The power of syndication: The innovation output of venture capital investments in the United States

Shuwaikh, Fatima; Dias, Joakim

doi:https://doi.org/10.59876/a-6r7m-ahrf

To improve the management of their funded companies and to mitigate risk of investments, Venture capital (VC) investors follow the syndication strategy (Dimov and Milanov, 2010; Manigart et al., 2006). In VC industry, much investment activity takes place in syndicates (Braune et al., 2019) expecting a joint payoff (Keil et al., 2010). A syndicate describes a group of VC investors who together back their companies (Gompers and Lerner, 2000) at different points in time or the same investment round (Brander et al., 2002). Syndication raises the likelihood of the successful exits of companies (Jääskeläinen, 2012) and builds value for them by aggregating networks, expertise, and different skills (Tian, 2012). Receiving funds from various investors carries a clear sign of the companies’ quality (Dai and Nahata, 2016). Jääskeläinen (2012) illustrated that syndication is path that aids companies know possible acquirers. Nevertheless, VC investors have various objectives and incentives (Brander et al., 2002). Different types of VC investors have distinguishing favorites for exit strategies (Bertoni et al., 2013) and the innovation of the companies (Shuwaikh and Dubocage, 2022).

Rather than investing alone, VC investors join together to provide superior returns for independent venture capital (IVC)[1] and corporate venture capital (CVC). Given the strategic performance advantages of CVC investors, it is not surprising that IVC investors are more likely to invite corporate investors to syndicates. Building upon the relational view of strategic management, investors enter interorganizational relationships to gain competitive advantage by accessing firms’ unique resources (Eisenhardt and Schoonhoven, 1996). The unique resources of CVC investing firms explain how CVC investors are able to be attractive to IVC investors and inter the VC syndication. Sorenson and Stuart (2001) and Wright and Lockett (2003) argue that VC investors might choose their investors to syndicate with based upon the unique resources that these investors hold.

Current literature argues on what type of VCs offer the most value for these companies in terms of achieving an innovative high output. Some scholars say that CVC offers an increased worth to their entrepreneurial companies (Alvarez-Garrido and Dushnitsky, 2016) while others state that CVC just do not offer additional benefit compared to IVC (Bertoni et al., 2013). Additionally, there are times when VCs are motivated to create a syndicate to achieve better future deal flow or diversification (Manigart et al., 2006). When a VC firm enters a business, it usually has the capability to transform it using its sector experience, knowledge, and other resources (Gompers and Lerner, 2000). In this paper we test how the participation of different types of VC investors impact investor influences the innovation output of their ventures. Specifically, we examine the power of syndicates that optimally nurtures innovation for entrepreneurial companies while considering possible mechanisms as the location fit to their entrepreneurial companies and the industry fit between the VCs and the funded company.

To test our hypotheses, we gathered a sample of 5182 US entrepreneurial companies that received funding from both CVC and IVC investors during 1998 and 2017. We explore how the relative influence among a heterogeneous group of VC investors (CVC and IVC) in a syndicate affects the likelihood of the entrepreneurial company’s successful exit and innovation output. When a syndicate consists of multiple types of VC investors with different incentives and objectives, company’s performance largely depends on which investor’s demands the company is serving. We focus on entrepreneurial companies funded by syndicates composed of both IVC and CVC investors. With whom a VC firm will form a value creation syndication is the question we focus on in this study. This research finds consistent support for the benefit of CVCs in syndicates. It also highlights the vastly positive impact that is created when investors and entrepreneurial companies operate within the same industry as well as location. For practitioners in the VC industry, we highlight the role and value of unique resources provided by IVC and CVC investors in VC syndicates.

Our study aims to contribute to the literature through three essential approaches. First, our study contributes to the best structure for nurturing innovation in entrepreneurial companies (Alvarez-Garrido and Dushnitsky, 2016; Ma, 2020). Lerner (2012), in describing the architecture of innovation, shows that the available corporate model may not be the best organizational structure for nurturing innovation. Additionally, he notes that, while IVC firms may have numerous innovative ideas, they have only performed well in some targeted businesses (Chemmanur et al., 2014). Increasingly, corporate investors are proving to be an essential alternative to IVC as outlined by Dushnitsky (2012). Lerner (2012) suggests that the best design to drive innovation is possibly a “hybrid” model, such as a CVC program that merges characteristics of backed companies with those of corporate research laboratories “within a powerful system that consistently and efficiently produces new ideas”. So, we respond to the call by Lerner (2012) and we demonstrate the power of the CVC fund in value creation.

