1988–1999 to enhance the relationships between the scientific world and production sectors. OTRIs engage in a wide range of R&D activities. However, only 35 are involved in the creation and development of spin-offs. While university spin-offs can be created by individuals or teams, the spin-offs examined in this research were created by teams that included at least one academic member from a university.

With the help of the OTRIs, a database of 862 spin-offs was created from which 181 responses were received (21% of the research population) to a web-based survey. All respondents were members of the founding teams and held positions on the executive board of the spin-offs. The spin-offs represented various sectors: information, computing and telecommunications (33.8%), engineering and consultancy (16.1%), medicine and health (15.3%), agriculture and biotechnology (15%), energy and environment (8.9%), aeronautics and the automotive industry (4.3%), electronics (3.4%) and other industries (3.2%). Most of the spin-offs (98%) were created in university incubators after 2003. The breakdown is as follows: 20% in 2009, 16% in 2010, 14% in 2006, 13% in 2008 and 2007, 7% in 2005, 5% in 2011 and 2004 and 7% in 2003 or earlier.

Measurements

To ensure the content validity of the measurements, this study uses questions that employ a seven-point Likert scale that was adopted from previous entrepreneurship and management studies (Antoncic and Hisrich 2001; Tsai and Ghoshal 1998). To assess the early-stage fundraising ability of a university spin-off, the questions required respondents to self-report on a variety of issues related to a founding team’s capabilities and social networks during the spin-off’s creation phase.

Early-stage financial sources

The study employs the suggestions of Shane (2004), Harrison and Mason (2000) as well as Greene and Brown (1997) to construct early-stage fundraising measurements, including follow-on investors (private investors or “angels”, venture capitalists, government grants and strategic partners) and new investors (the initial public-offering respondents, employees and customers).

Capabilities

The capability construct is derived from previous research (Antoncic and Hisrich 2001; Lumpkin and Dess 2001; McGrath 1997) and employs measures for entrepreneurial technology, organizational viability, human capital, strategy and the commercial resources of founding teams. More specifically, in terms of technology, the respondents were required to answer six questions regarding the ease of imitation, scope, continuity and the market signals of their entrepreneurial technology (McGrath 1997). To measure organizational viability, we adapted the measurements from Leonard-Barton (1992), Zahra (1993) and Antoncic and Hisrich (2001) to construct five questions that relate to the internal communication mechanisms, formal control mechanisms and organizational support within founding teams during the creation period. To measure human capital, a four-item measurement that evaluated industrial, managerial and entrepreneurial experience adapted from Alvarez and Busenitz (2001) and McKelvie and Davidsson (2009) was used. Questions that investigate innovation, proactiveness, risk-taking and competitive aggressiveness (Lumpkin and Dess 2001) were employed to measure entrepreneurial strategy-making. Finally, four questions regarding customer relationships, staff technology training and process design were used to measure the commercial resources of the founding teams (Nadherny 1998; Powell and DentMicallef 1997).

Social networks

By adapting prior management research, eight social-network measurements were constructed in the areas of ties, density, centrality, reputation, reciprocity, trust, information quality and diversity. The strength of the founding-team’s ties is measured by constructs that examine the willingness to engage in discussions related to social, political and family matters (Parks and Floyd 1996). The density of a network is measured by three-item scales that evaluate interactions within networks (Marsden 1993). Centrality is measured based on the measurements of Rowley (1997), which evaluate the location of actors within information flows using four questions on how directly respondents communicate with others within networks. To measure the quality of information within social networks, five questions developed by O’Reilly (1982) are employed, which evaluate the accuracy, relevance, reliability, specificity and timeliness of information. The availability of business-relevant information is used to measure the diversity of information within networks: market data, product designs, process designs, marketing know-how and packaging design or technology (Gupta and Govindarajan 2000). In addition, we measure trust using four questions, whereby respondents were required to self-report on how trustworthy they are perceived to be by other members within their networks (Tsai and Ghoshal 1998). Adapting Uzzi (1996) and Shane and Stuart (2002), a four-item measurement to evaluate the founder’s reputation was constructed to obtain the views of other participants within networks. Reciprocity is measured by four questions regarding the level of support, accumulation of favours and fairness in the relationships among members (Miller and Kean 1997).

