Special Acknowledgment

Abstract

Abstractased on a sample of 15 European countries, this survey analyses various features of the European IPO (Initial Public Offering) market over the period from 1995 to 2004: listing requirements, IPO-mechanism choices, performance and secondary market liquidity.First, the comparison of national primary market regulations, in spite of the commonly observed segmentation between Main, Parallel and New Markets, shows a wide diversity in listing requirements and reveals that the primary market's mechanisms are almost always monitored by investment banks, which then control the initial pricing and allocation of new issues. The examination of issuers' practices looks at the increase in the different types of IPO mechanisms in the late nineties and the widespread use of the book-building mechanism nowadays. Second, our empirical analysis of IPO short-term and long-term performance confirms, with a few exceptions, widely recognised patterns, but also show discrepancies between countries, periods, sector and primary listing mechanisms. The average initial underpricing amounts to 22% over our pan-European sample and is observed at various levels in each of the 15 countries of the sample. Empirical evidence on long-term performance is less clear. Results are not benchmark-dependent but sometimes differ from one measurement method to another. However, in line with previous studies, significant underperformance is found at the 3-year horizon with all methodologies and in all countries, except Greece and Portugal.Finally, using a sample of IPOs launched on Euronext between 1995 and 2004, our study examines the relationship between initial returns and post-listing liquidity in the short and in the long-run. We support the 'illiquidity-compensation hypothesis'. Initial underpricing is positively linked to information asymmetry in the after-market. It produces higher turnover immediately after the IPO but has no effect on trading volumes after the first year of trading, so that this liquidity effect cannot be put down to ownership structure but is more likely attributable to the interest underpriced stocks generate.