SEQUENTIAL DIVESTITURE THROUGH INITIAL PUBLIC OFFERINGS Jeffrey J. - - PDF document

SEQUENTIAL DIVESTITURE THROUGH INITIAL PUBLIC OFFERINGS Jeffrey J. Reuer Fisher College of Business Ohio State University 2100 Neil Avenue Columbus, OH 43210 USA

• Tel. (614) 292-3045

Fax: (614) 292-7962 e-mail: reuerj@cob.ohio-state.edu Jung-Chin Shen Strategy and Management Department INSEAD Boulevard de Constance 77305 Fontainebleau Cedex, France Tel.: (33) 1 60 72 44 73 Fax: (33) 1 60 74 55 00 E-mail: jung-chin.shen@insead.fr March 2002 JEL Codes: G30, D80, L10 In developing this research, we have benefited from comments and suggestions from Jean Helwege, Konstantina Kiousis, Mitchell Koza, and Subi Rangan. We also gratefully acknowledge financial support from INSEAD’s R&D Department.

SEQUENTIAL DIVESTITURE THROUGH INITIAL PUBLIC OFFERINGS Abstract An initial public offering (IPO) is often seen as a natural end state in a firm’s development, whereby ownership rights are reallocated to accomplish a financing objective. However, the decision to go public is also an important corporate strategy choice that can have implications for the transfer of control rights. The analysis situates IPOs within an extended M&A process and considers private firms’ decisions to undertake an IPO prior to divestiture rather than undergo an

• utright sale. We develop the argument that IPOs can ameliorate ex ante transaction costs in the

market for corporate control when search costs are nontrivial and when information asymmetries increase the risk of adverse selection. The empirical evidence suggests that sequential divestiture is more likely in industries with spatially-dispersed firms and for firms with significant intangible resources. Investments in strategic alliances attenuate the impact of intangibles on the propensity of firms to divest sequentially through IPOs. 2

INTRODUCTION Despite the dramatic rise in initial public offerings (IPOs) in recent years and the potential significance of the going public decision in a firm’s evolution, IPOs have not figured highly on the research agenda of strategy scholars. Rather, IPOs have tended to be seen as vehicles to finance growth or cash out following an entrepreneurial firm’s development. Based

• n traditional finance theory, IPOs can also be assessed in terms of the tradeoffs between the

benefits from relaxing owner-managers’ liquidity constraints and the agency costs that arise after the firm is publicly held (e.g., Jensen & Meckling, 1976). These conventional perspectives therefore focus on the firm’s access to capital and the transfer of ownership rights. In short, the IPO tends to be seen as a natural end state that addresses a purely financial objective. Recent descriptive findings and anecdotal evidence challenge these views, however. They indicate that IPOs are often part of a larger process of transferring control rights in

• rganizations (e.g., Mikkelson, Partch, & Shah, 1997). For instance, in newly-public Italian

firms, the controlling stake is sold to an outsider 13.6 percent of the time, which is roughly twice the rate for the Italian economy in general (Pagano, Panetta, & Zingales, 1998). Not only are control changes more likely to happen in newly-public firms than private firms, but newly-public firms experience a greater rate of acquisition than established public firms (Field & Mulherin, 1999). Practitioners also suggest that firms can use an IPO as a means of “teeing up” a company for subsequent sale (Rock, 1994). Taken together, these results indicate that transactions in capital markets and in the market for corporate control are not independent as is often assumed. An important question that follows is why private firms would incur the substantial costs

• f using IPOs as part of a sequential divestiture process rather than just selling the firm outright.

For instance, Ritter (1987) finds that registration and underwriting costs alone represent 14 3

percent of the proceeds raised. The indirect costs associated with underpricing, or the positive abnormal returns in early trading that represent wealth transfers from existing to new shareholders, are 15 percent on average (Ibbotson, 1975; Ibbotson, Sindelar, & Ritter, 1988; Ritter, 1984; Smith, 1986). Other costs relate to significant managerial time and organizational disruption surrounding the going-public event (Blowers, Griffith, & Milan, 1999). Several formal models have been recently developed to explain why IPOs might be an attractive component of firms’ divestiture efforts. Zingales (1995) suggests that a firm can use a two-stage sale to maximize total proceeds by relying on the capital market to auction off cash flow rights and the market for corporate control to negotiate the sale of the private benefits of

• control. In his model, the seller can obtain the value of the cash flow rights by selling them in a

competitive market to dispersed investors rather than through direct bargaining with a single

• bidder. Therefore, the IPO transaction allows the firm to avoid haggling over a portion of its
• value. Mello and Parsons (1998) emphasize the information an owner obtains by going public

regarding the value of the firm, and Ellingsen and Rydqvist (1997) develop a model focusing on the screening function of the stock market for firms undergoing divestiture. Despite this theoretical progress, Pagano, Pannetta, and Zingales (1998) conclude that “[t]he decision to go public is one of the most important and least studied questions in corporate finance” (p. 27). However, in research that has focused on the decision to go public, the decision tends to be considered in terms of “two major alternatives: to remain private or to become a public corporation” (Smith, 1986: 19). Although scholars have recently begun to frame the going-public decision based on the choice between private and public financing (e.g., Chemmanur & Fulghieri, 1999), the models discussed above and the emerging findings on the higher propensity of newly-public firms to be acquired underscores the value of assessing IPOs 4