Second, we participate in the growing discussion in the literature on entrepreneurial finance about the CVC outcomes (Braune et al., 2019; Ma, 2020). Previous studies have shown risks associated with CVC, and the distinctive value added that the CVC provide (Dushnitsky and Shaver, 2009). Our paper links CVC investment to value creation in the innovation output, showing that the CVC participation change the VC syndication output, which has been associated with higher innovation performance of new ventures. Our study respond to the call by Basu et al., (2011) to further study relational antecedents of CVC investment relationships and contributes to research on industry fit (Cefis et al., 2020; Howell, 2020) and location fit (Cumming and Dai, 2012;Tian et al., 2020) between new ventures and incumbent firms more broadly.

Third, our empirical conclusions bring new insights to the theoretical literature on the role of VC syndications in fostering innovation and corporate innovation (Braune et al., 2019; Ferrary, 2010; Keil et al., 2010; Tian, 2012). The variety of syndicate members is also positively correlated with the likelihood of the successful exits of the entrepreneurial companies due to expanding the diversity of the range of the resources given to the entrepreneurial companies (Colombo and Murtinu, 2017). Additionally, knowledge exchange between syndicate members is essential for improving entrepreneurial companies innovation performance (Anokhin et al., 2011). Hence, the behavior of VC investors in a syndicate increases its value creation potential (Braune et al., 2019; Tian, 2012). Literature focuses on importance of syndication and Dai and Nahata (2016, p. 140) point out “syndicate formation itself has received much less attention.” We respond to the call of Dai and Nahata’s (2016), and we address how the participation of a CVC affect investor concentration in syndicated investments. This paper reveals that CVC participation supports a company’s innovation output as long as CVC investors are in the syndicate.

This paper is structured in the following manner. Section 2 presents the literature review followed by the methodology in section 3, which explains the data collection and treatment process as well as the methods used. Section 4 expresses the results from this analysis which will, later on, be interpreted in the discussion section. Lastly, there is a conclusion of the research made and an insightful avenue for future research.

Literature review and hypothesis development

CVC Complementary resources

As investors do not have to work by themselves with their skillset as a limitation, a business association between CVCs and IVCs can be created—VC Syndicate. This is a type of alliance (Wright and Lockett, 2003) where they will take a joint-equity stake in a venture and work together in an attempt to maximize their potential to try to create additional value/innovation for the venture as well as post-IPO operating performance (Tian, 2012). It allows the syndicate members to not only pool their resources (Ferrary, 2010) but also to diversify risk (De Clercq and Dimov, 2004). According to Ferrary (2010), the motivations for creating a syndicate is two-fold. First, it is to diversify the risk during the seed stage, which is quite uncertain. Secondly, it is to create a heterogeneous community within that investment. On the other side, however, the VC syndicate may reveal some issues like the lack of dynamic stability or a dominant party within the decision making (Wright and Lockett, 2003). Due to the heterogeneity of the type of investors, one can wonder how it will impact entrepreneurial companies with their different skill sets. This diversity is originated all from the complementary assets and know-how (Gompers and Lerner, 2000; Maula et al., 2005; Dushnitsky, 2012) to the managers’ compensation (Gompers and Lerner, 2000) and strategy formulation (Maula et al., 2005). With this paper, the path that nurtures most innovation and value creation will be apparent in VC syndication.

VC firms syndicate their investments for a variety of reasons and the competitive advantage of complementary resources is one of the main motivation for such syndications. Wernerfelt (1984) presents the term “resource-based” and outlines that firms are groups of resources rather than sets of product-market positions. In a related vein, Rumelt (1984, pp. 557–558) adds “a firm’s competitive position is defined by a bundle of unique resources and relationships and that the task of general management is to adjust and renew these resources and relationships as time, competition, and change erode their value.” Barney (1991) describes that the key idea that drives the resource-based view (RBV) is the center for the competitive advantage of a firm which is related to tangible and intangible resources, competencies and specific skills of the firm. In addition to focusing on the internal resources of the firm, the RBV has been implemented in the examination of inter-organizational relationships of firms (Hamel et al., 1989). Stein (1997) explains that in the RBV, the primary argument for alliance structure is that firms decide to build a proper advantage in inter-organizational relationships by leveraging better resources they own with complementary resources. Das and Teng (2000) employ the RBV in the function of resource complementarities which leads to performance and the alliance formation. In line with this argument, we argue that VCs choose their syndication partners based upon different resources hold by these partners (De Clercq and Dimov, 2004; Manigart et al., 2006).

