پیچیدگی استراتژی های شبکه بین شرکتی در صنعت بانکداری سرمایه گذاری در کانادا

Using data on all underwriting syndicates in Canada over nearly 40 years, we examine whether, and if so to what extent, managers are aware of and strategic about their network positions by comparing the effects of partner selection on network position at two levels of complexity. Our findings show that when investment banks’ managers formed and joined underwriting syndicates, they improved their network positions by spanning more structural holes. They did not, however, distinguish between constrained and unconstrained structural holes, which would require a more complex understanding of the network. Our study suggests that models of network-based competitive advantages and network change need to consider more fully firms’ network strategies and the cognitive limits of the managers enacting them.

Recent research on interfirm networks has contributed to corroborating the theoretical claim (Burt, 1992) that, in certain environments, actors spanning structural holes realize network-based advantages over firms occupying positions spanning fewer or no structural holes (Rowley, Behrens, & Krackhardt, 2000; Walker, Kogut, & Shan, 1997). A firm spans a structural hole if it is linked to two partners that are not tied to one another and, consequently, gains information access and control benefits not available to other network positions. Firms spanning structural holes gain brokerage power by acting as intermediaries between disconnected partners who rely on them to facilitate exchange flows across the network. Competitive advantage also accrues in the form of unique information from non-redundant contacts, which economizes on the number of ties required to access timely and diverse information (Burt, 1992). Although there are other types of social capital serving different purposes (Burt, 1998), in uncertain and dynamic environments requiring firms to explore their surroundings for new information and know-how to continually adapt to their environments, structural holes represent a competitive advantage (Rowley et al., 2000; Rowley & Baum, 2002). Our study is motivated by two aspects of this research. First, social network research suggests that a firm's partnering choices, which determine in part its network position, contributes to the strength of a firm's competitive advantage (e.g., Podolny, 1993). Second, although it is increasingly clear that there are advantages associated with occupying particular interfirm network positions (e.g., Powell, Koput, & Smith-Doerr, 1996), it is unclear whether, or to what degree, managers are capable of conceptualizing their firms’ networks, comprehending the nature of network-based advantages, or using collaborative arrangements strategically to improve their network positions. There are two contrasting views on managers’ cognitive skills with respect to their firms’ networks and network strategies—the network strategy perspective and partnering strategy perspective ( Rowley & Baum, 2002). According to the network strategy perspective, managers understand the benefits realized from certain network positions and comprehend their networks sufficiently well to enable them to use their partnering decisions to maneuver strategically into beneficial network positions. Burt, for example, characterizes actors as alert to the locations of structural holes and their benefits, and so as actively seeking structural opportunities: “The task for a strategic player building an effective network is to focus resources on the maintenance of ties [spanning structural holes]” (1992, p. 30). Thus, as Burt (2001, p. 13) puts it, “the social benefits of [structural holes] are an incentive to build them.” Doreian agrees that actors should be and are strategic about social capital advantages: “Networks have instrumental character for network members as these members have structured goals and some goals are achieved through network choices” (2002, p. 95). One source of empirical evidence of firms engaging in network strategies is found in studies of interfirm status homophily, which show managers using information on the relative positions of firms in their network as a basis for partner selection (e.g., Podolny, 1993). The partnering strategy perspective, in contrast, views managers as more boundedly rational, focusing on selecting the most capable and reliable partners to create dyadic competitive advantages for specific alliance objectives, rather than considering how selecting particular partners affects the firm's network position and advantage. Much of the research on partner selection suggests that managers perform localized, myopic searches, evaluating potential allies based on first-hand knowledge and endorsements from common third parties regarding their capabilities and reliability, and showing a strong propensity to choose past partners and partners’ partners ( Chung, Singh, & Lee, 2000; Gulati, 1995; Gulati & Gargiulo, 1999; Uzzi, 1996). Thus, the partner selection perspective suggests that firms focus on establishing effective partnerships, and that effective network positions are emergent phenomena that result from these choices ( Gulati & Gargiulo, 1999). In this study, we examine whether and, if so to what extent, managers are strategic about their network positions. Specifically, we ask: Do managers make partner-selection decisions that are consistent with the notion of the network strategy perspective? And, if so, how sophisticated are the network strategies managers enact? We confront the former question by empirically examining whether or not firms’ managers improve their firms’ network positions when they select partners. We address the latter question by comparing the effects of firms’ partner selection on their network positions at two levels of complexity. At the first level of complexity, we examine whether managers use partnering opportunities to span structural holes, which requires managers to have a deeper understanding of their firm's network and network advantages than is implied by the partnering-strategy perspective. At the second level, we investigate whether managers are able to differentiate between constrained and unconstrained structural holes, the latter of which provide superior returns, when selecting partners. Following Burt (1992), we define constraint as the extent to which a firm has invested in ties with partners (and partners’ partners) that span many of the same structural holes as the firm, and thus provide more limited brokering advantage. Differentiating structural holes on the basis of their constraint represents a more sophisticated network strategy that requires managers to have a more detailed comprehension of their networks to enact. We examine firms’ network strategies in a study of all underwriting syndicates formed by investment banks in Canada between 1952 and 1990. This empirical setting provides an opportunity to comprehensively track firms’ partnering behavior over time and to assess how their partnering decisions shape their network positions. Notably, these syndicate relationships are meaningful beyond the task of underwriting new capital issues. These ties act as conduits to new underwriting opportunities and diverse groups of investors, and thus contributing to banks’ reputations and performance (Podolny, 1993).