in a comparative institutional light. Related research on mergers and acquisitions has done just that in many studies on firms’ choices between M&A, internal development, and alliances (e.g., Chatterjee, 1990; Hennart & Reddy, 1997; Mitchell & Singh, 1992; Shaver, 1998). However, this stream of research has also generally neglected sell-side considerations in acquisitions (e.g., Bergh, 1995; Duhaime & Grant, 1984; Harrigan, 1981; Klein, 1986; Ravenscraft & Scherer, 1991) and the potential role that IPOs in particular play within a more extended M&A process. In this paper, we empirically examine private firms’ decisions to use sequential divestitures through IPOs rather than outright sales. We suggest that a firm may use an initial public offering prior to divestiture in order to increase the firm’s visibility vis-à-vis potential acquirers, and this can be efficient when parties’ understanding of the identity or availability of potential exchange partners is incomplete. Sellers may also have to contend with buyers’ valuation difficulties stemming from asymmetric information. Specifically, when the seller is unable to credibly and efficiently demonstrate the value of the firm, it may not be able to obtain a selling price that reflects the business’ value. The process of going public not only increases the information available on a firm, but it can provide signals on the firm’s quality that can mitigate the effects of asymmetric information. The empirical evidence suggests that firms use sequential divestiture in the presence of search costs as well as when intangible resources increase the risk

• f adverse selection. Strategic alliances offer alternative signals to would-be acquirers and

attenuate the impact of intangibles on a firm’s proclivity to use IPOs prior to divestiture. THEORY AND HYPOTHESES Search for Exchange Partners In many markets, including the M&A market, potential buyers and sellers, prices, and

• ther conditions routinely change, stimulating search processes in order to locate potential

5

exchange partners and obtain favorable prices (Stigler, 1961). New firms enter the industry,

• ther firms exit or are acquired, firms’ strategies are altered, and the value of firms can change
• ver time for other reasons. Searching for potential exchange partners can be costly, with

expenses relating to the scanning of sellers and buyers, the encoding and decoding of signals (Arrow, 1974), and the employment of agents, among others. Nevertheless, many exchanges continue to take place through a market interface, due in part to the presence of institutions such as advertising and specialized traders to reduce these costs, and these institutions themselves emerge when profit opportunities justify their additional costs. The search costs noted above become manifest and tend to increase as the spatial dispersion of potential exchange partners grows. Such costs create “a powerful inducement to localize transactions as a device for identifying potential buyers and sellers” (Stigler, 1961: 216). Early organized markets, for instance, physically brought together buyers and sellers. By locating themselves close to one another, exchange partners reduce the costs of canvassing buyers and sellers and the costs of sending supply or demand signals. By contrast, as distance increases, trade tends to fall off dramatically, more so than can be explained by transportation costs and other factors (e.g., McCallum, 1995). As one example in the international context, the sluggish and small responses by some firms to opportunities created by exchange rate movements can be explained by their lack of knowledge about the existence and location of exchange partners, which tends to improve through the accumulation of external ties through foreign direct investment (Rangan, 2000a). A sequential divestiture strategy may therefore become attractive relative to an outright sale when firms in the industry are geographically dispersed and the IPO serves to raise the profile of the firm. Going public can enhance the visibility of the firm in a number of ways. 6

Prior to the IPO event, for instance, the firm engages in an intensive marketing effort lasting several months. During so-called road shows and subsequent registrations and offerings, firms not only present themselves to the investment community and attract media attention, they also tap indirectly into underwriters’ business relationships. After the firm has gone public, stock price, accounting, and other information are widely disclosed on the firm. By incurring the costs

• f increasing its visibility through an IPO, the seller can enter the feasible set of more potential

buyers, which increases the likelihood of an attractive bid. On the part of potential buyers, they are able to economize on their attention (March & Simon, 1958) and avoid redundant search efforts in pursuing acquisitions. We therefore posit: Hypothesis 1: The greater the spatial dispersion of firms in the industry, the greater the likelihood that a firm will divest sequentially using an IPO rather than undergo an outright sale. Asymmetric Information and Adverse Selection Ex ante transaction costs in the M&A market are tied not only to search costs but also to the costs related to valuing targets. Specifically, Akerlof’s (1970) classic market for lemons in product markets is instructive and may be in operation in the M&A market under certain

• circumstances. Suppose, for instance, that a seller has superior information as to its value, yet it

is unable to credibly convey its true worth to potential buyers. It follows that buyers’ offers will reflect the possibility that the seller will overstate its value and that the buyer may end up with an acquisition of lower quality. In the absence of signals to differentiate high and low quality firms

• r in the absence of institutions such as warranties or trust, in the extreme the market may fail

due to adverse selection or at least be subject to inefficiencies due to more extensive information gathering, negotiations, and so forth, which can contribute to a lower level of acquisitions in equilibrium. 7