Control variables

To ensure that one person from the founding team was employed by or was a student at a university, a binary code was used: one for at least one founder in the team at the creation time and zero for no member. To offset the potential negative effect on the early-stage fundraising ability of a spin-off created outside the university’s incubator, the study includes a dummy variable coded one if the spin-off was created inside the parent’s incubator and zero otherwise.

Validity and reliability

To reduce common method bias, previously validated measurements were employed (Spector 1987). In addition, a pilot test on five spin-offs from the University of Granada was conducted, which resulted in the survey avoiding potential question confusion on the part of respondents. An error can be generated using respondent self-reporting, particularly because many of our measures are complex and require post hoc assessment. To reduce this issue, Harman’s one-factor test was employed on all variables. The results suggest that the relationships revealed in this study among social network, capability and early-stage fundraising factors are unlikely to be caused by this common method bias. In addition, to avoid measurement errors, the study conducted proper survey measures and used a construct-validation test (the empirical indicators actually measure the construct) for validity (convergent and discriminant) and reliability. The results demonstrate that the study’s measurements are valid and reliable (see Appendix 1).

Results

Model estimation and fit

First, exploratory factor analysis (EFA) was used to construct the research indicators (see Appendix 2). The EFA results for the network structure model revealed that the item loadings were mostly significant (over 0.5). The four items that had loadings under 0.5 (trust, information quality and diversity, and strategy) were eliminated. EFA is not considered a sufficient means to evaluate the dimensions because it cannot test models with higher order factors (Rubio, Berg-Weger, and Tebb 2001). Therefore, in this study, we use first-order confirmatory factor analysis (CFA) to construct the lower order factors and second-order CFA to construct the higher order factors using the AMOS programme. The research employs CFA based on the maximum likelihood method to test the hypotheses because the normality test revealed that all of the observed variables have significant kurtosis and skewness p-values and the relative multivariate kurtosis is within an acceptable range (1.036). In addition, the sample size (181) exceeds the minimum requirement for the CFA (models with latent variables require at least 150 observations for normal distribution with no missing data; Muthen and Muthen 2002).

However, in a CFA model with fewer than 200 observations, a goodness-of-fit (GFI) test must be used (Barrett 2007). For this purpose, a combination of the ratio chi-square/degrees of freedom (CMIN/DF <3), RMSEA (<0.08), GFI (>0.9), NFI (0.9), and CFI (0.9) was employed to test the model (Ping 2004).

Before constructing our structural model, the average scores of eight first-order factors of social networks were estimated using all items identified in the first-order CFA of the structure, governance and content models. The first-order CFA results from the social network model revealed an acceptable fit, and all factor loadings (density, centrality, ties, reputation, reciprocity, trust, and quality and diversity of information) were significant at 0.01 levels (Table 1). The results also demonstrate that these structure, governance and content factors are valid and reliable (CR > 0.7 and AVE > 0.5 > SIC) indicators of the social network variable. Thus, these factors can be used as observed variables that construct the social network endogenous latent variable.

Table 1. First-order CFA of the social network model.

Paths	Loadings	CR	AVE
Network structure →		0.7678	0.5249
Density	0.756**
Centrality	0.739**
Ties	0.676**
Network governance →		0.7776	0.5416
Reputation	0.621**
Reciprocity	0.829**
Trust	0.743**
Network content →		0.7219	0.5650
Information quality	0.736**
Information diversity	0.767**

Notes: Model fit (CMIN/DF = 1.416, RMSEA = 0.048, NFI = 0.946, CFI = 0.980, GFI = 0.961).

^** Loading significant at the 0.01 level.

Second, we computed the average scores of the remaining seven first-order factors (technology, organizational viability, human capital, strategy, commercial resource, and follow-on and new investors) from the first-order CFA of capability and early-stage fundraising factors. By combining these factors with three social network variables, it was possible to construct a measurement model. The first-order CFA of the measurement model revealed an excellent fit (the ratio chi-square/degrees of freedom is less than two; RMSEA is less than 0.8; and all fit indexes are greater than 0.9; Table 2). In addition, the factor loadings are greater than 0.5 and significant at 0.01 levels, CR > 0.7 and AVE > 0.5 > SIC. The construct passes the validity and reliability tests, and thus, all constructs are adequate for use to test the research hypotheses.

