THE IMPACT OF SEPARATED VOTING AND CASH FLOW RIGHTS ON FRENCH IPO VALUATION: A PRINCIPAL AGENT PERSPECTIVE

Transcription

1 Frontiers of Entrepreneurship Research Volume 31 Issue 11 CHAPTER XI. GOVERNANCE Article THE IMPACT OF SEPARATED VOTING AND CASH FLOW RIGHTS ON FRENCH IPO VALUATION: A PRINCIPAL AGENT PERSPECTIVE asma fattoum EMLYON Business School, afattoum@em-lyon.com Frederic Delmar EMLYON Business School Recommended Citation fattoum, asma and Delmar, Frederic (2011) "THE IMPACT OF SEPARATED VOTING AND CASH FLOW RIGHTS ON FRENCH IPO VALUATION: A PRINCIPAL AGENT PERSPECTIVE," Frontiers of Entrepreneurship Research: Vol. 31: Iss. 11, Article 1. Available at: This Paper is brought to you for free and open access by the Entrepreneurship at Babson at Digital Knowledge at Babson. It has been accepted for inclusion in Frontiers of Entrepreneurship Research by an authorized administrator of Digital Knowledge at Babson. For more information, please contact digitalknowledge@babson.edu.

2 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 351 THE IMPACT OF SEPARATED VOTING AND CASH- FLOW RIGHTS ON FRENCH IPO VALUATION: A PRINCIPAL AGENT PERSPECTIVE Asma Fattoum, EMLYON Business School, France Frédéric Delmar, EMLYON Business School, France Abstract This paper examines how the use of mechanisms separating voting rights from cash flow rights such as dual share classes, pyramid structures and voting pacts agreements influence the valuation of firms at IPO. We claim that these mechanisms, adopted by founder-ceos to insulate them against the market for corporate control and to maintain private benefit extraction, have negative effects on IPO valuation. However, we also claim the negative impact of these disconnecting mechanisms on IPO valuation to be moderated by (1) firm size and age and (2) founder-ceo s status and equity retained at the IPO. We examine our claims using a unique hand-collected dataset of all 258 IPO s involving founder-ceos completed in the French capital markets from 1995 to We find support for our first claim. We found only mixed support for our second claim with retained equity moderating the effect. Introduction An initial public offering (IPO) is a critical mile stone for an entrepreneurial firm. To advance a privately held firm to the public market provides not only access to substantial financial resources needed for future investments but it also enhances the firm s and the founder-ceo s prestige, reputation and visibility (Nelson, 2003). However, this transition may expose founder-ceos to several threats. Indeed, going public increases pressure on CEO-founders as new investors are likely to exercise strong control over them, to require full disclosure of sensitive information and to claim important dividend payout (Roell, 1996). More critically, at the IPO founder-ceos are no longer able to choose new shareholders, ownership structure becomes more diluted, extraction of private benefits at the expense of minority shareholders becomes more costly, and as a result, founder-ceos may run the risk of losing control over the firm (Corman & Catalan, 1989). To insulate against this potential loss of control, founder-ceos may implement several mechanisms disconnecting voting rights from cash flow rights. These mechanisms include dual-shares classes, pyramid structures, and majority shareholders pact agreements. Such devices reinforce the founder-ceo control over firm corporate governance and decrease the pressure from the market for corporate control, from internal corporate governance monitoring structures and from other new minority shareholders. More importantly, disconnecting mechanisms may help the founder- CEO to take decisions regarding firm size and scope that are not in line with new minority shareholders interests and reduce the cost of private benefits extraction. These private benefits may take the form of excessive compensation, related parties transactions or special dividends (Cronqvist & Nilsson 2003; Claessens et al., 2002). Such mechanisms are seen as problematic form the perspective of the public market as this might lead to opportunistic behaviour from the side of the Frontiers of Entrepreneurship Research

