Discussion of Explaining the Short‐ and Long‐Term IPO Anomalies in the US by R&D

Abstract

Guo, Lev and Shi (2006) (hereafter GLS) set out on the formidable task of attempting to find a common explanation for the short- and long-term 'anomalies' commonly observed of firms that undertake an Initial Public Offering (IPO). Specifically, the excessive positive first-day returns of IPO firms (referred to as underpricing) and the subsequent poor long-term performance of these firms are among the well-documented anomalies in the finance literature. These findings persist through time (although of varying magnitude) and across jurisdictions despite differences in institutional arrangement and litigation environment. As the authors state, prior literature has presented theoretical explanations for these anomalies based on either information asymmetry or to investors' systematic mispricing explanations. Whilst both sets of theoretical work predict short-term underpricing and long-term underperformance, fewer empirical studies explicitly consider the links between these anomalies. GLS focus their study on the role of a specific source of information asymmetry and uncertainty at the time of public offering - research and development (RD and Rock, 1986) and that some firms will use credible signals to reduce adverse selection problems between buyers and sellers and hence lower the cost of capital (Allen and Faulhaber, 1989). Prior studies attribute the poor long-term performance of IPO firms to investor optimism (Purnanandam and Swaminathan, 2004) or earnings management prior to the offering (Teoh et al., 1998) or to misspecification in measuring returns in long-run horizons (Brav and Gompers, 1997). Whilst numerous studies provide compelling evidence to explain these phenomena, as GLS suggest 'it is fair to state that both anomalies are still unresolved'. GLS base their reconciliation of the anomalies by arguing that managers of R&D intensive IPO firms are privy to proprietary information on the potential of R&D projects, hence they are willing to 'leave more money on the table' than firms without R&D ('No-R&D') IPO firms because they believe they can recover these costs in subsequent equity offerings. They also predict that investor optimism in R&D intensive IPOs is dampened by the expected uncertainty on the prospects of R&D activities, as a result these firms are not expected to underperform as poorly as No-R&D IPO firms. GLS provide some empirical evidence to support their conjectures and further suggest that R&D is a major source of information asymmetry as it 'subsumes' the importance of other previously known proxies for information asymmetry such as the presence of venture capitalists. My discussion of this paper summarises and expands on the points raised at the JBFA conference. I discuss the use of R&D costs as a valid proxy for information asymmetry and also offer some comments on the results and analyses presented by the authors in support of the above conjectures.