Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The proposed model, by incorporating both (1) banker screening of new issues and (2) costly evaluation by investors, is the first to admit endogenous double-sided information production. It demonstrates a nontrivial link between these two sides: the banker wishes to structure a sale conducive to investor research because selling to an uninformed pool would result in his own shirking. One application of this paradigm indicates that, contrary to the findings of most IPO models, larger investor pools are not always better. This result resolves the "participation restriction puzzle" of why bankers do not open sales to all bidders even when doing so would maximize competition and reduce underpricing. Copyright 2005, Oxford University Press.