Table 2. First-order CFA of the measurement model.

Paths	Loadings	CR	AVE
Social network →		0.9196	0.7930
Structure	0.904**
Governance	0.799**
Content	0.961**
Capability →		0.8436	0.5292
Technology	0.682**
Organizational viability	0.821**
Human capital	0.520**
Strategy	0.915**
Commercial resource	0.725**
Early-stage financing →		0.8894	0.8027
Follow-on investors	0.989**
Potential investors	0.792**

Notes: Model fit (CMIN/DF = 1.186, RMSEA = 0.032, NFI = 0.940, CFI = 0.990, GFI = 0.945).

^** Loading significant at the 0.01 level.

The result of the null model test revealed that the goodness-of-fit is not acceptable (CMIN/DF = 13.402). Therefore, the null model, in which no relationships are posited, is rejected. The analysis results of the hypothetical model (Figure 2) also reveal an acceptable goodness-of-fit (CMIN/DF = 1.324, RMSEA = 0.042, NFI = 0.931, CFI = 0.982, and GFI = 0.938). Thus, it is appropriate to test hypotheses 1, 2 and 3 with the research data.

Figure 2. Result model (**p < 0.01; *p < 0.05, all error terms omitted for clarity).

Hypothesis tests

Hypothesis 1 states that the social networks of founding teams positively affect the capabilities of the teams. The results indicate that the path between social networks and capabilities is positive and significant (0.291, p < 0.01). Therefore, hypothesis 1 is supported. The results reveal that the relationship between the social networks of a founding team and its early-stage fundraising ability is not significant (0.133, p > 0.05). Therefore, hypothesis 2 is rejected. Hypothesis 3, which states that the capabilities of a founding team positively influence its early-stage fundraising ability (0.142, p < 0.05), is also supported (Figure 2). To understand how a founding team can exploit its social networks to improve its capabilities and enhance its early-stage fundraising ability, the indirect paths of this model are analysed.

The results (Table 3) suggest that social networks are likely to exert indirect influences on all aspects of capabilities but fail to demonstrate the indirect effects of social networks on the follow-on investor and new investor factors. Consistent with hypothesis 1, social networks appear to positively and significantly influence capabilities with respect to technology (0.265, p < 0.01), organizational viability (0.320, p < 0.01), human capital (0.185, p < 0.01), strategy (0.362, p < 0.01) and commercial resources (0.362, p < 0.01). Capability appears to have significant positive indirect effects on the follow-on investor and new investor factors of early-stage fundraising ability (0.184, 0.196, p < 0.01; Table 3).

Table 3. Path analysis results: direct and indirect effects.

Paths	Standardized direct effects	Standardized indirect effects
Social network ↔ capability	0.198**
Social network → early-stage fundraising	0.166
Capability → early-stage fundraising	0.184*
Social network → early-stage fundraising		0.067
Social network → existing investors		0.067
Social network → potential investors		0.071
Capability → follow-on investors		0.184**
Capability → new investors		0.196**
Social network → entrepreneurial technology		0.265**
Social network → organizational viability		0.320**
Social network → human capital		0.185**
Social network → strategy		0.362**
Social network → commercial resource		0.362**
Control
Within incubator → early-stage financing	0.191

^* p < 0.05

^** p < 0.01; two tailed significance.

From these results, we construct a mediation model that considers the mediate role of a team’s capabilities between its social networks and early-stage fundraising. That is, founding teams exploit their social networks to improve their capabilities during start-up and subsequently enhance their early-stage financing activities (Figure 3).

Figure 3. Optimal model (All estimates are significant at the 0.01 level; all error terms omitted for clarity; Model Fit: CMIN/DF = 1.337, RMSEA = 0.043, NFI = 0.930, CFI = 0.981, and GFI = 0.936).

Control variables

In this study, all spin-offs were created by academic teams and received support from their universities. In addition, a spin-off’s creation location (i.e. in a university incubator) does not significantly influence its early-stage fundraising (Table 3). Thus, these control variables do not affect the analysis of the relationships among a founding team’s social network and capability and early-stage fundraising factors.

