The French Initial Public Offering market and the role of venture capitalists

The general objective of this thesis is to improve the understanding of how specific characteristics of Venture Capital (VC) backing explain the economics of the Initial Public Offering (IPO) market. We choose to carry our analysis on the French market in order to expand the European evidence, which is still relatively scarce compared to the US one, but also because the French VC market is one of the most active markets in Europe. Within this setting one of our principal objectives is to check the validity of previously documented characteristics of VC-IPO backing but also to test some novel ones. Thus we distinguish between different types of VC funding, namely: Early-stage funding, Development funding, and the funding of Management buy-outs and buy-ins (MBO/MBI), and investigate, to our knowledge for the first time, the impact of VC specialisation on the performance of IPOs. Our second main objective is to provide an alternative investigation of the long researched and still controversial issue of VCs' certification, by looking, again for the first time, at the association between the presence of VCs and the quality of IPO prospectus forecasts. Finally, we provide the first analysis for France of the determinants of the lock-up choice and the performance of IPOs at lock-up expiry, placing emphasis on trying to establish the virtue or otherwise of VC backing. First, we find that IPOs with a potential conflict of interest between sponsors with an affiliated VC and investors are associated with more underpricing. We argue that the latter may be evidence that the market fears the opportunistic behaviour of affiliated sponsors and as a result requires greater compensation to take shares of those firms. However, there is no evidence in our data that those EPOs issue more opportunistic forecasts or that their stock market performance is worse than other IPOs over the long-run. Second, we uncover substantial differences between the characteristics of VC backing at IPO time as a function of the type of funding received. MBO/MBI and Early-stage-IPOs both have substantial degrees of VC involvement but only the former are associated with lower underpricing. It could be that the positive effect on investors' sentiment of good monitoring for Early-stage-VC-IPOs is offset by the negative adverse selection effect identified in the literature. In line with the idea that investors may be concerned by adverse selection issues in Early-stage-VC-IPOs we find a large price drop at the first lock-up expiry of those IPOs where VCs are unlocked. However, in the long-run we find no evidence that Early-stage-VC-IPOs perform worse on average. The evidence of lower underpricing for MBO/MBI-IPOs is not the only finding that tends to suggest that deeper VC involvement matters. When VCs have a blocking minority interest we find that the prospectus forecast displays more prudence. Although not associated with less underpricing, we find that the reputation of VC backers may also proxy for the better certification ability of VCs since the latter back firms that tend to issue more accurate forecasts. In contrast, IPOs backed by low reputation, less experienced, VCs who may be keen to build a reputation for bringing firms to the market are found to issue less accurate forecasts. Finally, we also show that the presence of VCs or rather VC sub-groups is not the only characteristic of the capital and ownership structure of IPO firms at work. For instance, we report that managing shareholders are more likely to enter a discretionary lock-up agreement when their shareholding is going to be substantially diluted as a result of the IPO. Moreover, in the long-run there is evidence that the performance of IPOs increases with the post-IPO shareholding of entrepreneurs. To counterbalance the latter, however, we also find that more specialisation by VC backers improves long-term performance. If specialised VCs are able to make better investment decisions and monitor portfolio firms more effectively, our evidence suggests that this is not anticipated by the market at IPO time since the IPOs of specialist VCs are not associated with less underpricing.