3 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 founder-ceo. Hence, principal agent theory suggests that the use of such mechanisms should be punished by the market. The market will to offset this risk by valuing the firm at a lower price than if the mechanisms where present. Hence, while the use of mechanism disconnecting voting form cash-flow rights might be attractive to the founder-ceo, the market is likely to punish such behavior to limit their risk according to principal agent theory. However, both stewardship theory and signalling theory suggest that the initial reaction of investor to lower the value of the IPO firm can be partly offset by characteristics of the firm and specific behavior adopted by founder-ceo to offset the risk of opportunistic behaviour. This raises several interesting theoretical questions about the role of such mechanism at IPO. For instance, are firms with founder-ceos more likely to use disconnecting mechanisms, such as dual share classes, pyramid structures, and voting pacts agreements, than non founder-ceos? Finally, does the adoption of disconnecting mechanisms impact negatively firm IPO valuation, and as a result, lead founder-ceos to leave money on the table at the IPO? Does the strength of this impact, if any, depends upon some firm specific and founder specific characteristics? Previous studies have largely focused on only one disconnecting mechanism, have overlooked the role played by founder-ceos, and have taken as a primary empirical context common law countries. Empirically, this is even more interesting when we know that the use of disconnecting mechanisms varies considerably across countries. Such mechanisms are rarely used in the United States and the United Kingdom, while they are popular in France and Italy (Zingales, 1994, La Porta et al., 1998). The propensity to adopt disconnecting mechanisms is linked to the protection minority shareholders (Bruton et al., 2010; la Porta et al 1998, Nenova, 2003). The limited protection awarded to minority shareholders as well as the resulting wide use of disconnecting mechanisms in French civil law countries, makes France an ideal setting to examine the impact of those mechanisms on both firm and founder-ceo welfare. This paper attempts to address this gap in the context of the French capital markets, through the investigation of all 258 IPOs undertaken by founder-ceos from 1995 to We claim that disconnecting mechanisms which are more likely to be put in place in firms with founder-ceo are negatively related to firm IPO valuation, but there are some important moderators. This paper is structured as follows. The first section presents our main theoretical arguments suggesting that founder-ceo s use of disconnecting mechanisms has a negative effect on firm IPO valuation, and that this negative effect is moderated by both firm s and founder-ceo s characteristics. The second and third sections describe the methodology used to test our hypotheses and the major findings of our regressions models. Finally, we conclude by discussing theoretical and empirical implications of our study. Theory and Hypotheses The following statements represent the cornerstones of our theoretical framework: The founder-ceo might have an incentive to implement mechanism to disconnect cash flow rights from voting rights at the IPO for two reasons. First, such mechanisms allow founder-ceo to maintain control over major strategic decisions in shareholders meetings even after reducing ownership at the IPO. Second, disconnecting mechanisms pro- Posted at Digital Knowledge at Babson 2

4 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 353 vide founder-ceos immunity against takeovers and legal actions initiated by minority shareholders, which facilitates private benefits extraction. The market at IPO might punish the use of such mechanisms by valuing the firm at a lower price (underpricing) because it leads to lower protection for minority stakeholders, increase information asymmetry and the risk of opportunistic behavior. This is especially the case in France which offer little legal protection to minority stakeholders therefore forcing them to use the price mechanism to insure themselves against such risks. The risk of receiving a lower price at IPO might be moderated by (a) firm characteristics and (b) CEO-founder behavior. Stewardship theory suggests that this negative impact is reinforced with increasing firm size and age as founder-ceos are more likely to behave opportunistically whit increasing firm age and size. Signaling theory suggests that this negative impact diminishes with the quality of the founder-ceo s human and social capital and with the amount of shares he/she retains after the IPO. This is because founder- CEO s human and social capital and detained shares are strong signals of reputation and commitment offsetting the risk of opportunistic behavior. These three arguments are developed in the following sections focusing on how the factors moderating the valuation at IPO when mechanisms separating voting form cash-flow rights are used. France is here especially interesting as the use of such mechanisms is popular and not legally prohibited. The impact of the legal framework on the use of disconnecting mechanisms The use of disconnecting mechanisms varies considerably across countries. They are rarely used in the United States and the United Kingdom, while widespread in France and Italy (Zingales, 1994, La Porta et al., 1998). Several studies have showed that the propensity of disconnecting mechanisms adoption around the world is strongly linked to the characteristics of the institutional context protecting minority shareholders (Bruton et al., 2010; la Porta et al 1998, Nenova, 2003). For example, La Porta et al., (1998) found that French civil law provides an internationally unique low level of protection for minority shareholders, while common law countries offer more rigorous protection for minority shareholders. More specifically, France scores only 3 on la Porta et al., (1998) investors protection index while the United States and the United Kingdom obtain a score of 5. Similarly, the World Bank (2008) index which measures investor protection rank the United Kingdom at a higher position (score of 8) than France (score of 5.3). Using a similar index, Nenova (2003) found that general investor protection is poorer in France than in other common law countries. The limited protection awarded to minority shareholders as well as the resulting wide use of disconnecting mechanisms in French civil law countries, makes France an ideal setting to examine the impact of those mechanisms on both firm and founder-ceo welfare. The use of mechanism is widespread, the legal protection is low for minority stakeholders and therefore the market place must most likely adjust for the risk of opportunistic behavior through the use of the price mechanism. This opportunistic behavior is driven by the fact that private benefits are fully captured by the founder-ceo while efficiency gains generated by value maximizing strategies will be shared with other shareholders in the form of residual claims. Such private benefits extraction may consist of non-pecuniary consumption, private utilization of scarce resources, disproportionate compensation levels and special dividends. The incentive and ability to extract private benefits is constrained Frontiers of Entrepreneurship Research