The level of asymmetric information for an individual divestiture rests on the type of resources being divested on the sell side. Concerning undifferentiated physical assets, for example, sellers may be able to convey their value with relative ease, and ex ante transaction costs will therefore tend to be relatively low. By contrast, adverse selection may be more likely to occur in the sale of intangible assets. This is the case because financial data may provide little information on these resources’ true value. Further, the knowledge held by the seller on the resources’ quality is more difficult to convey and verify, which increases the likelihood of ex ante misrepresentations. Consistent with these arguments, Coff (1999) finds that in knowledge- intensive industries, M&A negotiations tend to be more lengthy and buyers respond by offering lower bids. The process of going public can be responsive to such adverse selection problems and can deal with the underlying information asymmetries in two different ways. First, to some extent IPOs can directly reduce asymmetric information by increasing the quantity of information available on a firm. For example, firms going public face new information disclosure requirements for registration and subsequent listing. The aggregated information and collective judgment of investors may help buyers calibrate their bids (e.g., Hellwig, 1980). In an empirical analysis of the Securities Act of 1933, Simon (1989) finds that increased financial disclosure requirements reduced investors’ forecast errors. Second, an IPO can reduce the effects of a given level of asymmetric information by signaling firm value (Spence, 1974). Higher quality firms will tend to be better able to bear the costs of underwriting fees, legal and accounting expenses, and auditing charges as well as the indirect costs of underpricing and managerial time commitments than lower quality firms. The fact that a firm was able to go public therefore helps buyers assess its quality (Ellingsen & 8

Rydqvist, 1997). Moreover, investment bankers and other institutions involved in the IPO process can certify the quality of participating firms because the repeated nature of their business encourages them to refrain from opportunism and build up their reputation capital (e.g., Beatty & Ritter, 1986). Thus, to the extent that IPOs can reduce asymmetric information directly or mitigate its effects by offering signaling opportunities, this can reduce buyers’ downside risks and can enhance the efficiency of the M&A market. Hypothesis 2: The greater a firm’s intangible assets, the greater the likelihood that it will divest sequentially using an IPO rather than undergo an outright sale. The arguments presented above suggest that IPOs can enhance the efficiency of the market for corporate control in the presence of asymmetric information. The question that naturally arises is whether other actions by the firm might either be cheaper or more effective than taking the firm public and subjecting it to the costs and disruption associated with an IPO. For instance, warranties, hostages, and insurance provide remedies to failures in many product markets otherwise subject to adverse selection, though such institutions are not developed for many one-time decisions made by organizations. In view of the limitations of such remedies, Rangan (2000b) discusses situations in which firms come to rely heavily on their social networks for search and deliberation in economic exchange. Strategic alliances are an alternative form of organizational investment discussed in the literature on external corporate development that are thought to redress adverse selection in the market for firm resources (Balakrishnan & Koza, 1993; Reuer & Koza, 2000). The logic is that firms can use alliances to avoid the ex ante transaction costs associated with an acquisition under conditions of adverse selection. The alliance permits firms to pool together resources on a piecemeal basis. Such arrangements avoid the transfer of ownership rights and a terminal sale, and therefore can be reversed at lower cost. While misrepresentations are still possible, repeated 9

contracting and the threat of termination can promote information revelation, and such relationships afford opportunities for first-hand experience with the resources in question. Beyond reducing information asymmetries directly for collaborators, alliances can also mitigate the effects of asymmetric information for other potential acquirers. Under the assumption that taking on alliance partners will be more difficult for lower quality firms, alliance investments may provide signals that help potential buyers differentiate attractive from unattractive targets. Based on the discussion above that adverse selection tends to become more problematic as intangibles increase, we posit that alliances will attenuate the effect of intangibles on the likelihood of sequential divestiture through an IPO. Hypothesis 3: The likelihood of sequential divestiture through an IPO rather than a direct sale will be negatively related to the interaction between a firm’s intangible assets and its investments in strategic alliances. METHODS Data

• Sample. In order to test the hypotheses developed above, a sample of private firms

selling to other firms or staging their divestitures was developed. We first used the M&A module of the Securities Data Corporate (SDC) database to obtain a sample of private firms in the manufacturing sector (i.e., SICs 2000-3999) that were acquired by firms during the time period 1996-1999. The sample was comprised of firms acquired by U.S. bidders, and LBO firms were excluded. The new issues module of the SDC database was then used in order to obtain a second sample of private firms that had gone public and were acquired within three years of the initial public offering. Specifically, a sample of public targets was developed, and the new issues module of the SDC database was used to determine if the publicly-held targets had become public in a particular time frame period prior to the acquisition. Thus, by using the M&A and 10

new issue modules of the SDC database, we were able to develop a sample of private firms that were acquired outright versus acquired shortly after going public. As before, for the subsample

• f public targets identified as newly public in the new issues module of the SDC database, our

focus was on targets in the manufacturing sector (i.e., SICs 2000-3999). Additionally, all acquirers were based in the US, LBO firms were excluded, and the acquisitions of such firms also took place during the years 1996-1999. We considered only initial public offerings and excluded related transactions such as equity carveouts and spinoffs of business units. Other studies tracking newly-public firms and their propensity to be acquired have used time windows as long as ten years in length (e.g., Mikkelson, Partch, & Shah, 1997; Field & Mulherin, 1999), but for our purposes the assumption of an intent to divest sequentially becomes more problematic as the time window lengthens. Moreover, the selection of the three-year time frame is consistent with the time window employed in prior research on acquisitions following initial public offerings (Pagano, Panetta, & Zingales, 1998). For our particular sample, the average time span between the issue date of the offering and the announcement date for the acquisition was roughly 1.3 years. The sectoral distribution of the outright divestitures of private firms and sequential divestitures through IPOs is given in Table 1. The propensity of private firms to use an IPO prior to their sale averages 4.8 percent and ranges from zero to twenty-five percent for the industries considered. For the industries with more than one hundred transactions, the propensity to use initial public offerings prior to divestitures is above average in the following industries: chemicals and allied products (i.e., 9.4 percent); measuring, analyzing, and controlling instruments (i.e., 6.1 percent); and electronic and electrical equipment (i.e., 5.4 percent). 11

