IPO underpricing in Italy

Abstract

This article analyses the first-day return of 182 IPOs listed on the Italian Stock Exchange from 1985 to 2001. It finds a significantly mean positive underpricing (21.87%). Contrary to the evidence detected in the USA by Loughran and Ritter and Ljungqvist and Wilhelm, it highlights that on the main board of the Italian Exchange IPO underpricing decreased in the late 1990s. It claims that such a pattern can be accounted for by two determinants: (i) the evolution of pricing strategies, from fixed-price IPOs to bookbuilding, (ii) the segmentation of the Italian Exchange with the birth of a new board for high-growth and technology firms (Nuovo Mercato). It shows that IPOs are intentionally underpriced: both public and private information available at the IPO is only partially incorporated in pricing the shares. The results suggest that negative feedback learned during the preselling is more fully incorporated into the offer price than positive information. Finally, it shows that price revisions are partially predictable on the basis of public information at the time of the offering.

Acknowledgements

The authors wish to thank, for helpful suggestions on earlier versions of this article, Andrea Beltratti (Università Bocconi di Milano), Mara Faccio (Università Cattolica di Milano), Luca Filippa (Borsa Italiana), Rocco Mosconi (Politecnico di Milano), Jay Ritter (University of Florida), Peter Roosenboom (Erasmus University, Rotterdam), James W. Wansley (University of Tennessee), Kent L. Womack (Tuck School of Business at Dartmouth) and the participants at the 1999 APFA Annual Meeting in Melbourne, 2000 FMA Conferences in Edinburgh and Seattle, 2000 EFMA Conference in Athens. Last but not least, they thank the Italian Stock Exchange (Borsa Italiana) and CONSOB for providing some of the data. The usual disclaimer applies. Research grants by CNR and MURST are acknowledged, as well as assistance from Andrea Randone and Fabio Bertoni.

Notes

¹Loughran and Ritter (2000) showed that US IPOs significantly underperform both the market and comparable firms in the first three years after listing, and similar results are reported in other countries (a cross country survey is contained in Giudici and Paleari, 2001). A significantly overperformance is highlighted in emerging countries. Yet, detecting long-run abnormal performance raises several methodological concerns (Ritter and Welch, 2002).

²Booth and Smith (1986) and Carter and Manaster (1990) sustained the ‘certification hypothesis’: the more reputable the underwriter, the lower the underpricing. More recently this hypothesis has been refused by Michaely and Shaw (1994) and Beatty and Welch (1997). Cooney et al. (1999) argued that in high-demand IPOs high-reputation underwriters are able to exploit their superior bargaining position to underprice the IPO more severely.

³See Megginson and Weiss (1991), Barry et al. (1990). More recent evidence of an apparent reversal in this relationship is provided by Francis and Hasan (2001) and Habib and Ljungqvist (2001), explained by a conflict of interests between the venture capitalist, the underwriter and the entrepreneur, especially when the venture capitalist is affiliated with the underwriter.

⁴In this case the investment bank requires that insiders agree to refrain from selling their stock in the aftermarket for a period of time following the IPO.

⁵The new issue process is regulated by a public authority, CONSOB, which performs a role that is comparable to the SEC in USA, and by a private company, Borsa Italiana SpA that manages the Stock Markets in Italy. CONSOB (http://www.consob.it) has to be informed in advance of the offering conditions and has to certify that the issuer provides adequate information to the public (collected in an officially approved prospectus). Borsa Italiana (http://www.borsaitalia.it) deliberates the admission to the listing, after having verified all the necessary requirements.

⁶Until 1997 income realized by newly listed small- and medium-size firms (issuing new shares) had been levied at a reduced rate equal to 21%. In 1997 a further tax reform allowed all Italian companies to apply a reduced tax rate equal to 19% (dual income tax) at the income deriving from new equity capital raised or ploughed-back profits. In order to induce firms, particularly SMEs, to go public a particular disposal has been introduced for companies newly listed on stock markets: for three years the relief of 19% can be reduced to 7%.

⁷Details about general characteristics and ownership structure of the Italian listed companies may be found in the CONSOB and Borsa Italiana SpA Internet pages (see note 5). Although in the last years the number of listed companies has not significantly increased, a relevant turnover has considerably reset the stock market outline.

⁸In the second period the privatization process in the banking sector has been relevant; nevertheless, it has been realized through public offerings of secondary shares held by the Government and already listed on the stock market.

⁹In fact, the phenomenon is imputable to the process of ‘financial dismantling’ and separation between ownership and control experienced in Italy during the 1980s by large business groups.

¹⁰For example, Habib and Ljunqvist (2001) referred to the US market and report a mean age equal to 14 years. In Europe a higher comparable mean age is reported by Vandemaele (1999) for the French market (44 years), Roosenboom et al. (1999) for the Netherlands (35 years), Holmen and Högfeldt (1999) for Sweden (31 years).

¹¹Cooney et al. (1999) found 67.4% in their US sample, Lee et al. (1999) 53.0% for the Australian market, Goergen (1998) 76.4% and 62.6% for the German and UK market respectively, Roosenboom et al. (1999) 64.6% for The Netherlands.

¹²Among the 182 sample firms, only five were acquired by outside bidder companies in the three years following the IPO.

¹³As the t-test implies a normal distribution of the stochastic variable and this may not be justified in this case, the tests were validated through use of the Tchebyceff inequality.

¹⁴http:www istat.it

¹⁵Derrien and Womack (2003) stated that the auction mechanism is associated with less underpricing and thus more efficient, since this procedure is able to incorporate more information from recent market momentum into the pricing of the IPO. Kandel et al. (1999) examined the IPO auctions in Israel; they stated that in cases of auctioned IPOs investors gain information about the elasticity of the demand for stock, revising the prices of securities according to the new information. In this case the underpricing is entailed by the uncertainty about the demand elasticity, which is assumed to be important to determine the stock value.

¹⁶Note that in Italy the offer price has never been set beyond the maximum file limit.

¹⁷In other words, when volatility increases, one expects a higher divergence of opinion among investors. Therefore, the most optimistic investors should be more willing to purchase IPO shares. As a consequence, intermediates may reduce the initial underpricing.

¹⁸According to the hypothesis, intermediates are given the option to adjust the offer price with respect to the initial file range. This option is more valuable, the larger the uncertainty of the offering success, which is larger for small, risky firms and for larger offerings. In such cases, the incentives to postpone optimistic revisions to the future are higher.

¹⁹A study about IPO underpricing in European ‘new markets’ (comprising the NM) has been carried out by Giudici and Roosenboom (2002) who also compared such IPOs with new listings on the national exchanges’ main boards.