5 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 by the cost of expropriation. This cost varies across countries such as in countries with legal environments offering weak protection to minority shareholders, the cost of expropriation is limited, and hence the likelihood of the founder-ceo engaging in private benefits extraction should be high (Claessens et al., 2000). On the other hand, in countries with legal environments providing strict enforced rules buffering minority shareholders, the cost of expropriation is high, and the likelihood of the founder-ceo behaving opportunistically should be limited. La Porta et al., (1998) examined the set of legal rules protecting minority shareholders and the severity of enforcement of these rules in 49 countries. Their results suggest very important variations across these countries on both dimensions, with common law countries having the strongest protection to minority shareholders, German and Scandinavian civil law countries offering moderate levels of protection and, finally, French civil law countries providing the lowest level of protection for minority shareholders. The impact of country s legal protection on private benefits extraction has been confirmed by Zingales (1994, 1995), which found that private benefit extraction by controlling shareholder/manager is at least 30 percent in Italy while it is only 4 percent in the U.S.A. The weakness of legal protection offered by French civil law countries to minority shareholders is materialized along several dimensions. For instance, the ability of minority shareholders to call an extraordinary meeting is considerably constrained by the high percentage of share capital required. Moreover, the ability of minority shareholders to challenge directors decision in court is very much limited by law. Other mechanisms, often forbidden in common law countries but legally accepted in French civil law countries facilitate private benefits expropriation by entrenched the founder-ceo as they separate cash flow rights from voting rights. Mechanisms disconnecting voting rights from cash flow rights and IPO valuation The most popular mechanism is the dual-class shares structure, which implies that the firm issues shares with different voting rights. For instance, a firm may issue two classes of shares. A first class of shares (A) which provide 1 voting right for the owner and a second class of shares (B) which provide, for example, 10 voting rights for the owner (Nenova, 2003; Villalonga, & Amit, 2006). Dual-class shares are generally established through IPO or recapitalizations. There are two main arguments generally suggested by founder-ceo to justify use of dual class shares. First, dual class shares allow founder-ceo to remain in command over corporate control, and a result, be able to make long term investments without being concerned by hostile reaction of risk-averse investors (DeAngelo and DeAngelo, 1985). Second, dual-class device make successful attacks by external raiders less likely, particularly, in undervalued high growth firms (Lehn, et al., 1990). Issuing classes of stock with differential voting rights may allow the founder-ceo, to largely disconnect their stock ownership from their voting rights ownership. In this case, the founder-ceo may keep full control of the firm without committing to additional capital investment in the firm s equity (Hoi and Robin, 2010). Pyramid control structures, which allow the founder-ceo to hold shares in the firm through one or more intermediate structures such as trusts, funds and holdings, are other mechanisms often used by founder-ceo to decouple firm controlling structure from ownership structure (Villalonga, & Amit, 2006; Faccio and Lang, 2002; Claessens et al., 2002). By using pyramid structures, the founder-ceo is able to keep control of the firm through a network of holding companies at a lower cost than direct ownership, particularly, if the holding companies are located at the top of the pyramid (Boubaker & Labegorre, 2008, Bebchuk et al. 2000). Finally, voting pact agreements Posted at Digital Knowledge at Babson 4

6 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 355 involving the founder-ceo and family members may also be used to enhance the level of control in the hand of the founder, which in turn, further facilitates private benefit expropriation. Dual class shares, pyramid structures and voting pact agreement may decouple stock ownership from voting rights ownership, and as a result, accentuate the entrenchment of the firm founder-ceo from the market for corporate control (Villalonga, & Amit, 2006). For instance, using these mechanisms a founder-ceo may stay in place as the CEO of the firm even in the situation in which firm value maximization would suggest replacing him/her by another CEO. Similarly, such mechanisms may allow founder s family to keep control of executive positions in the firm even though it is in the best interests of minority shareholders to recruit outsider executives from an efficient labor market. Hence, potential investors are likely to pay close attention to the presence of mechanisms disconnecting cash flow rights from voting rights in their valuation of the firm in the context of an IPO. More specifically, potential investors are likely to discount the value of the firm at the IPO as a result of the higher potential of agency costs. Therefore, we suggest that: Hypothesis 1: There is a negative relationship between the founder-ceo use of disconnecting mechanisms and firm IPO valuation. Although investors should interpret separation between cash flow rights and voting rights at the IPO, as an indication of a serious opportunity for entrenched founders to take advantage of private benefits, several contextual factors may moderate the magnitude of such under-valuation. In the following two sections, we develop a rationale suggesting that IPO valuation may be moderated by firm specific factors such as size and age as well as founder-ceo specific characteristics like human and social capital as well as retained equity at the IPO. The moderating effect of firm characteristics Stewardship theory suggests the founder-ceo is unlikely to engage in opportunistic individual goals such as private benefits expropriation at the expenses of minority shareholders, but rather he/she tends to behave as steward, especially when the firm is young and small (Le Breton Miller and Miller, 2009; Davis et al., 1997). The main argument is that founders generally view their firms as an extension of themselves, and therefore, are likely to behave as stewards putting the welfare of their company higher in the hierarchy than their own welfare and self-interest (Corbetta and Salvato, 2004; Hambrick & Mason, 1984). Founders perceive their personal prestige and selfesteem are closely tied to the ones of their firms. They are also highly committed and attached to their firms (Davis et al., 1997). Accordingly, founders are likely to exert an extraordinary power to increase value creation for their firms and are less likely to expropriate minority shareholders significant private benefits, a behavior which risks to damaging their long-term career as managers. For example, Wasserman (2006) found that steward founders accept lower cash compensation than non-founder CEOs. Stewardship theory suggests that founder-ceo may refrain from opportunistic behavior, it also suggests that the likelihood of such stewardship behavior decreases as firm size increases and as firm ages (Wasserman, 2006). Indeed, stewardship theory suggests that founders behavior is closely related to their level of attachment, commitment and psychological identification with their firms (Hambrick & Mason, 1984). These levels of attachment, commitment and psychological identification are likely to be stronger during initial phases of firm life as founders are involved Frontiers of Entrepreneurship Research