Alkhanbouli et al. (2020) summarize how a business creates and captures value. They illustrate that an innovative business model refers to the innovation of one or more parts of the business (investor-venture). IVCs invest in companies that have important technology, industry models, and operational risks to achieve financial gains. To that end, IVCs initiate a dedicated fund in the form of a limited liability partnership, building funds from third parties or limited partners, such as pension funds, high net worth individuals and families (Alvarez-Garrido and Dushnitsky, 2016). IVCs give their backed companies various added value assistance, such as networking managers with investors and possible acquirers, staff recruitment and support for plan formulation (Sapienza, 1992). IVCs invest nearby as a requirement of personal interaction to carry out scouting and nurturing (Sorenson and Stuart, 2001). The CVC investors are typically tightly connected with their backed companies by providing it access to many corporate resources (Alvarez-Garrido and Dushnitsky, 2016). Usually, CVC funds are co-invested along with IVC funds (Ferrary, 2010). For instance, large pharmaceutical firms own complementary resources along the industry value chain, including experimental labs and experimentation equipment, in addition to clinical examination sites (Friedman, 2006). Keil et al. (2010) find that CVC investing usually leverages access to complementary resources and organizational personnel to help and advance the progress of companies.

Alvarez-Garrido and Dushnitsky (2016), outline that entrepreneurial companies can use organizational laboratories and research tools. These resources can be utilized to support, advance and examine additional encouraging inventions, and the advantages are not restricted to the application of material support. The CVC program can leverage the competences, capacities and assets of the firm, which, in turn, promotes and facilitates the evolution of the company (Block and MacMillan, 1993). Before considering creating a syndicate and its potential partners, VCs decision is dependent on their developed investment strategy and the intrinsic characteristics of the VC itself (De Clercq and Dimov, 2004). Taking into consideration that CVCs are supposed to boost the resources while reducing risk (Ferrary, 2010), it is hypothesized that, CVC/IVC-backed entrepreneurial companies experience higher innovation outputs. We assume that CVC investors and, more specifically, their complementary resources influence a company’s innovation output. As a result, entrepreneurial companies experience distinct degrees of innovation due to being funded by venture capitalists with various complementary resources characterizations. Based on these insights, we expect the CVC/IVC-backed companies to be more innovative once they are able to access and leverage the complementary resources of the firms providing the CVC funding.

Hypothesis 1 — The presence of CVC investors in a syndicate positively impacts the innovation output of CVC/IVC-backed entrepreneurial companies.

Location fit

Location fit enables more interaction between the parties involved creating a more personal relationship, allowing a smoother due diligence process, operational assistance, and relationship nurturing (Sorenson and Stuart, 2001). This location fit has particular importance when it comes to the CVCs complementary assets. These complementary assets are any non-financial benefit a corporation can share with their investments. Using the example of Alvarez-Garrido and Dushnitsky (2016) while studying this effect on biotech ventures, both laboratories and scientists are complementary assets of paramount importance for these ventures. The need for R&D and other necessary facilities for the ventures are both expensive and scarce, giving a competitive advantage to the CVCs. The emphasis on the impact that complementary assets have on the output for innovation (Chemmanur et al., 2014) highlights the importance of location fit. Complementing the previous statement, the probability of a breakthrough for researchers is higher once the parties are neighboring each other (Catalini, 2018). The opposite is also true, however, where distant ventures access minimal resources, making the CVCs lose their main competitive advantage (Chemmanur et al., 2014). Therefore, taking into consideration the potential boost of the CVCs, it is expected that location fit will contribute to a higher innovation output for the CVCs.

