Essays on The Role of Network Centrality in The Australian IPO Market

Abstract

Modern financial systems are highly interconnected. Institutions or individuals form connections with each other in different ways for different purposes. Networks, broadly understood as a collection of nodes and connections between nodes, can therefore be a useful representation of financial systems. Using social network analysis (SNA), this thesis provides novel empirical evidence on the effects of network centrality, which refers to an actor’s position and level of connectedness in a network, on Australian initial public offering (IPO) firms that went public between 2001 and 2017. The thesis includes three essays that aim to examine: (1) the relationship between investment bank networks and the accuracy of IPO prospectus earnings forecasts, (2) the overinvestment problem associated with excess cash in seasoned versus newly public firms, and whether investment bank networks affect the performance of IPO firms with excess cash, and lastly, (3) the relationships between board networks, cash holdings, and IPO firms’ performance. The first empirical essay investigates whether investment banks with more central network positions could improve earnings forecast accuracy in the context of Australian IPOs issued during the 2001-2017 period. Using SNA to construct investment bank network centrality, I show that IPOs backed by more central investment banks tend to have higher absolute offer price revision, exhibit greater forecast accuracy, and are associated with more pessimistic forecasts. Taken together, the results support the idea that, relative to their counterparts that are peripheral in the network, more central investment banks may produce more information in the IPO process and certify the quality of IPO issuers by encouraging more conservative and accurate earnings forecasts. This study makes an important contribution by adding to the burgeoning literature on network studies in the context of IPO investment banking. To the best of my knowledge, this study is the first to examine the relation between investment bank networks and earnings forecast accuracy, particularly in a setting where forecast disclosure is a voluntary and reasonably common practice (such as that of Australia). This study also has significant implications for key participants in the Australian IPO market. For investment banks, working in syndicates when managing an IPO is essential not only for risk-sharing purposes but also to gain knowledge and experience that enhance their certification ability. For issuers, hiring more central investment banks may help them prepare more accurate forecasts and to avoid increased exposure to costly litigation risks. As for potential investors, the presence of more central investment banks in an IPO could provide some assurance of the issuer’s forecast accuracy, which is useful for investment decision-making. The second empirical essay analyses the overinvestment problem associated with excess cash holdings in ASX-listed firms for the period between 2001 and 2017. First, I show that firms with higher excess cash are associated with higher capital expenditures and lower firm performance, indicating that cash-rich firms tend to invest their excess cash in less-profitable projects. I find that this overinvestment problem is more pronounced among newly public firms that have undergone an IPO within the prior three years. Next, I test whether investment bank networks could reduce this overinvestment problem. I find that IPOs with excess cash managed by more central investment banks experience lower capital expenditures but higher firm performance. The evidence shows that investment banks benefit from being well connected in the network; hence, the banks can better monitor and advise the IPO firms, constraining managerial opportunism regarding the use of excess cash. This study contributes to the literature as the first study that brings together two streams of research which have not overlapped—that is, the literature on excess cash holdings and the role of investment banks in IPOs. Connecting the two research streams gives new insights into the largely unexplored area of excess cash in IPO firms, and further our understanding of investment bank monitoring and advising roles in the post-IPO period. Finally, the third empirical essay examines how board networks formed by interlocking directors impact the use of cash in Australian IPO firms. I find that well-connected boards tend to reduce their firms’ cash holdings. Further, high-cash firms with well-connected boards often underperform, which is indicative of their suboptimal investment decisions. As excess cash accumulates, firms with well-connected boards prefer to spend cash on capital expenditures rather than returning it to shareholders through dividends. In light of two competing views on connected directors—i.e. the reputation hypothesis and the busyness hypothesis—the findings are closer to the latter. This study represents the first attempt in the literature to address the impact of board networks on cash holdings in IPO firms. The evidence presented in this study adds to the ongoing debate regarding the benefits versus the drawbacks of board networks, casting new light that IPO firms with more connected directors are associated with less efficient use of cash. In the absence of regulation, the evidence may lead us to call for policies that limit board networks in Australia. Nonetheless, it is worth acknowledging that prescribing the appropriate numerical limit for multiple directorships is challenging for regulators since firms’ sizes and the complexity of the issues their boards face vary widely from one firm to another.