7 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 in all firm operations. However, as the firm grows formal routines systems and standard procedures as well as control systems are crated or simply emerge, which reduces the level of involvement of the founder in the firm. Moreover, as the company s size increases the complexity of managing the firm become more important and time consuming, and hence, it reduces the level of involvement of the founder in the day-to-day management of firm operations. Accordingly, founder s intrinsic motivation and psychological commitment may decrease with firm size and age, and hence, his/ her degree of stewardship is also likely to decrease. For instance, Wasserman (2006) found that the difference in cash compensation between founder and non-founder CEOs is reduced by firm size and age. Moreover, the opportunities for substantial private benefits expropriation as well as costs associated with monitoring controlling shareholders/ founders are likely to increase with firm size and age. Thus, as founder-ceo s stewardship behavior is more likely to be observed in small and young firms in which founder s level of attachment, commitment, involvement and psychological identification is high, and as, this stewardship behavior is likely to decrease with firm increasing size and age, we suggest the following hypotheses: Hypothesis 2: Firm size moderates the impact of founder-ceo use of disconnecting mechanisms on firm IPO valuation: the larger the firm, the greater the negative impact of disconnecting mechanisms on firm IPO valuation. Hypothesis 3: Firm age moderates the impact of founder-ceo use of disconnecting mechanisms on firm IPO valuation: the older the firm, the greater the negative impact of disconnecting mechanisms on firm IPO valuation. The moderating effect of founder-ceo status and behavior Signaling theory suggests that the founder-ceo may play an important role in altering the magnitude of initial public offering valuation by his or her status and behavior. More specifically, founder s human and social capital as well as the importance of his/her shareholding after the IPO may affect firm valuation (Cohen & Dean, 2005) because they are strong signals that the founder- CEO is less likely to behave opportunistically. In attempting to assess firm value at the IPO, investors face important information asymmetry problems. Indeed, on the one hand, incumbent shareholders have an incentive to over-evaluate the quality of the assets and the future prospects of their firm. On the other hand, liability of market newness implies that investors have little information on the real quality of firm s assets, have difficulties evaluating precisely its expected future returns, and its ability to satisfy successfully the expectations of financial markets (Certo, 2003). This difficulty increases when incumbent shareholders misrepresent available information and when regulation does not require full disclosure of information on firm s assets and prospects. Under such conditions, investors may rely on social indicators of firm value rather than economic indicators, which are often manipulated or misrepresented by incumbent shareholders (Carter & Manaster, 1990; Carter et al., 1998). In the context of an initial public offering, investors may formulate their valuation of the firm on indicators capturing the status of founder-ceo rather than solely financial information (Certo, 2003; Podolny, 1994). The status of founder-ceo may represent a reliable signal for investors in evaluating firm future expected returns because it combines two important criteria: it is observable by all potential Posted at Digital Knowledge at Babson 6

