The influence of top management team heterogeneity on the capital raised through an initial public offering.

Tenure can be viewed as a proxy for its commitment to the status quo, informational diversity, and risk propensity, and the TMT's tenure may affect organizational outcomes (Finkelstein & Hambrick, 1990). Long-tenured groups have been associated with increased cognitive rigidity and commitment to the status quo (Bantel & Jackson, 1989), standard ways of communicating (Katz, 1982), persistent strategies, and strategies that conform to those of the industry (Finkelstein & Hambrick, 1990).

Bantel (1993) argued that members with similar tenure form a cohort that influences consensus, and that homogeneity in tenure will be positively related to the TMT's ability to reach consensus on strategic decisions. However, Bantel found little support for her argument. Heterogeneity in the tenure of the TMT was found to be positively related to firm performance, strategic change, and the degree of international diversification (Hambrick et al., 1996; Murray, 1989; Tihanyi et al., 2000; Wiersema & Bantel, 1992). Hambrick et al. found tenure heterogeneity among the TMT to positively affect a firm's competitive action propensity and the scope of its competitive response. Williams and O'Reilly (1998) argued that groups with greater tenure heterogeneity have less social integration, higher turnover, and poorer communication than groups with less heterogeneity.

Potential investors may perceive TMT tenure heterogeneity as a signal that indicates whether the firm will stick to past strategies or be flexible in its strategic approaches. Greater tenure heterogeneity may lead to greater strategic flexibility and hence an ability to address the challenges experienced in transitioning from private to public ownership (Bantel & Jackson, 1989; Certo, 2003; Wiersema & Bantel, 1992). Alternatively, tenure homogeneity may lead to strategic rigidity, and hence a firm with a homogeneous (i.e., tenure) TMT may experience problems once the IPO is complete.

In addition, heterogeneity in the tenure of the TMT may be due in part to the presence of founders on the team and recent additions to the team. Founders are central to the creation of the firm (Schein, 1983), and in technology-based firms, they may be central to the invention of the technology used by the firm (Chandler & Hanks, 1998). Ucbasaran et al. (2003) found that entrepreneurial founder teams do not remain static over time and argued that the entry to and exit from entrepreneurial teams is related to the functional heterogeneity of founding teams. A team considered as balanced at start-up may not be considered balanced as the venture develops. Managers with short tenures may be additions to the team made shortly before the IPO to round out the team, especially with regard to the functions represented on the team (Ryan & Hise, 2001). These additions bring with them experience in other firms as well as a fresh view of the firm they join. However, teams that are more homogeneous, either containing only founders or new managers, may lack specific firm experiences and/or broader perspectives and contacts (e.g., Glick et al., 1993; Hambrick, 1994; Hambrick & Mason, 1984). Thus, it appears that tenure heterogeneity might provide a signal to potential investors about the quality of the IPO firm, and hence be associated with greater capital accumulations.

Hypothesis 4: Heterogeneity in the TMT's tenure is positively related to the amount of capital raised at IPO.

Methods

Sample and Data Collection

The sample used in this study are firms undergoing their IPO of stock during the time period of January 1, 1993 through December 31, 1997. In an effort to examine a large sample of young firms in a single industry, the prepackaged software industry was examined, i.e., Standard Industrial Classification (SIC) code 7372. Examining a single industry prevents industry effects from confounding relationships between the independent and dependent variables (Dess, Ireland, & Hitt, 1990). The prepackaged software industry was selected because of the large number of software firms that went public during a recent period of significant IPO activity and during a period that preceded the dot-com bubble euphoria: a period of market exuberance in which investment metrics were turned upside down (Mudambi & Zimmerman Treichel, 2005). A total of 243 U.S.-based software firms were identified by IPO Reporter and IPO Data as having an IPO during the period of January 1, 1993 through December 31, 1997.