=================== Insert Table 1 about here =================== We also compared the distribution of the observations over the 1996-1999 timeframe with the divestitures reported in Mulherin and Boone (2000). The subsample of firms that went public prior to divestiture had the same distribution as in their study (χ2=2.51) while the subsample of firms that divested outright had a different distribution (p<0.001) inasmuch as fewer sales occurred in 1996 and proportionately more in the remaining years. The similarities for firms that went public and the differences for private firms that divested outright can be explained by the fact that Mulherin and Boone (2000) focus on divestitures by publicly-traded firms appearing in the Value Line Investment Survey. In the subsection that follows, the measures used in the multivariate analysis are

• described. After accounting for missing data for these variables and outliers, 820 observations

were available for analysis, 120 of which were private firms that were acquired after undergoing an IPO. Given this reduction in sample size due to the paucity of data for private firms, we reran the multivariate models without the firm size control. This increased the sample size from 820 to 3463, and the same results were obtained for the theoretical variables. Additional details on the sample are contained in the results section. Model Specification and Measures The basic structure of the multivariate model used to differentiate firms that went public just prior to divestiture from those that sold outright is as follows: (1) Initial Public Offering = β0 + β1Spatial Dispersion + β2R&D Intensity + β3Strategic Alliances + β4R&D Intensity · Strategic Alliances + β5Firm Size + β6Industry Incumbents + β7Industry Uncertainty + ε. 12

Given the binary nature of the dependent variable, equation (1) was estimated using logistic regression. The first explanatory variable captures search costs in the industry’s acquisition market, as reflected in the geographic dispersion of firms in the industry (i.e., Spatial Dispersion). The next two explanatory variables – R&D intensity and Strategic Alliances – and their interaction incorporate adverse selection problems from information asymmetries created by intangible resources as well as the role of alliances in signaling the value of intangibles. While our objective was to develop a parsimonious model to differentiate firms using sequential divestiture through IPOs from those relying on direct sales, we introduced controls for the size of the firm (i.e., Firm Size) as well as the number of firms in the industry (i.e., Industry Incumbents) and the uncertainty of the industry in which the firm is embedded (i.e., Industry Uncertainty). The measurement for the regressors and motives for the firm- and industry-level controls are taken up below. Initial public offering. As described earlier, we determined whether each private firm was sold outright or the divestiture was preceded by an initial public offering. Therefore, Initial Public Offering equals one if the firm went public up to three years prior to being acquired by another firm, and zero if the firm was divested outright to another firm. Data for this variable were obtained from the new issues and M&A modules of the SDC database. Explanatory variables. The first explanatory variable captures the geographic dispersion of firms within an industry. Let Pij be the proportion of firms in firm i’s 2-digit SIC industry in state j (i.e., j∈J) in 1996. Since the sum of the squared proportions yields a measure

• f the geographic concentration of firms, and this measure has a lower bound of zero and an

upper bound of one, the spatial dispersion of firm i’s industry was defined as follows: 13

(2) Spatial Dispersion = 1 . P

J j 2 ij

∑ −

∈

Thus, Spatial Dispersion ranges from zero to one and increases in the geographic dispersion of firms in the focal firm’s industry. Data for this variable were obtained from the U.S. Census Bureau and the Office of Advocacy of the Small Business Administration. We used the intangible assets of the focal firm’s industry to develop a proxy for the asymmetric information faced by potential acquirers. This approach was motivated by the fact that very little firm-level data are available on private firms that comprise the sample. Specifically, the R&D intensity of the focal firm’s industry provides a measure of selling parties’ intangible assets and the ex ante valuation challenges would-be acquirers confront. We measured R&D Intensity as the ratio of R&D expenditures to sales for the year prior to the

• divestiture. Data for this variable were obtained from the National Science Foundation’s

Research and Development in Industry as published in the U.S. Census Bureau’s Statistical Abstract of the United States. Data are available at the 2-digit SIC level, with the exception of the transportation equipment industry (i.e., SIC 37) since data for motor vehicles and motor vehicle equipment (i.e., SIC 371) are separated from data for aircraft and missiles (i.e., SICs 372, 376) at the 3-digit SIC level. To measure the firm’s investment in strategic alliances in order to test its interaction with R&D intensity, we used a separate module of the SDC database containing data on alliance formation to count the number of alliances the firm entered (i.e., Strategic Alliances). The SDC database includes collaborative agreements such as equity joint ventures; minority purchases; R&D contracts; joint manufacturing, marketing, or supply agreements; and licensing and value- added resale agreements as strategic alliances. Data on these alliances were assembled from 1986, the year the database began tracking alliances. 14

Control variables. In addition to the theoretical variables discussed above, several additional regressors were incorporated into the specification to control for other firm- and industry-level effects that might influence the route the firm follows in pursuing a divestiture and also relate to the theoretical variables of interest. First, we introduced a control for the size of the

• firm. The uncertainty surrounding the products and technologies of smaller, entrepreneurial

firms might pose valuation difficulties for potential acquirers. Moreover, if firms go public to understand their value, smaller firms rather than large organizations should use IPOs (Mello & Parsons, 1998). However, prior research has also found that economies of scale exist for the direct costs

• f going public (Jenkinson & Ljungqvist, 1996), that the indirect costs due to underpricing are

greater for smaller issues as well (Beatty & Ritter, 1986), and that the market for investment banking services is segmented (e.g., Carter & Manaster, 1990) such that the most reputable underwriters tend to issue shares of large and prestigious clients, which involve less risk. Firm size was measured as the firm’s number of employees (in thousands). Data for this control were

• btained from SDC, Compustat, and Standard and Poors.