Polanyi (1967) describes tacit knowledge as “knowing more than we can tell”. Fundamentally tacit knowledge is observed through an individual’s activities according to Polanyi (1967) which cannot be seen through precise descriptions of what an individual knows. Scholars indicate under the lens of the knowledge-based view that tacit knowledge can form the foundation of a sustained competitive advantage because tacit knowledge is almost immobile and complex to imitate (DeCarolis and Deeds, 1999). Location fit proximity facilitates the profitability of tacit knowledge once the partners engage in the innovation process. Based on these insights, we expect to find the location fit as a dimension that enhances the innovation production of the CVC-backed companies. Catalini (2018) observes when the partners have location fit, the possibility of creating an invention is more significant, and as a result, the company can benefit from the firm’s complementary resources. Fleming et al. (2007) explain that research is mysterious and concerns different reconstructions of ideas because it demands repeated interpersonal communication, which is facilitated by location fit. We expect that syndications with CVC investors where there is location fit between the CVC investors and the funded companies will experience more innovation production than solely IVC-backed companies.

Hypothesis 2 — The location fit between the CVC investors and the funded companies positively moderates the innovation output of the CVC/IVC—backed entrepreneurial companies.

Industry fit

The industry compatibility between the VC investors and the ventures can impact the innovation output as they have specialized expertise (Gompers et al., 2009) like market knowledge and sector connections. Sometimes, parent corporations try to outsource their internal projects to increase the speed of innovation and competitive advantage (Fulghieri and Sevilir, 2009). Consequently, to accomplish this, they outsource through ventures with the same sector-specific goals. Compared with IVCs, they are sometimes focused on a specific sector to capture this beneficial innovation and compatibility (Anokhin et al., 2011). There is opposing literature regarding who takes the most benefit from the industry fit. Due to the resources of the corporate parent, CVCs have access to a competitive advantage, which is sector-specific expertise (Chesbrough, 2002). In contrast, the efficient resource allocation, compensation schemes, and venturing experience with IVCs make them better at nurturing innovation (Gompers et al., 2009). To that end and as Fulghieri and Sevilir (2009) demonstrate, the firms change their plans to move from an internal to an external direction.

Investors who seek to implement policies via portfolios need to adhere to the industry fit in their investments, which is of essential importance for their success. Industry fit is regarded as an essential condition that supports all other success criteria. Incremental learning is supported through the sharing of the essential characteristics of related industries (Cefis et al., 2020). Industry fit can be operationalized using many aspects that the sectors offer, including the customers they serve, the products and services the industries provide, their methods of organizing operations, the technologies they use, and executive decisions driven by the cognitive framework (Silverman, 1999). Industry fit should allow the investing firm and the company to exchange knowledge more efficiently (Howell, 2020). The similarities in cognitive frameworks and the fit in the business logic mean that partners look for new knowledge in similar places and are affected by the same aspects of the future evolution of technologies and markets. We expect more knowledge refining and building and incremental learning with more industry fit. We aim to examine whether the industry fit between the entrepreneurial company and the corporate investor is a mechanism that impacts innovation output. In particular, we test the CVC/IVC-backed companies’ innovation production and whether industry fit better contributes to these outcomes. We assume that CVC/IVC-backed entrepreneurial companies with industry fit will experience more innovation outcomes than will solely IVC-backed companies.

Hypothesis 3 — The industry fit positively moderate the innovation output of the CVC/IVC—backed entrepreneurial companies.

Methodology

Sample and data collection

We construct an unbalanced panel of 5182 U.S. companies between 1998 and 2017. Our sample is IPO backed companies to have access to their financial data and ownership information. We choose 1998 as the start of our study period because it corresponds to the increase in CVC investments among venture capitalists and companies. We focus on companies funded by two leading venture capitalists: IVC and CVC. We exclude all companies funded by fewer than two IVC investors in order to consider the syndication dynamics between CVC and IVC investors. We identify CVC investors with a unique corporate parent. From this entire dataset, 924 companies are funded by CVC and IVC syndicates while 4258 companies are funded by IVC investors only. We obtain complete investment information from VentureXpert, ThomsonOne database, an extensive database on venture investments that has been broadly applied in prior studies (Wadhwa et al., 2016). To obtain the necessary financial and accounting information, we use Standard and Poor’s Compustat database. The United States Patent and Trademark Office (USPTO) database is used to obtain patent application information. To capture the time of knowledge creation, we use the application date of the granted patent, we end the sample at 2018 because of the lag between the patent application date and its granting. Finally, Datastream, our fourth database, is used to find the ticker symbols, which help us obtain data from Compustat because of the differences in the names of companies in the databases. To combine the data from the Thomson One, USPTO, Compustat and Datastream databases, the names of the companies were hand-matched in all the databases. We obtain our final sample at the company level.