Research at the intersection of strategy and social network analysis consistently shows that network structure and position are important determinants of firm performance—those firms better positioned in their networks outperform their rivals. We too find that network structure matters for firm performance (H1, H2)—firms in particular network positions experience higher performance in the subsequent period. And, in the investment banking industry, where acquiring novel information and timely identification of opportunities are paramount to success, our analysis indicates that unconstrained structural holes are positively related to performance. Thus, investment banks compete for market share through collaborative strategies and those banks that seek partnerships that span structural holes will outperform those that follow partnering strategies (seek embedded ties in dense structures). By forming cooperative ties with unconnected others, banks gain a competitive advantage over their partners. A key managerial implication of this body of research is that managers should pursue network strategies, selecting partners that improve the firm's network position. The idea that managers are aware of their firms’ networks and the types of positions that provide social capital advantages—core assumptions underpinning the network strategy perspective—remains relatively unexplored ( Rowley & Baum, 2002). The cognitive demands required to enact a network strategy are greater than those required for a partnering strategy, in which boundedly rational managers select partners based on localized search and evaluation of potential partners’ reliability and capabilities based on first-hand knowledge and partners’ endorsements. Our primary goal in this study was to provide evidence regarding whether or not, and at what level of sophistication, firms’ managers are aware of their firms’ networks, the benefits of occupying certain positions in the network, and so act strategically to improve their firms’ network positions when they select partners. Because network positions rich in structural holes enhance market share performance in our empirical setting, banks’ managers face a decision tension that is useful for uncovering how they view their networks. According to the partner selection literature, banks are predisposed to select past partners for future collaborations because they have first-hand knowledge of these partners’ capabilities and trustworthiness. At the same time, however, banks may gain performance advantages by spanning structural holes that form ties with partners beyond their local networks of past partners and partners’ partners. In our models, the highly significant positive coefficient for lagged network-effective size (see Table 2) indicates that a bank's network position was strongly inertial and that their managers tended to reconstitute past relationships. Despite this tendency, however, banks’ managers also formed and joined underwriting syndicates in ways that lead them to increase the number of structural holes they spanned (H3a, H4a). Thus, our sample banks’ managers appeared to consider network-level factors beyond the creation of dyadic competitive advantages when selecting partners. Our findings also reveal the limits of these managers’ cognitive abilities, however. Specifically, while apparently able to identify and span structural holes, the managers did not discriminate among their banks’ structural-hole opportunities by forming and joining syndicates whose partners (and structural-hole opportunities) were relatively less constraining (H3b, H4b). Avoiding constraint in their networks requires managers to understand the subtle complexities of structural holes. Thus, the limits of managers’ cognitive abilities appear to fall somewhere between the identification of structural holes and the more difficult task of differentiating among them. These limits are particularly meaningful, given the importance of syndicate ties motivating banks’ managers to attend to them, and the availability of information about syndicates relationships making it possible for them to do so. While our study reveals some new insights into the network-strategy perspective, much work remains to be done. Our findings, suggest to us several promising directions for future research in this area. Krackhardt (1990) found that managers exhibited differential abilities to comprehend their intraorganizational networks. Do managers’ abilities to grasp their interorganizational networks differ similarly? If so, a useful extension of our study would be to identify factors—individual, organizational and network—explaining these managerial differences. Do certain types of managers pursue network strategies demanding sophisticated knowledge of their networks, such as an ability to recognize value differences across structural hole opportunities? Further research directions stem from several limitations of our study. Investment banking has features that do not characterize all industries, and this likely affects, potentially greatly, the logic of banks’ network strategies. Generalizations should be bounded to settings where structural holes and sparse network positions are a source of competitive advantage. Consequently, further research is needed in settings like ours, but where managers’ incentives and/or opportunities to pursue network strategies are weaker, for example, because information on interfirm ties is not publicly available. Work in settings where cohesive ties and dense network positions are a source of competitive advantage is also needed. In such settings, network strategies and partner strategies should work in concert. Does this render firms’ network strategies ‘redundant’, or do their partnering choices still reveal influences beyond partner-specific concerns? Additionally, while our empirical findings provide evidence of network strategies in the context of richly specified empirical models accounting for a wide range of alternative explanations, they do not provide direct evidence that banks’ managers are aware of and pursue network strategies. Therefore, to further examine the plausibility of this interpretation of our results, we conducted a series of semi-structured interviews with managers responsible for forming syndicate relationships at 13 Canadian investment banks, to gain qualitative insight into the syndicate formation process and banks’ partner selection. Consistent with the network-strategy view, the bankers we interviewed understood well the information advantages associated with relationships. The manager of a small investment bank that acts primarily as a co-lead, for example, indicated that for their firm, “information and access to it are king … being close to the source is the name of the game … I don’t have time to know everyone, but I need to be close to those that have the best contacts.” And, as another manager explained, “the best players in the industry build reputations by getting the biggest clients and controlling information, and carefully passing it out to others. It makes you a hot commodity, like a hot concert ticket or restaurant—everybody wants some.” Of course, such anecdotes do not provide systematic corroboration of managers pursuing network strategies for their banks. However, along with our quantitative findings we hope this interview evidence inspires future research designed to examine more systematically investment bank managers’ mental models of their interfirm networks (e.g., Porac, Thomas, Wilson, Paton, & Kanfer, 1995). Although evidence is mounting on the performance implications of firms’ network positions, our study provides the first empirical evidence regarding whether or not, and at what level of sophistication, firms’ managers enact network strategies. Overall, our study indicates that to improve models of the nature of network-based competitive advantages and network change, researchers will need to more fully consider firms’ network strategies and the cognitive limits of the managers’ enacting them.