At the firm level, we also incorporated a control for the direct effects of alliances since we are interested in testing the interaction between alliance investments and intangible resources. Inclusion of the control for alliances is also based on the observation that firms enter into alliances to achieve high growth (Dutta & Weiss, 1997) and on findings showing that companies tend to go public following a period of expansion (Pagano, Panetta, & Zingales, 1998). Prior research has also reported that firms with alliances tend to go public sooner and obtain higher IPO valuations (Stuart, Hoang, & Hybels, 1999). 15

At the industry level, we first introduced a control for the number of firms in the focal firm’s industry (i.e., Industry Incumbents). Potential buyers experienced in the same line of business as the seller tend to be in a better position than others to make such judgments concerning the value of the acquisition. Thus, when the number of firms in an industry increases, the set of potential acquirers for which information asymmetries are low also tends to increase, as will the spatial dispersion of firms in the industry. However, profit opportunities due to concentration and resulting entry barriers can stimulate acquisitive entries despite possible information asymmetries (e.g., Yip, 1982). Data for this variable were obtained from the U.S. Census Bureau and the Office of Advocacy of the Small Business Administration. Finally, we included a control for industry uncertainty to address the unpredictability of demand conditions as well as product-market factors potentially shaping firms’ investments in internal development projects and external alliances (e.g., Bergh & Lawless, 1998; Eisenhardt & Schoonhoven, 1996). Such uncertainty may lead to greater IPO costs in the form of underpricing (Beatty & Ritter, 1986) as well as valuation problems for potential acquirers. Industry uncertainty was calculated as an ex post measure of the volatility of net sales in each industry using regression analysis over the five-year time period 1992-1996 (Dess & Beard, 1984; Keats & Hitt, 1988). The specification used to develop the proxy for industry uncertainty was as follows: (3) Industry Sales = γ0 + γ1Year + ε. Industry uncertainty was then measured as the standard error of the slope parameter divided by the mean of industry sales. Data required for estimating the industry-specific regressions and calculating the proxy for industry uncertainty were obtained from Compustat. 16

RESULTS Table 2 presents descriptive statistics and a correlation matrix. For the final sample, twenty percent of the firms engaged in an IPO prior to divestiture. This high percentage relative to the base sample described in the methods section can be ascribed to the paucity of firm-level employment data for private firms that divested outright. However, the sectoral distribution of

• utright divestitures indicated that the final sample was representative of the base sample (i.e.,

χ2=6.89, n.s.). As noted above, we also reran the models without the firm size control and

• btained the same results for a much larger sample of firms that divested outright. The average

R&D intensity for the industries considered is 3.86 percent and ranges from 0.40 to 7.70 percent. The descriptive statistics indicate that firms’ alliance experiences were modest. The average number of alliances formed by the sampled firms was 0.08, yet one firm had as many as seven

• alliances. Analysis of variance indicated that alliance usage varied across industries (F=2.39,

p<0.01), yet this inter-industry heterogeneity reflects in part firm’s decisions to divest

• sequentially. No significant differences in alliance usage across industries were observed for the

two subsamples of firms that sold outright (F=1.41) or that went public first (F=1.18). The average firm had 502 employees, and inter-industry differences in firm size were not evident for the firms that went public prior to divestiture (F=0.97), but such firms tended to be larger in size (t=2.68, p<0.01) and more varied in size (F=123.9, p<0.0001) than firms that divested outright. =================== Insert Table 2 about here =================== The correlation matrix reveals modest correlations among most of the variables. As might be expected, not only are most industries spatially dispersed, but the spatial dispersion of firms in an industry tends to increase as the number of firms in the industry rises (p<0.001). The 17

R&D intensity of spatially-dispersed industries tends to be low, while industries with firms that are geographically concentrated are more R&D intensive (p<0.001), which is consistent with research on localization and knowledge spillovers (Harrison, Kelley, & Gant, 1996; Jaffe, Trajtenberg, & Henderson, 1993). Reflecting the use of alliances to exploit existing capabilities or to tap into others’ skills in the presence of contractual hazards (Mowery, Oxley, & Silverman, 1998; Teece, 1992), firms situated in R&D intensive industries tend to be more engaged in collaborative agreements with

• thers (p<0.01). Similar to Dutta and Weiss (1997), we note that no relationship is evident

between industry uncertainty and firms’ investments in alliances, despite arguments that alliances provide firms with needed flexibility in such settings (e.g., Kogut, 1991). Although the correlations among the regressors are generally low, inclusion of the interaction between R&D intensity and strategic alliances in the multivariate model resulted in a maximum variance inflation factor (VIF) of 10.6, above the rule-of-thumb threshold value of ten used to indicate multicollinearity problems in regression models (Neter, Wasserman, & Kutner, 1985). To alleviate concerns of multicollinearity, the variables R&D Intensity and Strategic Alliances were standardized prior to forming the interaction term, which reduced the maximum VIF to 2.3. Estimation results for models using these transformed variables appear in Table 3. Model I presents a baseline specification consisting of three controls. Model II augments this model with the direct effects of the theoretical variables, and Model III represents the full model incorporating the interaction effect between R&D intensity and alliances. All three models are significant at the 0.001 level. Log-likelihood ratio tests demonstrate that Models II and III have significantly more explanatory power than Model I (both p<0.001), and Model III has 18