Dependent variables

Citations and Patents: It is the natural logarithm of the patent count for company (i) at year (t), ln (Patents_i,t), is our first proxy that captures the quantity of innovation output. Particularly, the variable ln (Patents_i,t) includes the number of patent applications filed in the granting year (Zhang et al., 2019). We build the second proxy, ln (Citations_i,t) which indicates the patent quality by including the number of citations collected by each patent (Chemmanur et al., 2014). Following the literature, we create patent variables based on the year of patent application because it is nearer to the event of the exact innovation, which is outlined by Griliches (1990).

Independent variables

To evaluate the effect of the presence of CVC funding, we use three measures for the degree of CVC participation in the syndication.

CVC funding: The variable CVC funding is a dummy variable taking the value of one when the syndication has CVC funds in the syndication (CVC/IVC) and zero if the company is funded by IVC investors only. The objective with this variable is to analyze whether any extra impact is created depending on if it is a CVC and IVC syndicates or IVC syndication only.

CVCs Number: Number of CVC investors is a variable which count the number of CVCs in the VC syndicate.

% CVC: The variable percentage of CVC measures the percentage of CVC investments in the VC syndicate.

Location Fit: The purpose of the development of the Location Fit variable was to access the location aspect in order to understand if geographic fit fosters firms’ innovation output if financed by CVC investor. This variable is a dummy variable meaning it is one when the funded company’s location matches the CVC investor’s location and zero otherwise. Since some states are vast in the US, it was decided that this dummy would be developed using metropolitan statistical area (MSA) (firm, fund, company). This way, the CVC investing firm, fund and the funded company had to have their headquarters in the same city. This is most notable when looking to California, where several firms are based in San Jose or San Francisco. However, despite being in relative fit, it would be considered as a different location.

Industry Fit: This variable was created to track the sector compatibility between the backed company and the CVC investor. This variable was build based on the Standard Industrial Classification (SIC) codes of both parties (Shuwaikh and Dubocage, 2022). The SIC code is a 4-digit number that indicates the sector and business activity in which the company operates. The first two digits represent the major group the company works in while the first three digits represent the sector group, and the whole four digits combined represent their division. The SIC codes for the entrepreneurial companies were taken from the DataStream. Simultaneously, for the investors, the information was extracted with web-scabbing mechanisms and the EDGAR system from SEC. This index takes the value of 1 in the case of a matching SIC code and zero in the case of an entirely different SIC code. In prior research, very similar measures have commonly been used to measure industry fit. We divide firms into 17 industries using the classification given by Kenneth French on his web site (Rosen, 2006).[2]

Control variables

Following the innovation literature (Chemmanur et al., 2014; Shuwaikh and Dubocage, 2022) it is relevant to control the variables that impact the innovation output. Our control variables that are considered are: ASSETS, R&D/ASSETS, REVENUES/ASSETS, LEVERAGE, CAPEX/ASSETS, ROA and Tobin’s Q when regression with patents or citations. Furthermore, as literature has shown, including Chemmanur et al. (2014) and Alvarez-Garrido and Dushnitsky, (2016), the backed company’s age also plays a significant role in their outcome and growth, so we add Ln (Age at IPO).

Results and discussion

Descriptive statistics

Table 1 reports the descriptive statistics of the variables in our sample. Panel A presents the innovation output of IPO backed companies. The average patenting of the company is (3.59) patents per year. After breaking down our sample into sub samples IVC- and CVC/IVC-backed companies we find that CVC/IVC backed companies have higher number of patents (5.13) patents per year while it is (2.75) patents per year for IVC-backed companies. Additionally, the average citations the company is receiving (3.39) citations per patent. CVC/IVC backed companies have higher number of citations (4.31) citations per patent while it is (2.89) citations per patent for IVC-backed companies. Panel B reports summary statistics of the control variables. The average assets of the IPO backed company is $121.17 million, ratio of R&D to assets is 11%, revenues 34%, leverage 11%, CAPEX 5% and ROA of -2%.

