The dramatic growth of equity markets throughout the last decade has, to a large extent, been a result of the increasing number of companies that chose to list their stock on public markets via an initial public offering (IPO). The number and frequency of IPOs has risen impressively in the US, in Europe and in many developing countries. This is part of a worldwide shift away from private or bank finance, towards funding via public security markets. The process is commonly explained by the decreased cost of equity capital and the increased availability of equity finance associated with more integrated capital markets and faster information linkages. Improvements in the design and performance of primary equity markets may also have contributed to this process. Among these are the diffusion of book-building techniques, better disclosure rules, greater expertise and competition among investment banks and, possibly, competition among stock markets.

Held in Capri on 16/18 June 2000, a joint CEPR/CSEF/NYSE/TMR Conference assessed the improvements in the design and performance of primary equity markets and their subsequent implications. The insights that could be taken away from the papers presented fall broadly in two classes: those concerning the microeconomic aspects of IPOs, and those concerning the overall performance and macroeconomic impact of the primary equity market.

The Microeconomics of IPOs

It is well known that 'IPO underpricing' is a key determinant of the cost of equity capital for companies that tap the stock market for the first time. Typically, the price at which the shares of an IPO are placed with investors is below the level that they reach on the market a few days or weeks later, when more complete public information is available. Most research by financial economists relates such underpricing to the presence of some investors endowed with superior information about the value of the IPO shares, or to the need to compensate professional investors for the costs of producing information about the company.

The conference added three important insights to this much-researched topic. First, IPO underpricing is lower when other companies go public, because each IPO generates beneficial information externalities for other companies that are about to go public. Second, designing the IPO procedure also matters: bookbuilding allows substantial cost savings, but these savings materialize only if underwriters are willing to let the issue price vary outside the range initially chosen in response to demand. The third insight concerns the motivation itself of the IPO process: the going public decision is influenced by firms' ownership structure. When their shares are held by only one owner and when banks own shares, then companies are more likely to prefer private rather than public sales of equity.

The first paper, entitled 'Evidence of Information Spillovers in the Production of Investment Banking Services' by Lawrence Benveniste (Carlson School of Management, University of Minnesota), William Wilhelm (Boston College) and Xiaoyun Yu, highlighted various implications for information externalities in the IPO process: bunching by industry, implicit subsidies from the leader to the followers and the tendency for monopolistic financial intermediaries to engage in less underpricing when many companies go public. These predictions are consistent with the evidence from US IPOs, where the contemporaneous number of IPOs affects the estimated proceeds in the pre-offer period, and IPO underpricing is reduced by clustering and by firm-specific information that was not publicly available. Moreover, the information spillover is twice as large for information-sensitive industries than for other industries.

The second paper, entitled 'Has the Introduction of Bookbuilding Increased the Efficiency of International IPOs?' and co-authored by Tim Jenkinson (University of Oxford and CEPR), Alexander Ljungqvist (University of Oxford and CEPR) and William Wilhelm, used a large cross-country data set to show that bookbuilding can reduce costs if certain other conditions are also met. The authors illustrate that bookbuilding has higher costs but countervailing benefits when the IPO is marketed by US banks and sold to US investors. Jenkinson et al attribute these benefits to the US underwriters' willingness to price outside the initial range: placements directed to US investors are often priced outside the initial range by as much as 30%. This may reflect the lack of legal or regulatory impediments as well as greater transparency and competition for issues marketed by US banks to US investors. Non-US banks appear reluctant to respond to unexpectedly high demand by raising prices outside the initial range, possibly due to the power of large investors, and this undermines bookbuilding.

The above result was confirmed by Francesca Cornelli (London Business School and CEPR) and David Goldreich (London Business School) in their paper 'Bookbuilding: How Informative is the Order Book?'. The two authors explore the actual order books for 64 international issues sold by a European investment bank. The results show a high percentage of issues priced at the top of the initial price range, with a substantial oversubscription at the issue price (the average IPO being oversubscribed 10 times).

The discussion of these two papers brought to the fore the idea that underpricing depends not just on the sale method (bookbuilding versus other mechanisms) but also on the objective function of underwriters and their regulatory constraints. If underwriters try to design the optimal mechanism to maximize the seller's objective function and are unconstrained by regulation, then almost by definition they should do best with a bookbuilding mechanism, which allows them to extract information useful to set the issue price, as shown by Cornelli and Goldreich.

The real question is whether underwriters have the 'right' objective function, face tough competition and are unrestricted by regulatory constraints. For the US underwriters, this seems to be the case, which helps explain why they dominate the IPO industry. By the same token, insufficient competition among bidders and collusion between investment bankers and bidders may explain the higher IPO underpricing when US banks are not involved. This may change in the future, partly because more sophisticated auction methods via the Internet may enhance the competition among bidders and the transparency of IPO sales.

The conference also added new insights into the motives of companies that go public. The identity of the initial owners of the company appears to play an important role in this decision. Ekkehart Böehmer (US Securities and Exchange Commission) and Alexander Ljungqvist, in their paper 'The Choice of Outside Equity: Evidence on Privately-held Firms', analyse 266 German firms that have pre-announced their intention to go public and show that firms that issue new shares are more likely to complete the IPO process. In contrast, other companies, and in particular those that have majority owners or a bank among the shareholders, tend to use the pre-announcement to signal their willingness to find new partners but eventually remain private.