8 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 357 investors and it is costly to imitate (Spence, 1973). It is observable because potential investors can get access to rich information on founder-ceo s education, experience and external board mandates in public documents, such as IPO prospectus. It is costly to imitate because founder- CEO who lack degrees from prestigious educational institutions, or industry specific experience or strong ties to board interlock networks cannot convey to potential investors signals of rich human and social capitals (Certo, 2003). As such, founders status may represent a credible signal to the financial market about their ability to use own experience, expertise and social capital to create value for the firm (Certo et al., 2001; Nelson, 2003). Even in the case in which private extraction opportunities are readily available for the controlling founder through dual share classes, pyramid structures and cross equity ties, investors may perceive that founder s rich human and social capital may generate efficient investments that may offset private benefits extraction. In other words, as long as investors perceive that potential value created by founder-ceo s rich human and social capital minus private benefit extraction undertaken by the founder-ceo is larger than the amount of value potentially created by a professional outsider manager with lower managerial talents, the negative impact of mechanism decoupling ownership structure (cash flow rights) from controlling structure (voting rights) in favor of founder-ceo will be lower. Therefore, we suggest the following hypothesis: Hypothesis 4: Founder-CEO s human and social capitals moderate the impact of his/her use of disconnecting mechanisms on firm IPO valuation: the greater founder-ceo s human and social capitals, the lower the negative impact of disconnecting mechanisms on firm IPO valuation. An additional factor, which may attenuate the negative effect of disconnecting mechanisms on firm valuation, is the behavior of the EO-founder as indicated by the amount of shares retained by the founder-ceo at the IPO. Indeed, founder- CEO s shares ownership may provide a signal to the financial market that he/she is unlikely to engage in private benefits extraction (Gomes, 2000). This signal is likely to be positively received by potential investors because in the case where the founder-ceo engages in important expropriation of private benefits just after going public, firm stock market valuation will decrease accordingly, and the founder-ceo remaining stocks will sell at lower prices (Gomes, 2000). Thus, founder-ceo stock ownership may represent a credible signal for reduced opportunistic behavior because it is readily observable and it is also costly to imitate (Spence, 1973). That is, the cost of private benefit expropriation on founder-ceo s welfare, represented by the discount in market valuation of his/her remaining stocks, is significant. Therefore, we suggest the following hypothesis: Hypothesis 5: Founder-CEO s retained equity moderates the impact of his/her use of disconnecting mechanisms on firm IPO valuation: the greater retained equity, the lower the negative impact of disconnecting mechanisms on firm IPO valuation. Sample and Data Collection Database & Methodology We test our hypotheses using a unique dataset of all French IPO s from January 1995 to December Out of a total 429 IPOs, 258 IPOs involved founder-ceos and 171 IPOs were led by non-founder CEOs. Table 1 describes the major characteristics of these two groups. Firms involving founder-ceos are considerably smaller and younger than other firms. They are more likely to use disconnecting mechanisms, particularly dual class shares and pyramid structures. In terms of ownership structure, four elements could be highlighted. Founder-CEOs retain more Frontiers of Entrepreneurship Research

9 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 equity at the IPO than non founder CEOs. In contrast, venture capital and private equity shareholding is lower in firms involving founder-ceos than other firms. There is no difference between the two groups in terms of public equity released at the IPO. Finally, founder-ceos are more likely to combine executive and chairman positions than non founder-ceos (81 percent vs. 61 percent). Because we focus on the impact of founder-ceos use of disconnecting mechanisms on IPO valuation, we restrict our sample to the 258 IPOs involving founder-ceos. Our primary data sources for IPO valuation include Datastream, Bloomberg and Thomson One banker databases. We collected data on founder-ceo equity, venture capital/private equity shareholding, public equity released, and CEO duality from IPO prospectus and reference documents filed with the French financial markets authority (AMF). We obtained founder-ceos biographies from the Who s Who in France and Factiva. Measures IPO Valuation is our dependent variable. Following previous research, we measure IPO valuation using IPO underpricing (Barry et al., 1990; Ritter and Welch, 2002; Ljungqvist, 2007; Certo et al 2009; Certo et al, 2007; Boulton et al 2010; Filatotchev & Bishop, 2002). We defined IPO underpricing as the percentage difference between the end of the third day of trading stock price and the offer price, adjusted for market index return: IPO Underpricing = ([(P1 P0) / P0] [(M1-M0) / M0]) x 100 Where: P1 is the firm s stock price value at the end of the third trading day. P0 is the firm s stock offer price, the price at which shares are initially offered to investors. M1 is the closing value of the market index SBF250 observed on the third trading day. M0 is the opening value of the market index SBF250 observed on the first trading day. Disconnecting Mechanisms is our independent variable. Three main devices can be used by founder- CEOs to disconnect cash flow rights from voting rights. Table 1 shows that the most popular device is dual class shares (66 percent) with multiple voting rights followed by pyramid structures (38 percent) and voting pact agreements (20 percent). Founder-CEOs may adopt simultaneously several disconnecting devices. In our sample, 19 percent of founder-ceos did not implement disconnecting devices. In contrast, 45 percent of founder-ceos used one mechanism, 31 percent combined two mechanisms and only 5 percent put in place the three mechanisms concurrently. We operationalized disconnecting mechanisms using two dummy variables. The first dummy variable labeled one mechanism takes the value of 1 if founder-ceo adopts one disconnecting mechanism and 0 otherwise. The second dummy variable labeled two or plus mechanisms takes the value of 1 if founder-ceo adopts at least two disconnecting mechanisms, and 0 otherwise. The group of firms not implementing any disconnecting mechanisms is the reference group. Our hypotheses are based on a number of interactions related to firm and founder characteristics. We captured firm characteristics using firm size and age. Firm size was measured using the logarithm of firm total capitalization on the basis of the offer price in euros (Certo et al. 2001; Filatotchev, 2006; Bruton et al., 2010). Firm age was defined as the number of years separating its founding date and IPO date. To measure founder-ceo s human and social capital we relied primarily on signals capturing his/ her education. In line with previous research, we considered that founder CEO has valuable human and social capital if he/she graduated from prestigious educational institutions or has earned an MBA degree (Maclean et al., 2006; Pollock et al., 2009). We measured founder CEO s human and social capital using a dummy variable which takes a value of 1 if he/she holds an MBA or another degree from a Posted at Digital Knowledge at Babson 8