Regression analysis

Our study aims to compare the innovation output of IVC- and CVC/IVC-backed companies. We begin to examine the innovation output of pre-IPO IVC- and CVC/IVC-backed companies. We consider a cumulative innovation over three-year period prior to the IPO date. To evaluate the effect of the presence of CVC funding, we use three measures for the degree of CVC participation in the syndication. CVC funding is a dummy equal to one if the CVC fund is participating in the syndication and zero if the company is funded by IVC investors only. CVCs Number, which count the number of CVCs in the VC syndicate.% CVC measures the percentage of CVC investments in the VC syndicate. Table 2, Panel A presents the baseline findings of the dependent variable which is the total number of patents. Model 1 highlights the positive and significant impact that CVC funding add to the syndication. Economically and based on the coefficient estimate of CVC funding in the three years prior to IPO, CVC/IVC-backed IPO company experience 27% more patents than IVC-backed IPO companies. Models 2 introduce the number of CVCs in the VC syndicate. Based on the coefficient estimate of CVCs Number, the patenting output increases by 16% for one additional CVC investor in the syndicate. In the same vein, Model 3 reports the significant impact of the percentage of CVC investments in the VC syndicate which positively impact the patenting output. Panel B demonstrate the results of the quality of the innovation output of the backed companies in the three years prior to IPO. The three variables evaluating the impact of the presence of CVC investors in the syndication are significant. Model 4 highlights the positive and significant impact that CVC funding add to the syndication. Economically and based on the coefficient estimate of CVC funding in the three years prior to IPO, CVC/IVC-backed IPO company experience 18% more citations compared to IVC-backed IPO companies.

Table 3 reports the analysis of the innovation output of post-IPO IVC- and CVC/IVC-backed companies. the dependent variables patents and citations are based upon the innovation output of the IPO years and the three years following the IPO. Regarding the quantity of the patents, Panel A suggest that CVC/IVC-backed companies have higher innovation output tested by the patents post IPO. Model 1 highlights the positive and significant impact that CVC funding add to the syndication. Economically and based on the coefficient estimate of CVC funding in the four years post IPO, CVC/IVC-backed IPO company experience 55% more patents than IVC-backed IPO companies. Models 2 introduce the number of CVCs in the VC syndicate. Based on the coefficient estimate of CVCs Number, the patenting output increases by 32% for one additional CVC investor in the syndicate. In the same vein, Model 3 reports the significant impact of the percentage of CVC investments in the VC syndicate which positively impact the patenting output. Panel B demonstrate the results of the quality of the innovation output of the backed companies in the four years post IPO. The three variables evaluating the impact of the presence of CVC investors in the syndication are significant. Model 4 highlights the positive and significant impact that CVC funding add to the syndication. Economically and based on the coefficient estimate of CVC funding in the four years post IPO, CVC/IVC-backed IPO company experience 24% more citations compared to IVC-backed IPO companies.

Table 2

**Pre-IPO innovation output of IVC and CVC/IVC backed companies**

Table 3

**Post-IPO innovation output of IVC and CVC/IVC backed companies**

Table 4 reports an additional analysis comparing pre- and post-IPO innovation output of our sample. This is an analysis of VC investments round by round at the funding stage. We add control variables that take each stage into consideration. It has been shown that earlier-stage VC-backed companies are expected to be more likely to achieve a successful exit (i.e., acquisition or IPO) than later-stage VC-backed companies are (Nahata,R., 2004). Espenlaub et al. (2012) further argue that the time to exit is quicker for companies funded at later stages than for companies funded at earlier stages. We control for the impact of the stage of the company when the first funding occurs by including an investment stage dummy variable. Startup/seed investment stage is a dummy variable that is equal one 1 if the funded company is at the seed stage when receiving the first investment and is 0 otherwise. Early investment stage is a dummy variable that is equal one 1 if the funded company is at the early stage when receiving the first investment and is 0 otherwise. Expansion investment stage is a dummy variable that is equal one 1 if the funded company is at the expansion stage when receiving the first investment and is 0 otherwise. Later stage is a dummy variable that is equal one 1 if the funded company is at the later stage when receiving the first investment and is 0 otherwise. Panel A and B report the results of the analysis of the CVC/IVC investments syndication. We consider the first investments rounds. We find a positive and significant impact at the 1% level for all funding stages.