significantly more explanatory power than Model II (p<0.01). Model III correctly predicts whether or not the firm used an IPO for 683 of the 820 transactions (or 83.3 percent). This hit rate improves upon alternatives such as random assignment (i.e., 50 percent) or the value that would obtain if all transactions were assigned to the dominant category of outright divestiture (i.e., 80 percent). =================== Insert Table 3 about here =================== The first hypothesis concerned the potential use of IPOs to raise the visibility of firms in industries with firms that are geographically dispersed. Consistent with this prediction, in industries characterized by greater spatial dispersion, firms have a greater tendency to divest sequentially via IPOs (p<0.05 in Model II and p<0.01 in Model III). Outright divestitures are more likely in industries with firms that are geographically more concentrated, holding everything else constant. In order to be able to interpret the effect further, we re-specified the model using standardized data, which yielded a point estimate of the partial odds ratio for the spatial dispersion measure of 1.52, which indicates an increased probability of sequential divestiture of 0.6 for a one standard deviation increase in spatial dispersion, holding constant the

• ther covariates.

Consistent with the second hypothesis, the results suggest that the tendency of firms to use IPOs prior to divestiture is greater in industries that are R&D intensive (p<0.001). For firms with fewer intangibles, outright divestiture is more efficient by avoiding the costs associated with going public. In order to consider the relative importance of the motives of using an IPO to raise the firm’s visibility and to reduce valuation problems, we re-estimated the direct effects model using standardized data. This resulted in a standardized beta of 0.31 for the spatial dispersion 19

variable and 0.62 for the R&D intensity measure, indicating that a one standard deviation increase in the latter has roughly twice the impact of a similar increase in the former. The third hypothesis posited that strategic alliances can lessen the effects of intangible resources since alliances can signal quality to would-be acquirers. The estimation results appearing in Table 3 suggest that the effects of R&D intensity are attenuated by private firms’ investments in strategic alliances. The interaction between R&D intensity and strategic alliances is negative and significant at the 0.01 level. Thus, in the absence of alliances, increases in intangibles are most problematic and tend to stimulate sequential divestiture. In order to explore the sensitivity of the findings to the measurement of intangible resources, we constructed two additional proxies for intangibles. First, we implemented a measure of Tobin’s Q to provide a proxy for overall intangibles. Following Chung and Pruitt (1994), we approximated Tobin’s Q as the market-to-book ratio since this measure explains over 96 percent of the variance in a more sophisticated Tobin’s Q ratio that would require arbitrary assumptions about depreciation and inflation rates for the calculation of assets’ replacement

• values. The market value numerator is the year-end market value of common stock plus the

book value of preferred stock and debt. The book value denominator is year-end total assets, and the measure is calculated at the 3-digit SIC level using Compustat data. As before, to mitigate multicollinearity concerns, the variable was standardized prior to forming the multiplicative

• term. Second, we followed Kohers and Ang’s (2000) analysis of the relationship between

intangibles and the structuring of M&A deals by relying on the high-tech indicators in the SDC database to distinguish high-tech targets from firms in other industries. Estimation results appear in Table 4. 20

=================== Insert Table 4 about here =================== Models I and III provide direct effects models, and Models II and IV introduce the interaction effects using the additional proxies for intangibles. As before all four models are highly significant and explain significantly more variance than the control variables alone (p<0.001). Firms situated in industries with spatially dispersed firms tend to exhibit a preference for sequential divestiture over outright sale (p<0.10 in Models I and II; p<0.001 in Models III and IV). The direct effect of intangibles, whether measured in broader terms using Tobin’s Q or more narrowly based on a high-tech industry indicator, is positive (p<0.001 in all models), and alliances moderate the effects of Tobin’s Q (p<0.05 in Model II) and the high-tech indicator (p<0.10 in Model IV). These results indicate that the findings are robust to the measurement of intangible resources. The results for the control variables also deserve some comment. First, larger firms rather than smaller firms tend to rely on sequential divestitures through IPOs rather than outright sales (p<0.001 in all models). This may reflect factors such as economies of scale in going public (Jenkinson & Ljungqvist, 1996) as well as segmentation in the market for investment banking services (Carter & Manaster, 1990). Second, the direct effect of investments in strategic alliances is positive (p<0.001 in all models), which might be interpreted in several ways. Prior research has established that firms with such cooperative arrangements tend to go public sooner and raise greater proceeds (Stuart, Hoang, & Hybels, 1999). The positive effect of this variable might reflect a correlation with an omitted firm-level variable such as growth (Dutta & Weiss, 1997), but then the question arises as to why such firms are not equally attractive acquisition candidates absent other considerations such as greater valuation complexity. Other research has 21

suggested that firms defensively enter into alliances to mimic competitors’ moves (e.g., Caves, 1996), and such firms may also be mimicking the going public decisions by partners or others during this active period in the IPO market. Finally, at the industry level there is some evidence that more predictable industries tend to experience outright sales (p<0.10 in Model III in Table 3; p<0.05 in Models I and II in Table 4), as do industries with more incumbents (p<0.05 in Models I and II in both tables; p<0.01 in Model III in Table 3; p<0.001 in Models III and IV in Table 4). DISCUSSION The study has several implications for research and points to areas where additional work