10 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 359 prestigious institutions, as classified by Maclean et al. (2006) and 0 otherwise. To measure the extent of equity kept by founder-ceo at the IPO, we relied on the percentage of cash flow rights held by founder- CEO the first trading day (Gomes, 2000). Since IPO underpricing may be impacted by numerous factors other than devices disconnecting cash flow rights from voting rights, several control variables well-known from previous research were introduced in our statistical analysis. More specifically, we controlled for stock market conditions, industry effect, the amount of public equity released, venture capital and private equity shareholding, underwriter reputation, IPO procedure and CEO duality. Descriptive statistics and Pearson correlation coefficients for all variables included in our model are presented in Table 2. Results To test our hypotheses, we used multiple linear regression models. The results of linear regression models are presented in table 3. Model 1included all control variables capturing stock market conditions, industry and period effects, the amount of public equity released, venture capital and private equity shareholding, underwriter reputation, IPO procedure and CEO duality. The only control variable to have a significant effect on IPO underpricing was the industry dummy for consumer services sector. This industry had lower IPO underpricing than firms competing in the technology sector. In model 2, we added the four variables forming the basis for our interactions in later models (firm age and market value as well as founder-ceo s human and social capital and equity ownership at the IPO). We found IPO underpricing to be lower for firms led by founder- CEO s endowed with high human and social capital than firms which do not. This result highlights the strong impact of founder-ceo education on increasing firm IPO valuation. The following models test our hypotheses. In model 3, we introduced two dummy variables measuring the number of devices separating voting from cash flow rights used by founder-ceos. We found both dummy variables to be statistically significant at (β= 9.70; p<.01) and (β= 8.07; p<.05) respectively. This result suggests that IPO underpricing is significantly higher when founder-ceos adopt one or more disconnecting mechanisms than when they do not. This finding provides strong support for H1. In models 4 to 7, we provide tests for the interaction effects between disconnecting mechanisms and firm age and market value, founder-ceo s human and social capital and equity ownership at the IPO by adding corresponding interaction terms in each model. Models 4, 5 and 6 provided no support for H2, H3 and H4. There are no strong moderating effects of firm age and size as well as founder-ceo human and social capital on the disconnecting mechanisms IPO underpricing relationship. This result suggests that these firm and founder specific characteristics are not sufficient to offset the negative impact of disconnecting mechanisms on IPO valuation. Finally, in model 7 we added the two terms capturing interaction between founder-ceo retained equity and the use of disconnecting mechanisms. The increase in explained variance of model 7 compared model 3 by 2.33 percent is significant (p<0.05). In addition, the interaction term CEO equity x One mechanism is significant and negative (β=-0.29; p<0.05. This result suggests that the higher founder-ceo retained equity, the lower the negative impact of using one disconnecting mechanism on firm IPO valuation. The effect of the interaction term CEO equity x two or plus mechanism is not statistically significant at p<0.05, which suggests that higher founder-ceo equity retained at IPO is not Frontiers of Entrepreneurship Research