Discussion

This paper investigates the impact of the capabilities and social networks on the founding teams associated with the creation and development on early-stage university spin-offs on their financing. The literature has generally focused on new ventures (Zahra, Sapienza, and Davidsson 2006) and on the impact of the capabilities and social network associated with the new venture, not the founding teams (Walter, Auer, and Ritter 2006). This study is distinctive in its focus on university spin-offs, the use of teams as the unit of analysis and its scrutiny of the early-stage fundraising of university spin-offs in relation to founding team characteristics. The study posited that the capabilities and social networks of a founding team would be positively related to improvements in the early-stage fundraising ability of a university spin-off. This hypothesis was tested using survey data from 181 spin-offs of 35 universities in Spain. The results indicate that a founding team is likely to improve its capabilities by exploiting its social networks and that these improved capabilities can help a spin-off access early-stage financial resources. However, we could not find a significant direct relationship between the social networks of a founding team and the spin-off’s early-stage fundraising. Additionally, we found support for a mediating role of capabilities between social networks and a spin-off’s early-stage fundraising.

The ability to improve a founding team’s capabilities through the deployment of its social networks to support the development of a university spin-off is supported by research on new ventures per se (Tsai-Lung 2005). We suggest that a new venture’s relationships with various actors, for example, consultants, universities and other companies support the acquisition of technological knowledge. Deakins (1996) identified that information and knowledge received and learned from social networks also improve capability, which in turn helps enhance organizational viability. In addition, Yli-Renko, Autio, and Sapienza (2001) indicated that by exploiting business experience and market knowledge gathered from social networks, founders can increase their commercial resources to enable them to commercialize their products or services. Therefore, this paper indicates that as in other new ventures founding teams involved in university spin-offs can exploit social networks to improve their capabilities. Acknowledging this evidence, universities should support networking activities with industries through events, practical courses and research projects that involve academics and businessmen. These activities will stimulate the exchange of information and create relationships that benefit the future spin-off activities of universities.

Thus, this study agrees with the literature (Shane 2004; Vohora, Wright, and Lockett 2004a) in recommending that university spin-offs, like all generic new ventures, create founding teams that possess the necessary capabilities or can call on their wider social networks to enhance existing capabilities. To follow this recommendation, universities and policy-makers should develop and facilitate entrepreneurial communities that integrate academics, entrepreneurs, industry experts, the public sector and investors. We also suggest that these communities be established to share knowledge and experience as well as to discuss, identify and implement solutions to potential challenges in entrepreneurship.

The empirical tests reveal that the early-stage fundraising ability of university spin-offs can be improved by displaying the capabilities of the founding teams to investors as signals of value. This result supports the findings of Chen, Yao, and Kotha (2009), Rasmussen and Sørheim (2012) and Miloud, Aspelund, and Cabrol (2012) regarding the role of signals in the investment decisions of venture capitalists. These signals are used to convince investors that their investment will be profitable at an acceptable level of risk. These results support the findings of many scholars who have concluded that the resources of entrepreneurs improve the financing ability of new ventures (Chandler and Hanks 1998) and demonstrated the importance of human capital in the early-stage financing of new ventures (Brush, Edelman, and Manolova 2012). However, our results contradict the findings of Lindstrom and Olofsson (2001) who show that lower technological firms experience fewer problems in early-stage financing than technology-forefront ventures, and of Cassar (2004) who indicate that investors do not consider the education and experience of entrepreneurs to be financing preferences.

This study found that the social networks of founding teams during the creation phase do not generically directly relate to the early-stage fundraising ability of university spin-offs. This outcome partially contradicts the results of prior researchers (Heuven and Groen 2012; Rasmussen and Sørheim 2012) who emphasized the important role of the founder’s social networks in the early-stage financing of new ventures. These studies indicate that the founders of new ventures can quickly access public or private financial resources by utilizing their reputations (Mahto and Khanin 2013) and the strength of their network ties (Heuven and Groen 2012; Shane and Cable 2002; Shane and Stuart 2002).