The primary source of data used in this study was the firm's IPO prospectus. The IPO prospectus discloses information valuable to regulators, investors, underwriters, and other relevant parties (Beatty & Zajac, 1994). A number of studies have used IPO filings as a source of primary data (e.g., Beatty & Zajac, 1994; Daily et al., 2003; Deeds et al., 1997, 2004; Lester et al., 2006; Welbourne & Andrews, 1996). I was able to obtain IPO prospectuses for 172 of the firms either directly from the firm (I contacted them by letter and/or by phone to request a copy of their IPO prospectus) or by purchasing the prospectus from IPO Data. I was unable to secure prospectuses from the remaining 71 firms (29%). The sample used in this study (172 firms) was representative of the larger population: The average IPO value of the population was $30,616,536, and the average IPO value of firms in the sample was $32,659,133.

Dependent Measure

IPO Value. The dependent variable used in this study is the capital raised at IPO, measured as the total value of the capital for the firm raised through the firm's IPO less the underwriters' fees as noted on the cover page of the firm's prospectus (Deeds et al., 1997, 2004; Finkle, 1998; Gulati & Higgins, 2003). It is a measure not only of IPO performance (Gulati & Higgins, 2003) but also of how the market values a company at the time of its initial offering (Deeds et al., 2004; Finkle, 1998).

Independent Measures

The TMT was defined as those managers listed in the prospectus as composing the firm's management team (Lester et al., 2006; Shrader, Oviatt, & McDougall, 2000). This definition includes all of the C-level positions, e.g., CEO, chief financial officer, chief operating officer, as well as vice presidents, senior vice presidents, and other managers listed in the management section of the prospectus. Including these members enables us to include the most important organizational decision makers (Murray, 1989; Tihanyi et al., 2000). Data on the top managers were obtained from the managers' biographies presented in the IPO prospectus. TMT heterogeneity data were coded using the following guidelines:

Functional Heterogeneity. Functional background heterogeneity was calculated using Blau's (1977) heterogeneity index (1-[summation][i.sup.2]), where i is the proportion of the group in the ith category. A high score indicates variability in the functional backgrounds among team members, i.e., functional heterogeneity, and a low score represents homogeneity (Smith et al., 1994). The functional categories used to calculate the index were those functional categories frequently used in the study of heterogeneity--finance, human resources, general management, marketing/public relations, operations, engineering/R&D, strategic planning, and law (Boeker, 1988; Murray, 1989; Tihanyi et al., 2000), to which I added executive, information technology, and other.

Educational Heterogeneity. Educational background heterogeneity was measured using Blau's (1977) heterogeneity index as described earlier. I used the eight educational background categories used by Hambrick et al. (1996), i.e., engineering, science, business administration, economics, liberal arts, law, business other, and other. As was done by Wiersema and Bantel (1992), individuals with BS or MS degrees were classified as science specialists unless a discipline was listed.

Age Heterogeneity. Age heterogeneity was calculated as the coefficient of variation of the top managers' age (Murray, 1989; Richard & Shelor, 2002; Tihanyi et al., 2000), where a high score indicates age heterogeneity and a low score indicates age homogeneity.

Tenure Heterogeneity. Tenure heterogeneity was calculated as the coefficient of variation of the top managers' tenure (Murray, 1989; Tihanyi et al., 2000), where a high score indicates age heterogeneity and a low score indicates age homogeneity.

Control Variables

Year of IPO. Year of IPO was measured as the year in which the IPO took place beginning with 1993 as the base year--"l" for 1993, "2" for 1994, "3" for 1995, "4" for 1996, and "5" for 1997. It is used to control for the development of the software industry and the industry legitimacy that develops over time. According to Zimmerman and Zeitz (2002), the level of industry legitimacy is related to the ability of the firm to secure resources, i.e., capital.

Hot Market. The effect of periods of increased market activity has been shown to positively influence the IPO of stock. Hot Market was used to identify firms that went public during years of high IPO activity (Deeds et al., 1997, 2004; Ritter, 1984). The years 1993 and 1996 were two hot markets for IPOs (http://www.marketdata.nasdaq.com/asp/ Sec3IPO.asp). If the firms in the sample went public during either 1993 or 1996, they were coded as "1" and "0" otherwise (Deeds et al., 1997, 2004). This is similar to the period effect measured by Carpenter (2002).