11 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 sufficient to compensate for the negative effects on IPO valuation related to using two or more disconnecting mechanisms. This result provided partial support for H5. Discussion The main purpose of this study is to investigate how founder-ceos cope with threats related to potential loss of control of their firms resulting from the IPO. More specifically, this paper examined whether firms with founder-ceos are more likely to use dual share classes, pyramid structures, and voting pacts agreements than non founder-ceos. It also assessed the impact of disconnecting mechanisms on firm IPO valuation and tested whether this impact depends upon firm specific and founder-ceos specific characteristics. The objective is first, to better understand whether, why and how founder-ceos implement specific strategies and governance practices to deal with IPO threats and challenges, and second to evaluate the cost of such strategies and practices on founder-ceos welfare. By examining data drawn from a sample of all 258 IPOs involving founder-ceos undertaken in the French capital markets over sixteen years period, this study shows that disconnecting mechanisms are implemented more often in IPO firms with founder-ceos than in other IPOs. Indeed, over the period under study only 19 percent of IPO firms with founder-ceos did not use disconnecting mechanisms while 45 percent of other IPO firms refrained from using such devices. Moreover, 36 percent of IPO firms with founder-ceos implemented at least two disconnecting mechanisms, a ratio which is 50 percent higher than other IPOs. These results provide strong indication that dual share classes, pyramid structures, and voting pacts agreements are popular mechanism used by founder-ceos to isolate themselves from the market for corporate control. The findings of our study suggest, however, that the use of those popular devices by founder- CEOs does not come without costs. The valuation of IPOs with founder-ceos putting in place one, two or more disconnecting mechanisms are significantly lower than IPOs with founder-ceos not implementing disconnecting mechanisms. This result suggest that disconnecting governance practices used by founder-ceos to maintain control over the firm at the IPO are sanctioned by the capital markets, since founder-ceos need to undervalue the issue price to attract more investors. In other words, founder-ceos has to leave money on the table in return of maintaining control over the firm. The discounted stock issue price of firms using disconnecting mechanisms explains why minority shareholders accept to bear important agency costs created by those mechanisms and buy new stocks at the IPO. The financial sanction suffered by founder-ceos using disconnecting mechanisms does not seem to be attenuated or exacerbated by firm-specific characteristics such as size and age. Similarly, although our results suggest that founder-ceo s human and social capital has a positive direct impact on IPO valuation, founder-ceos status does not moderate the financial sanction inflicted by capital markets to founder-ceos using disconnecting mechanisms. However, the amount of founder-ceo s equity retained at the IPO seems to be perceived by capital markets as a credible signal capturing the extent of founder-ceos opportunistic behaviour if only one disconnecting mechanism is used. If two or more disconnecting mechanisms are adopted by founder-ceos, IPO valuation suffers even at high levels of retained equity. In conclusion, this study offers several contributions to entrepreneurship theory and practice. First, it provides an overview of the propensity of disconnecting mechanisms implementation at IPOs over the last sixteen years in France. Second, it sheds some light on the impact of dual share Posted at Digital Knowledge at Babson 10

12 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 361 classes, pyramid structures, and voting pacts on IPO valuation. Third, it examines whether firm specific and founder specific characteristics alter the perception of potential investors of the severity of agency problems generated by disconnecting mechanisms. Finally, it provides an explanation for why minority shareholders are willing to accept agency costs generated by the use of such a procedure disconnecting cash flow rights from voting rights. CONTACT: Asma Fattoum; afattoum@em-lyon.com; (T): ; (F): ; EMLYON Business School, 23 avenue Guy de Collongue, F Ecully Cedex, France. Selected References Full References Available From Corresponding Author Bruton, G.D., S. Chahine, & I. Filatotchev. (2009). Founders, Private Equity Investors, and Underpricing in Entrepreneurial Ipos. Entrepreneurship Theory Practice, 33(4): Certo, S. T. (2003). Influencing IPO Investors With Prestige: Signalling With Board Structures. Academy of Management Review, 28: Claessens, S., S. Djankov, J. P. H. Fan, & L. H. P. Lang. (2002). Disentangling The Incentive and Entrenchment Effects of Large Shareholdings. Journal of Finance, 57: Cronqvist, H. & Nilsson, M., (2003). Agency Costs of Controlling Minority Shareholders. Journal of Financial and Quantitative Analysis, 38: D Aveni RA. (1990). Top Managerial Prestige and Organizational Bankruptcy. Organization Science, 1: Florin, J., Lubatkin, M., & Schulze, W. (2003). A Social Capital Model of High-Growth Ventures. Academy of Management Journal, 46: Gimeno, J., Folta, T. B., Cooper, A. C., & Woo, C. Y. (1997). Survival of The Fittest? Entrepreneurial Human Capital and The Persistence of Underperforming Firms. Administrative Science Quarterly, 42: Gomes, A. R. (2000). Going Public Without Governance: Managerial Reputation Effects. Journal of Finance, 55: Hambrick D.C & Mason P.A. (1984). Upper Echelons: The Organization as a Reflection of Its Top Managers. Academy of Management Review, 9: Himmelberg, C. P., R. G. Hubbard, & D. Palia. (1999). Understanding The Determinants of Managerial Ownership and Performance. Journal of Financial Economics, 53: Jayaraman, N., Khorana, A., Nelling, E., & Covin, J. (2000). CEO Founder Status and Firm Financial Performance. Strategic Management Journal, 21: La Porta, R., F. Lopez-De-Silanes, A. Shleifer, & R. W. Vishny. (1998). Law and Finance. Journal of Political Economy, 106: Nelson, T. (2003). The Persistence of Founder Influence: Management, Ownership, and Performance Effects at Initial Public offering. Strategic Management Journal, 24: Podolny JM. (1994). Market Uncertainty and The Social Character of Economic Exchange. Administrative Science Quarterly, 39: Spence M. (1976). Informational Aspects of Market Structure: An Introduction. Quarterly Journal of Economics, 90: Villalonga, B., & Amit, R. (2006). How Do Family Ownership, Control, and Management Affect Firm Value? Journal of Financial Economics, 80: Wasserman, N. (2006). Stewards, Agents, and The Founder Discount: Executive Compensation In New Ventures. Academy of Management Journal, 49: Zingales, L. (1995). What Determines The Value of Corporate Votes? Quarterly Journal of Economics, 110: Frontiers of Entrepreneurship Research