• n IPOs might be conducted. At the most general level, there is value in considering IPOs in a

comparative institutional light given the importance of better understanding the decision by firms to go public rather than only attending to the benefits and costs of going public (Pagano, Panetta, & Zingales, 1998). By emphasizing some of the inefficiencies that arise in M&A markets in specific settings, we suggest that IPOs might play a part in reducing ex ante transaction costs. Thus, the costs borne by firms in equity markets can have ramifications for the efficiency for the market for corporate control. The theoretical arguments and evidence also highlight the relevance of considering IPOs within a more extended process of strategic decision-making in general, and M&A activity in particular, rather than as natural end states in the life cycles of private firms that address purely financial objectives. The relative neglect of the going-public decision by strategy scholars and

• ther researchers interested in the theory of the firm likely reflects the latter, more traditional
• view. However, emerging evidence on the propensity of newly-public firms to be acquired and
• ur findings on the tradeoffs between sequential divestiture through IPOs and outright divestiture

lead us to conclude that there is more scope for research on IPOs using other, more 22

interdisciplinary perspectives. The findings suggest that IPOs are often means rather than ends, strategic choices rather than only financial decisions. It is important to note that the analysis presented here relies on a discrete choice model and the conventional reduced form set-up that it implies. Thus, the paper is silent on the actual performance implications of sequential divestiture through IPOs versus direct sales. Extensions might go on to establish the performance implications of IPOs within different strategies pursued by firms. Such work could also examine related transactions designed for corporate restructuring, including equity carveouts and spinoffs. In the case of carveouts, an additional player – the parent firm – is involved in an ongoing and significant way. Corporate parents typically do not tend to grant autonomy to carved-out units (Slovin et al., 1995), and the modal

• utcome for carveouts is reacquisition by the parent corporation (Schipper & Smith, 1986).

Spinning off units independently may therefore resolve valuation problems and adverse selection in M&A markets in a manner that is more similar to IPOs. Moreover, by altering operational relations with other units in the corporation, such transactions might also reduce transaction costs

• therwise associated with post-merger integration.

Relative to recent theoretical developments in the IPO literature in particular, the findings

• ffer evidence that can inform the debate between new theories on IPOs and subsequent
• divestiture. For instance, the result that large firms rather than small ones tend to use IPOs prior

to divestiture appears to be more consistent with the view that owner-managers use IPOs to maximize their total proceeds from the separate sale of cash flow rights and the private benefits

• f control (Zingales, 1995) than the perspective that firms go public to develop an understanding

about their value (Mello and Parsons, 1998). However, it is also worth pointing out that this result might simply reflect scale economies in the direct costs of going public (Jenkinson & 23

Ljungqvist, 1996) or segmentation in the market for investment banking services (e.g., Carter & Manaster, 1990). Thus, future research might further distinguish the empirical implications of these different theories and examine their relative importance. The particular argument that we develop and test is that IPOs may mitigate inefficiencies in the acquisitions market in two ways. First, IPOs can reduce ex ante transaction costs attending acquisitions by raising the visibility of firms and thereby reducing search costs. We find that private firms embedded in industries that are spatially dispersed tend to use sequential divestitures through IPOs rather than outright sales. Second, IPOs may be responsive to adverse selection in the M&A market by reducing asymmetric information directly or by sending signals concerning the quality of the firm. Opportunities exist to examine other potential factors leading firms to choose IPOs over alternative forms of financing and organizational governance. For instance, while our focus has been on ex ante transaction costs in the market for corporate control, future research could also examine transaction costs arising from moral hazards and tradeoffs concerning incentive intensity and coordinated adaptation (Williamson, 1991). Based on the results presented here, it might be tempting to conclude that asocial remedies to inefficiencies in the market for corporate control such as IPOs are necessary or

• ptimal. However, we note that the acquisition of a private firm will often satisfy Rangan’s

(2000b) criteria that search and deliberation are problematic in order for social networks to influence the efficiency of economic exchange. Moreover, social relations figure into IPO processes, and while alliances may be seen in part as asocial mechanisms in providing signals to

• ther potential acquirers, they also provide indirect social linkages between firms in an industry

(e.g., Gulati, 1995). Therefore, because prior studies have tended to consider single remedies to 24

search and valuation problems, it would be attractive to assess simultaneously the various asocial and social alternatives that firms have at their disposal. Extensions might also explore the IPO process in greater detail by examining specific features of IPOs. For example, while we have suggested that IPOs can enhance a firm’s visibility and reduce valuation problems in general, specific signals can vary from deal to deal based on factors such as an underwriter’s reputation (Carter & Manaster, 1990) and whether the

• ffering proceeded on a firm commitment or best efforts basis. Models exist that link signals

from underpricing to higher prices in the aftermarket for remaining stakes (Allen & Faulhaber, 1989) and to more favorable seasoned equity offerings (Welch, 1989), so future work could examine how underpricing potentially affects subsequent acquisition. In broader terms, future research could track newly public firms over time in order to investigate different features of IPOs and their implications, the interplay between equity markets and the M&A market, and the information revealed during the process of going public. 25

REFERENCES Akerlof, G. A. 1970. The market for 'lemons': Qualitative uncertainty and the market

• mechanism. Quarterly Journal of Economics, 84: 488-500.

Allen, F., & Faulhaber, G. R. 1989. Signaling by underpricing in the IPO market. Journal of Financial Economics, 23: 303-323. Arrow, K. J. 1974. The limits of organization. New York; W. W. Norton. Balakrishnan, S., & Koza, M. P. 1993. Information asymmetry, adverse selection, and joint

• ventures. Journal of Economic Behavior and Organization, 20: 99-117.