Table 4

**innovation output of IVC and CVC/IVC backed companies (Investment stage)**

Univariate comparison approach

As our analysis is focusing on the IPO-backed companies, we address this concern by taking into consideration now the full sample of all VC-backed companies based on their current status and the exit outcomes. Table 5 presents the univariate comparison of innovation output of CVC/IVC-backed companies and IVC-backed companies taking the dependent variable number of patents. For all VC-backed companies, the difference for the patents output between CVC/IVC-backed companies and IVC-backed companies is statistically significant at the 1% level with 0.64 patents for CVC/IVC syndication. Then, based on the current status and the exit outcomes, we break down the sample into four categories. The first category is the active status of the VC-backed companies. The CVC/IVC-backed companies have higher innovation output of 0.35 patents than solely IVC-backed. The second category is the IPO based exit and the CVC/IVC-backed companies generate an average of 2 patents more than IVC-backed companies. The third category is the acquired based exit and the CVC/IVC-backed companies generate an average of 1 patent more than IVC-backed companies. The fourth category is the written-off companies and the CVC/IVC-backed companies generate an average of 0.37 patent more than IVC-backed companies.

Table 5

**Innovation by all VC-backed companies**

Location fit and industry fit mechanisms

Our analysis demonstrates so far that the presence of CVC investors in the syndication improves the innovation output of the backed companies. We move now to discuss the possible mechanisms that allow the syndication CVC and IVC investors to better improve the innovation output of their funded companies. Moving on to Table 6, we address the Location Fit and Industry Fit. Panel A, Model 1 reports the impact of CVC funding on two groups of backed companies with and without location fit. The first group is CVC with location fit where the companies receive CVC funding and have geographic proximity with the corporate investor. The second group is the companies that receive CVC funding and do not have geographic proximity with the corporate investor. We find that the gap in patenting outcomes between CVC-backed companies with geographic proximity and CVC-backed companies without geographic proximity is responsive to being located close to the CVC fund. As a result, CVC-backed companies with geographic proximity are more innovative than CVC-backed companies without geographic proximity and are similar to IVC-backed companies in the absence of geographic proximity. Our findings are line with Shuwaikh and Dubocage (2022) and Catalini (2018), who finds that when the partners have geographic proximity, the possibility of creating an invention is more significant, and as a result, the company can benefit from the firm’s complementary resources. This result implies that access to CVC investors’ R&D support is an essential channel by which CVC investors boost companies’ innovation output.

Table 6

**Innovation output of IVC and CVC/IVC backed companies (location and industry fit)**

Model (2) examines the impact of the CVC-backed companies regarding industry fit with the CVC investors. Model (2) splits CVC-backed companies into two groups. The first group is the companies that receive CVC funding and have industry fit with the CVC investors, while the second group is the companies that receive CVC funding and do not have industry fit with the CVC investors. The coefficient of CVC with location fit is positive and significant, whereas the coefficient of CVC without business similarity is not significant. Findings illustrate that CVC with industry fit boost the innovation output by 36% which is once again supported by the citation results of 46%. Additionally, Models 3 and 4 present the analytical results with regard to the second proxy of innovative outcomes, citations. The results are compatible and echoes with those in Model 1 and 2. This complements previous literature that mentioned the benefit of CVC funding (Chemmanur et al., 2014; Guo et al., 2015; Shuwaikh and Dubocage, 2022) regarding the positive benefits of complementary assets and strategy formulation of CVC investors. These results could even foresee a somewhat proportional relationship between the numbers of investors and their innovation output. Therefore, CVC investors in the syndication boost innovation compared to solely IVC syndications.

The location fit between entrepreneurial companies and their investor is proven as beneficial relating to the stimulation of innovation. Panel B Model 1 indicates that when both parties are within the same metropolitan area, the entrepreneurial companies’ innovation output is increased by 27%, supported by the citations’ analysis with 33%. This backs the theory developed by Alvarez-Garrido & Dushnitsky, (2016) regarding all the perks within complementary assets such as operational assistance and facilities, for example. Therefore, matching the findings by Catalini (2018) and Shuwaikh and Dubocage (2022) regarding the positive impact of Location Fit.