13 Frontiers of Entrepreneurship Research, Vol. 31 [2011], Iss. 11, Art FRONTIERS OF ENTREPRENEURSHIP RESEARCH 2011 Table 1 - IPOs with founder-ceo vs. non founder-ceo characteristics IPOs founder-ceos N=258 IPOs no founder-ceos N=171 Variable Mean S.D Mean S.D Firm age Firm market value Public Equity released CEO equity VC/Private equity shareholding CEO duality Dual Class Shares CEO pyramid structures CEO pacts agreements No mechanisms One mechanism Two mechanisms Three mechanisms Table 2 - Descriptive statistics and Pearson correlations Variable Mean S.D IPO underpricing Public Equity released VC/Private equity shareholding Underwriter reputation IPO procedure (fixed price) IPO procedure (fixed and auction) CEO duality Firm age Firm market value Founder-CEO Hum/soc capital Founder-CEO equity One mechanism Two or plus mechanisms Natural log transformation. N= 258 observations. Correlations greater than.12 are significant at p<.05. Posted at Digital Knowledge at Babson 12

14 fattoum and Delmar: Separated Voting & Cash Flow Rights on French IPO Valuation GOVERNANCE 363 Table 3- OLS regression predicting IPO underpricing Model 1 Model 2 Model 3 Variable Coeff. (S.E) Coeff. (S.E) Coeff. (S.E) Intercept 3.55 (11.10) (14.63) 5.37 (14.60) Public Equity released (0.12) (0.13) (0.13) VC/Private equity shareholding (0.08) (0.09) (0.09) Underwriter reputation 0.08 (0.08) 0.07 (0.08) 0.07 (0.08) IPO procedure (fixed price) 8.52 (8.77) 9.76 (8.81) (8.77) IPO procedure (fixed and auction) 1.29 (2.82) 0.76 (2.85) 0.76 (2.82) CEO duality (3.31) (3.51) (3.48) Firm age (1.51) (1.52) Firm market value (1.48) (1.47) Founder-CEO Human/social capital * (2.89) * (2.87) Founder-CEO equity (0.07) (0.07) One mechanism 9.70 ** (3.36) Two or plus mechanisms 8.07 * (3.66) Degrees of freedom R squared Δ R squared Natural log transformation. Significance levels: ** p<.01; * p<.05.. All models include four dummy variables capturing stock market conditions as well as seven dummy variables controlling for industry effects not shown. Table 3 (continued) - OLS regression predicting IPO underpricing Model 4 Model 5 Model 6 Model 7 Variable Coeff. (S.E) Coeff. (S.E) Coeff. (S.E) Coeff. (S.E) Intercept 9.65 ( (15.08) 3.64 (15.47) 2.72 (14.71) VC/Private equity shareholding (0.09) (0.09) (0.09) (0.09) Underwriter reputation 0.07 (0.08) 0.08 (0.08) 0.07 (0.08) 0.07 (0.08) IPO procedure (fixed price) (8.87) (8.81) (8.88) 9.20 (8.72) IPO procedure (fixed and auction) 1.10 (2.90) 1.03 (2.83) 0.90 (2.83) 1.15 (2.80) CEO duality (3.52) (3.49) (3.51) (3.49) Firm age (1.53) (2.97) (1.53) (1.52) Firm market value (2.68) 0.03 (1.48) (1.49) (1.47) Founder-CEO Human/social capital * (2.90) * (2.88) (6.81) ** (2.86) Founder-CEO equity (0.07) (0.07) (0.07) 0.14 (0.12) One mechanism 0.93 (13.64) 3.69 (7.66) 9.88 ** (3.87) ** (5.47) Two or plus mechanisms 6.47 (14.31) 9.39 (8.18) 9.22 * (4.04) 8.23 (6.81) Market value * One mechanism 2.12 (3.17) Market value * Two or plus mechanisms 0.26 (3.44) Firm age * One mechanism 3.14 (3.84) Firm age * Two or plus mechanisms (3.92) CEO H/S capital * One mechanism (7.86) CEO H/S capital * Two or plus mechanisms (8.43) CEO equity * One mechanism * (0.13) CEO equity * Two or plus mechanisms (0.14) Degrees of freedom R squared Δ R squared Natural log transformation. Significance levels: ** p<.01; * p<.05. All models include four dummy variables capturing stock market conditions as well as seven dummy variables controlling for industry effects not shown. Frontiers of Entrepreneurship Research