This paper investigates the determinants of survival, acquisition and non-survival for internet firms that have gone public at the NASDAQ stock exchange from December 1996 through February 2001. We analyze two models, survivors versus non-survivors and survivors versus acquired firms with the help of the Cox proportional hazard method. We have selected all internet firms that went public during the internet bubble. The paper presents a model that includes a number of new and earlier examined, explanatory variables simultaneously that are often studied separately and for noninternet firms in other studies. The key issue of the study is that our models provide a prediction of the rate of non-survival or acquisition of internet firms.The empirical results provide evidence that compared to both acquired firms and nonsurvivors, survivors' owners retain a larger share in the firm at IPO. Our study presents a number of novel findings. Survivors differ significantly from non-survivors in the number of risk factors in the IPO prospectus and underwriter reputation. Furthermore, our results are consistent with risk factors indicating the economic fundamentals of an internet IPO rather than being a mechanism against legal liability. Although internet firms belong to a relatively young branch of industry, the outcome provides evidence that top bracket lead managers are successful in selecting high quality internet IPOs. Also, our empirical results document that during the last phase of the internet bubble the industry was turning into a hot market with weaker firms going public due to reduced scrutiny of investors. Our findings hold under a number of robustness checks.

No abstract

Testing for mediation, or indirect effects, is empirically very important in many disciplines. It has two obvious symmetries that the testing procedure should be invariant to. The ordered absolute t-statistics from two ordinary regressions are maximal invariant under the associated groups of transformations. Sobel’s (1982) Wald-type and the LR test statistic are both functions of this maximal invariant and satisfy two logical coherence requirements: (1) size coherence: rejection at level α implies rejection at all higher significance levels; and (2) information coherence: more (less) evidence against the null implies continued (non) rejection of the null. The LR test statistic is simply the smallest of the two absolute t-statistics, and we show that the LR test is the Uniformly Most Powerful (information and size) Coherent Invariant (UMPCI) test. In short: the LR test for mediation is simple and best.

Summary In Fabrizio and Tsolmon (Review of Economics and Statistics, 2014, 96(4), 662–675) and Barlevy (American Economic Review, 2007, 97(4), 1131–1164) it was concluded that R&D investments are procyclical. Fabrizio and Tsolmon utilized a model based on Barlevy, but it differed in some respects and allowed for more heterogeneity. However, we doubt whether their implied trends are intended. Fabrizio and Tsolmon also set missing values for R&D equal to zero, leading to unrealistic jumps in investment and its first differences. We reconcile and replicate both the Fabrizio and Tsolmon and Barlevy papers by considering extensions that encompass both models. Furthermore, we treat missing values more appropriately and consider some alternative specifications to check the robustness of the results. Procyclicality is confirmed, but we find much less heterogeneity than Fabrizio and Tsolmon did. In particular, obsolescence and patent effectiveness are no longer important but external financing is.