In the following Models 2 and 4 of Panel B, we can see further support of the positive impact that the industry fit attributes to the funded companies along with their syndicated investors. Combining the syndicates with the industry fit, it presents evidence that when syndicates are created, the location plays a paramount role in the innovation output of the funded companies. This suggests that in terms of innovation productivity, CVC investors have a higher yield in innovation once joining the syndicate groups. Complementing with the industry pillar, Panel B model 2 demonstrates that when both the funded company and the investor share the same sector, it enhances the innovation output by 39%. Once again, it is supported by the citations Model 4 with a contribution of 48%. This industry-specific expertise, like market experience and connections, gives an additional edge for the CVCs (Gompers et al., 2009).

Propensity score matching analysis

Table 7 presents the difference in the observable characteristics of IVC- and CVC/IVC-backed companies. Our analysis aims to document the difference in the innovation output between IVC-backed companies and CVC/IVC-backed companies. We use the propensity score matching to allow us disentangling the treatment and the selection of CVC funding on the innovation output of backed companies based on their observable characteristics. These are the selection and treatment where corporate investors select companies with higher innovation output and CVCs have superior ability to nurture innovation. To minimize the selection bias based on observable, we proceed with the propensity score matching to mitigate these observable selection effects.

Using the propensity score matching, we follow the approach first developed by Rosenbaum and Rubin (1983) to apply the nearest-neighbor matching implementation. Based at the IPO company level, our dependent variable is the CVC/IVC-backed companies being a binary variable equal one if CVC funding join the syndication and zero for solely IVC syndicates. Our control variables are matched at the IPO year measurement. To absorb any industry and time heterogeneity not captured by company characteristics, we add the year and industry fixed effects. Table 7, Panel A presents the univariate comparison between IVC- and CVC/IVC-backed companies. We find that CVC/IVC-backed firms have less fixed assets, have higher Tobin’s Q, have higher R&D, less profitable and are larger. Panel B reports the probit model of funded companies containing non missing data for all explanatory variables. Prematch results are reported through Model 1 to Model 3. We then perform a nearest neighbor propensity using the propensity score from the Prematch. The Prematch column contains the parameter estimates of the probit estimated on the entire sample, prior to matching. This model is used to generate the propensity scores for matching. As our main control group companies, we use the three nearest neighboring IVC-backed companies that come from the same industry year IPO. This is due to that the number of IVC-backed companies are more than CVC/IVC backed companies. To assess the accuracy of the matching procedure we conduct diagnostic tests (Chemmanur et al., 2014). Models 4 and 5 presents the results of the univariate comparison between IVC—and CVC/IVC—backed companies for the matched pairs. The results are no more significant between IVC- and IVC/CVC-backed companies. The Postmatch column contains the parameter estimates of the probit estimated on the subsample of matched treatment and control observations after matching. The magnitude of the probit regression coefficient of the postmatch restricted to the matched sample decline dramatically as reported in Model 7 in Panel B. In both diagnostic tests, we are unable to match the ROA. We find that CVC/IVC backed companies are less profitable that are IVC-backed companies.

Table 7

**Propensity score matching: Diagnostic tests**

Conclusion

This paper demonstrates the impact of the participation of CVC in VC syndicated investments. As CVCs and IVCs bring various potential liabilities and value-added to the new venture, a closer look at the effect of the CVC and IVC engagement on the syndicate can be helpful to entrepreneurs and investors. We examine the dynamics of the impact of these syndications and the possible channels (industry fit and location fit) that improve innovation effectiveness. We provide a more comprehensive picture of venture capitalists ’ behavior by displaying when CVC and IVC are involved in syndications. Entrepreneurs should take into account the potential effect of IVC or CVC on syndicate partners’ investment performance when they get funding. The best strategy for companies interested in using VC investments to improve the outcomes of their innovation efforts is to consider CVC/IVC syndications besides location and industry fit.

Our results have several practical implications. For corporate investors, our conclusions propose that the competitive advantage of the complementary resources of the parent corporation has essential value for syndication. To leverage these resources, we discuss the importance of two channels: location fit and industry fit. For independent venture capitalists, our results highlight the differential value that corporate investors contribute. Syndicating with corporate investors provides value creation with partner venture capitalists through access to these unique resources. For practitioners in the VC industry, we highlight the role and value of unique resources provided by IVC and CVC investors in VC syndicates. As with all studies, our study is not without limitations. A selection effect could play a significant role, therefore controlling for it like Bertoni et al. (2013) would also enhance the results. Also, it is interesting to study the formal governance mechanisms related to each fun type.

Abstract

Résumé

Resumen