In 'Why do Governments List Privatized Companies Abroad?' Bernardo Bortolotti (FEEM, Università di Torino), Marcella Fantini and Carlo Scarpa (Università di Bologna) demonstrate that when a company is being privatized, political variables also play an important role in the decision to go public. In particular, the decision to list companies abroad appears to reflect a desire to lock them into a more investor-friendly legal framework, presumably in order to sell their shares at a better price. Examining 342 listings of privatized companies in 42 countries, the authors find that privatized companies from OECD countries tend to cross-list abroad in countries offering better legal protection of shareholders.

Overall Performance and Macroeconomic Impact of the Primary Equity Market

Even if market participants do as well as possible in designing the sale mechanisms of new stock issues, two important questions still remain. First, is it possible (and worthwhile) to try to encourage IPOs by fostering the venture capital industry and by setting up special markets such as the 'new markets' that have recently sprung up in Europe? And second, should the markets where these new issues are traded be designed and regulated in any special way?

Claudio Michelacci and Javier Suárez (CEMFI, Madrid, and CEPR) shed some light on the first issue with their theoretical paper on 'Business Creation and the Stock Market'. They show that the 'informed capital' of venture capitalists and the stock market play complementary roles: the stock market allows venture capitalists to recycle their scarce informed capital. The logic of their model is that new businesses require a special type of capitalist who can solve information and incentive problems, and thereby allow these firms to postpone their IPO until their profitability prospects are clearer. The scarcity of this informed capital therefore acts as a constraint on the rate of business creation. The lower the listing costs of firms, the faster venture capitalists can unload the firms they have catered for on the stock market and 'recycle' their informed capital with new businesses.

This suggests that any policy that can reduce the listing costs of new businesses will translate into faster recycling of informed capital and faster real growth. Within this framework, the difference between the IPO market (and the rate of business creation) in Europe and in the US could be attributed to higher listing costs and lower availability of venture capital in Europe. But, as Patrick Bolton (Princeton University and CEPR) noted, it is not clear if the European bottleneck lies in a scarcity of informed capital or in a scarcity of valuable and innovative projects to be funded.

Even assuming that listing costs are higher in Europe than in the US (an assumption for which currently there is no solid evidence), the evidence provided by Asher Blass and Yishay Yafeh (Hebrew University of Jerusalem) suggests that listing costs are not of crucial importance in the decision to go public. In their paper 'Vagabond Shoes Longing to Stray: Why Foreign Firms List in the United States', they show that high-tech, high-growth, export-oriented Israeli companies flock on to the Nasdaq, forgoing the substantial tax benefits of listing in Tel Aviv. These companies do not even try to reap such benefits by cross-listing their shares in Israel after listing in the US. In fact, Blass and Yafeh argue that these Israeli companies list in the US precisely because listing there is costlier than in Israel: they do so in order to signal their superior quality. Only firms with very large growth and profit opportunities can face the larger costs of an IPO in the US, in terms of lower private benefits of control, larger underpricing and underwriting fees, and forgone tax benefits in Israel.

A similar signalling story was told by Jörg Kukies (University of Chicago) to explain why the recent IPO boom in Germany was associated with the creation of the Neuer Markt (NM) in 1997. Firms admitted for listing to the NM must be first admitted to the traditional exchange of the Deutsche Börse. Additionally, they must fulfil other requirements, especially in terms of information dissemination and accounting rules. Therefore the companies that went public on the NM could have gone public before, but did not. In his paper on 'The Effects of Introducing a New Stock Exchange on the IPO process', Kukies argues that the NM's stringent information disclosure requirements provided a precommitment device that did not exist before. Listing on the NM acted as a signalling device for the most promising companies, in the same way as listing on Nasdaq does for the best Israeli companies, according to Blass and Yafeh.

However, as Oren Sussman (Ben-Gurion University of the Negev) remarked, the requirements imposed by the 'new markets' in Europe are not uniformly stricter. They are stricter about information disclosure but less restrictive about age, past profitability and size, and also allow companies to sell a smaller fraction of their shares to outside shareholders than traditional exchanges. Hence it is natural to ask which matters more, the strictness on disclosure or the greater leniency in these other dimensions? The evidence provided by Kukies is not based on a 'clean' experiment. At the same time as the NM was being created, US investment banks entered the German IPO market, German banks themselves became more supportive of IPOs and the demand for stocks rose dramatically, especially for high-tech stocks.

Whatever the intrinsic merits of the requirements imposed by the Nasdaq and the NM, it should be evident that both the design of these markets and the listing choices of the companies across markets are endogenous. Increasingly, stock markets tend to compete for listings with each other, and will tend to differentiate their listing requirements and trading mechanisms so as to soften such competition. This is illustrated by Thierry Foucault (HEC School of Management, Jouy en Josas, and CEPR) and Christine Parlour in their paper 'Competition for Listings'. For instance, a possible equilibrium configuration is one in which one market displays low trading costs but high listing fees while another does the opposite. The first market will be attractive for large companies, which will be ready to pay the high listing fees in return for a more liquid market for their shares, whereas the second market will specialize in smaller companies – an example strikingly reminiscent of the differences between the NYSE and Nasdaq.

Rapporteur: Marco Pagano (Università di Salerno and CEPR)

The Conference papers can be downloaded from www.cepr.org/meets/wkcn/5/554/