Conclusion

Generally, this research underscores the important role played by the capabilities of founding teams in early-stage fundraising and recognizes the indirect influences of the teams’ social networks in decreasing the problems of uncertainty and asymmetric information in the fundraising processes of university spin-offs. Therefore, academic entrepreneurs are recommended to identify their existing abilities and to determine which capabilities they must improve to form capable teams that possess technology, management and industry knowledge by learning from or employing external resources. In addition, it is suggested that universities and economic development agencies become involved in activities that support the founding teams of university spin-offs to enhance their capabilities. Universities can encourage staff and students to improve their entrepreneurial and managerial skills through relevant seminars, conferences and additional courses. Universities and economic development agencies should also support spin-off activities by establishing incubators, institutions and mentoring boards to provide low-cost facilities, services (e.g. R&D, product development, marketing, recruitment, accounting, and legal assistance) and managerial advice.

The network-based entrepreneurship literature has primarily employed ego-network analysis, which focuses on network structure. This study adopts a more holistic view which analyses three dimensions of social networks: structure, governance and content. The results of the quantitative analysis demonstrated that our measurements are a valid and reliable means to determine the roles of social networks in an entrepreneurship process. Thus, this paper consolidates the validity of the network approach not only in entrepreneurship studies but also in network-based management research. By employing capability and social network theory in university entrepreneurship studies, this paper broadens the context in which this relevant theory can be applied. Current resource-based entrepreneurship studies have primarily focused on the capabilities of spin-offs. However, this paper has highlighted the important role of a founding team’s capabilities. The capabilities of a founding team, which include technology, human capital, organizational viability, strategy and commercial resources, contribute importantly to the early-stage investment decisions of external investors. In part, this influence occurs through exploiting the benefits of social networks, which over time contribute significantly to the capabilities of the founding team. This enhancement of existing capabilities through the exploitation of social networks facilitates signalling the value of university spin-offs. Thus, this paper enriches university entrepreneurship theory by identifying factors and processes that underpin the successful creation and development of university spin-offs.

While the study’s findings are robust, it is acknowledged that there are areas within the research process that could impinge on their validity and reliability. Compared with SEM requirements, this study’s sample size was restricted because of the limited number of spin-offs from Spanish universities. However, the sample represents 21% of all spin-offs created in Spain between 2003 and 2010. The survey is also based on a non-random sample. The respondents were selected based on their potential to provide a level of detail that could enhance our understanding of the phenomena in the judgement of OTRI officers in Spain. The data were collected using an Internet survey, which has a potential for misinterpretation. However, these issues were carefully examined during the pilot phase of the empirical research. It is also possible that survey respondents may exhibit cognitive bias based on post hoc rationalization. Our respondents were asked to comment on the constructs of capabilities and social networks of founding teams at start-up. However, they provided these evaluations at a later point during the spin-off’s development. To address this problem, we applied Harman’s one-factor test to all variables. The result indicated that this issue does not affect the study’s overall findings.

Appendix 1.

Validity and reliability

Convergent validity

We construct the CFA of 16 first-order factors: density, centrality, tie, reputation, reciprocity, trust, information quality, information diversity, technology, organizational viability, human capital, strategy, commercial resource, follow-on and new investors. These factors indicate five second-order variables: structure, governance and content of networks, capability and early-stage financing. The results revealed that both first- and second-order CFA of measurement models are acceptable fit, and each item loads on a single factor and is significant at 0.01 levels (Table A1).

To assess convergent validity, the extent to which the indicators of measurement converge to a high proportion of variances in common, we examine construct loadings and average variance extracted. The results from the first-order CFA of social network, capability and spin-off’s fundraising models reveal that all standardized loadings estimates are higher than 0.5 (Table A1). Moreover, all indexes of average variance extracted (AVE), the amount of construct variance relative to measurement error, are greater than 0.5 (Table A2) suggesting adequate convergent validity.

Discriminant validity

Discriminant validity (i.e. unidimensionality) is to test whether a construct is truly distinct from other constructs. The results revealed that all AVE estimates are larger than the corresponding squared interconstruct correlation estimates (SIC) (Table A2) inferring discriminant validity of the hypothesized structure is supported by our data.

Reliability

We compute the composite reliability, analogous to Cronbach’s alpha, of all first-order factors by the formula of Fornell and Larcker (1981). Most factors revealed sufficient composite reliabilities (above 0.70) except the reputation (0.632) and new investor factors (0.668) (Table A2). However, according to Hatcher (1994), the cut-off level of 0.6 is acceptable for a new conceptual variable. Thus, the measurements of this research are reliable.

