Résumé
In this study, we attempt to predict which IPOs will benefit from analyst coverage with information contained in their prospectus or revealed immediately after the first day of trading. We find that the predictive models are very accurate in classifying approximately 75% of the US IPOs over the period from 1994 to 2012 correctly. We identify best predictors for analyst coverage. These are (i) underpricing, (ii) high tech firms and (iii) bookbuilding method. For example, CART model classifies orphan IPOs with slightly worse accuracy (70.6%) than non orphan IPOs (81.1%). This result is evidence that analyst coverage can be predicted with only information contained in the IPO prospectuses or immediately after the first day of trading