Table A1. Factor loading of CFA

Social network

Reliving this spin-off’s creation period, evaluating these statements about relationships between your team and individuals, who you received advices or information related to process of your firm’s establishment, and among them (1: Not true … 7: Very true).

Table

Measures		First order loadings	Second order loadings
Structure
Density			0.769**
	Knowing each other by name	0.688**
	Talking to each other about business	0.941**
	Seeing each other regularly in business situations	0.933**
Centrality			0.797**
	We talked directly about business issues	0.67**
	We received directly helpful business information	0.712**
	We could call for advice about running our business	0.697**
	We were the first to receive new things in the group	0.781**
Ties			0.681*
	We would share personal matters with them	0.663**
	We might discuss family matters with them	0.917**
	We might ask them for advice about private matter	0.832**
Governance
Reputation			0.627**
	We generated a lot of enthusiasm	0.711**
	We had a forgiving nature	0.604**
	We persevered until the task is finished	0.742**
	We liked to play with ideas	0.775**
Reciprocity			0.755**
	People were generally pair in dealings with us	0.759**
	People were willing to do us a favour if asked	0.598**
	We did favours for each other from time to time	0.762**
	People patronized my business	0.87**
Trust			0.826**
	We were dependable by these people	0.888**
	People would say that we are sincere	0.917**
	They would say that we are trustworthy	0.604**
Content
Infor. Quality	Their information was usually accurate	0.878**
	Their information was relevant	0.916**
	Their information was specific	0.859**
	I quickly received their information	0.777**
Diversity Infor. (information used to be exchanged)	Market data	0.782**
	Product design	0.913**
	Process design	0.854**
	Marketing know-how	0.75**
	Packaging design/technology	0.744**

Notes: Structure model (CMIN/DF = 1.269, RMSEA = 0.039, NFI = 0.961, CFI = 0.991, GFI = 0.964).

Governance model (CMIN/DF = 1.149, RMSEA = 0.029, NFI = 0.950, CFI = 0.993, GFI = 0.963).

Content model (CMIN/DF = 1.288, RMSEA = 0.040, NFI = 0.973, CFI = 0.994, GFI = 0.965).

*Loading significant at the 0.05 level; **Loading significant at the 0.01 level.

Capabilities

Reliving spin-off’s creation period, evaluating these statements about what the founding team possessed (1: Not true … 7: Very true).

Table

Measures		First order loadings	Second order loadings
Technology			0.685**
	Hard to make a substitute for the technology	0.686**
	Our products might replace numerous existing one	0.78**
	Might replace other technologies in the industry	0.729**
	Potential to generate large economic returns	0.778**
	A platform for variety of commercial applications	0.598**
	Developed products with considerable demand in market	0.752**
Organizational viability			0.743**
	Team members were encouraged to improve working method	0.772**
	Team members had power to make decisions	0.770**
	Rewards and reinforcement were used	0.690**
	Individuals had time to incubate innovative ideas	0.600**
	Training in working techniques and attitudes was major emphasis	0.729**
Human capital			0.531**
	Good working experience	0.605**
	Good business management knowledge	0.767**
	Good industrial experience	0.712**
	Good entrepreneurial experience	0.856**
Strategy-making			0.923**
	Strong emphasis on R&D, technological leadership, and innovation	0.711**
	The first to introduce new products and services, administrative technologies, etc.	0.793**
	Strong tendency to be ahead of other competitors in introducing novel ideals and products	0.751**
	Strong tendency for high-risk projects with chances of very high returns	0.616**
Commercial resource			0.685**
	Building long-term customer relationships	0.605**
	Good plan to redesign management process	0.767**
	Good plan to redesign marketing and sales process	0.712**
	Focusing on customer satisfaction	0.895**

Notes: Model fit (CMIN/DF = 1.078, RMSEA = 0.021, NFI = 0.945, CFI = 0.990, GFI = 0.915).

*Loading significant at the 0.05 level; **Loading significant at the 0.01 level.

Early-stage financing

Describing how easy your new firm could access to these financial sources right after it was established (1: Much more difficult … 7: Much easier).

Table

Measures		First order loadings	Second order loadings
Follow-on investors
	Private investors/ angels	0.767**
	Venture capital	0.784**
	Government grants	0.693**
	Strategic partners	0.627**
New investors
	Initial public offering	0.741**
	Employees	0.685**
	Customers	0.711**

Notes: Model fit (CMIN/DF = 1.415, RMSEA = 0.048, NFI = 0.953, CFI = 0.985, GFI = 0.978).

*Loading significant at the 0.05 level; **Loading significant at the 0.01 level.

Table A2. Reliability and validity tests

Table

	Construct Reliability (CR)	Composite Reliability ^a	Average Variance Extracted (AVE)	Squared Interconstruct Correlation (SIC)
Social Network
Structure	0.7940		0.5634
Density	0.8949	0.888	0.7431	0.0751; 0.2025
Centrality	0.8076	0.736	0.5129	0.1475; 0.2052
Ties	0.8499	0.840	0.6576	0.0751; 0.1475
Governance	0.7825		0.5485
Reputation	0.8020	0.632	0.5054	0.1043; 0.1246
Reciprocity	0.8379	0.850	0.5678	0.1043; 0.3894
Trust	0.8523	0.879	0.6647	0.1246; 0.3894
Content	0.7220		0.5650
Infor. Quality	0.9182	0.926	0.7379	0.2767
Diversity infor.	0.9053	0.922	0.6580	0.2767
Capability	0.8427		0.5249
Technology	0.8668	0.839	0.5221	0.3204; 0.2927
Organizational viability	0.8384	0.794	0.5113	0.1069; 0.5083
Human capital	0.8279	0.808	0.5498	0.0320; 0.1069
Strategy	0.8109	0.702	0.5195	0.0600; 0.5083
Commercial resource	0.8135	0.708	0.5226	0.0841; 0.3881
Early-stage Fundraising
Follow-on investors	0.8108	0.724	0.5191	0.4045
New investors	0.7557	0.668	0.5079	0.4045

^aanalogous to Cronbach’s Alpha.

Appendix 2.

Means, standard deviation, ranges and correlations for variables in the measurement model

Table

	Variables
	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
(1) Density
(2) Centrality	.503**
(3) Ties	.294**	.445**
(4) Reputation	.087	.344**	.110
(5) Reciprocity	.297**	.643**	.351**	.400**
(6) Trust	.279**	.612**	.254**	.410**	.677**
(7) Info quality	.353**	.652**	.417**	.279**	.692**	.395**
(8) Info diversity	.371**	.496**	.517**	.036	449**	.318**	.564**
(9) Entrepreneur technology	.042	.161*	.109	.082	.182*	.171*	.149*	.215**
(10) Organizational viability	.070	.289**	.189*	.254**	.268**	.270**	.314**	.272**	.388**
(11) Human capital	.028	.156*	.150*	.136	.162*	.162*	.070	.201**	.190*	.393**
(12) Strategy	.050	.225**	.155*	.217**	.193**	.241**	.211**	.256**	.589**	.835**	.289**
(13) Commercial resource	.123	.160*	.035	.215**	.120	.134	.158*	197**	.553**	.558**	.333**	.729**
(14) Follow-on investors	.011	.167*	.012	−.024	.198**	.170*	.119	.119	.154*	.160*	.027	.186*	.154*
(15) New investors	.023	.108	−.006	.003	.119	.145	.095	.055	.178*	.106	−.061	.159*	.095	.784**
Mean	4.03	4.83	3.51	3.53	5.82	5.68	4.03	3.45	5.58	5.76	5.10	5.14	5.63	2.01	2.39
SD	1.77	.89	1.63	.36	.89	.69	1.01	1.27	1.13	.97	1.50	.90	1.25	.93	1.15
Min	.30	1.95	1.08	2.25	2.82	2.74	.83	.82	1.77	2.43	1.66	1.67	1.63	.89	1.00
Max	6.78	6.13	6.14	4.08	7.05	6.39	5.74	6.69	7.32	7.53	8.25	6.81	8.06	4.41	5.44

^*Correlation is significant at the 0.05 level (2-tailed).

^**Correlation is significant at the 0.01 level (2-tailed).