Beatty, R., & Ritter, J. R. 1986. Investment banking, reputation, and the underpricing of initial public offerings. Journal of Financial Economics, 15: 213-232. Bergh, D. D. 1995. Size and relatedness of units sold: An agency theory and resource-based

• perspective. Strategic Management Journal, 16: 221-239.

Bergh, D. D., & Lawless, M. W. 1998. Portfolio restructuring and limits to hierarchical governance: The effects of environmental uncertainty and diversification strategy. Organization Science, 9: 87-102. Blowers, S. C., Griffith, P. H., & Milan, T. L. 1999. The Ernst and Young LLP guide to the IPO value journey. New York: John Wiley. Carter, R., & Manaster, S. 1990. Initial public offerings and underwriter reputation. Journal of Finance, 45: 1045-1067. Caves, R. E. 1996. Multinational enterprise and economic analysis (Second Edition). New York, NY: Cambridge University Press. Chatterjee, S. 1990. Excess resources, utilization costs, and mode of entry. Academy of Management Journal, 33: 780-800. Chemmanur, T., & Fulghieri, P. 1999. A theory of the going public decision. Review of Financial Studies, 12: 249-279. Chung, K. H., & Pruitt, S. 1994. A simple approximation of Tobin's q. Financial Management, 23: 70-74. Coff, R. W. 1999. How buyers cope with uncertainty when acquiring firms in knowledge- intensive industries: Caveat Emptor. Organization Science, 10: 144-161. Dess, G. G., & Beard, D. W. 1984. Dimensions of organizational task environments. Administrative Science Quarterly, 29: 52-73. 26

Duhaime, I. M., & Grant, J. H. 1984. Factors influencing divestment decision-making: Evidence from a field study. Strategic Management Journal, 5: 301-381. Dutta, S., & Weiss, A. M. 1997. The relationship between a firm's level of technological innovativeness and its pattern of partnership agreements. Management Science, 43: 343- 356. Eisenhardt, K. M., & Schoonhoven, C. B. 1996. Resource-based view of strategic alliance formation: Strategic and social effects in entrepreneurial firms. Organization Science, 7: 136-150. Ellingsen, T. & Rydqvist, K. 1997. The stock market as a screening device and the decision to go

• public. Unpublished manuscript, Stockholm School of Economics.

Field, L. C., & Mulherin, J. H. 1999. Newly-public firms as acquisition targets: A comparison with established firms. Unpublished manuscript, Penn State University. Gulati, R., 1995. Social structure and alliance formation patterns: A longitudinal analysis. Administrative Science Quarterly, 40: 619-652. Harrigan, K. R. 1981. Deterrents to divestiture. Academy of Management Journal, 24: 306-323. Harrison, B., Kelley, M., & Gant, J. 1996. Innovative firm behavior and local milieu: Exploring the intersection of agglomeration, firm effects, industrial organization, and technological

• change. Economic Geography, 72: 233-258.

Hellwig, M. F. 1980. On the aggregation of information in competitive markets. Journal of Economic Theory, 22: 477-498. Hennart, J.-F., & Reddy, S. 1997. The choice between mergers/acquisitions and joint ventures: The case of Japanese investors in the United States. Strategic Management Journal, 18: 1- 12. Ibbotson, R. G. 1975. Price performance of common stock new issues. Journal of Financial Economics, 22: 235-272. Ibbotson, R. G., Sindelar, J. L., & Ritter, J. R. 1988. Initial public offerings. Journal of Applied Corporate Finance, 1: 37-45. Jaffe, A. B., Trajtenberg, M., & Henderson, R. 1993. Geographic localization of knowledge spillovers as evidenced by patent citations. Quarterly Journal of Economics, 108: 577- 598. Jenkinson, T., & Ljungqvist, A. 1996. Going public: The theory and evidence on how companies raise equity finance. Oxford: Clarendon Press. 27

Jensen, M. C., & Meckling, W. H. 1976. Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3: 305-360. Keats, B. W., & Hitt, M. A. 1988. A causal model of linkages among environmental dimensions, macro organizational characteristics, and performance. Academy of Management Journal, 31: 570-598. Klein, A. 1986. The timing and substance of divestiture announcements: Individual, simultaneous, and cumulative effects. Journal of Finance, 41: 685-696. Kogut, B. 1991. Joint ventures and the option to expand and acquire. Management Science, 37: 19-33. Kohers, N., & Ang, J. 2000. Earnouts in mergers: Agreeing to disagree and agreeing to stay. Journal of Business, 73: 445-476. March, J. G., & Simon, H. A. 1958. Organizations. Oxford: Blackwell. McCallum, J. 1995. National borders matter: Canada-U.S. regional trade patterns. American Economic Review, 85: 615-623. Mello, A. S., & Parsons, J. E. 1998. Going public and the ownership structure of the firm. Journal of Financial Economics, 49: 79-109. Mikkelson, W. H., Partch, M. M., & Shah, K. 1997. Ownership and operating performance of companies that go public. Journal of Financial Economics, 44: 281-307. Mitchell, W., & Singh, K. 1992. Incumbents' use of pre-entry alliances before expansion into new technical subfields of an industry. Journal of Economic Behavior and Organization, 18: 347-372. Mowery, D. C., Oxley, D. E., & Silverman, B. S. 1998. Technological overlap and interfirm cooperation: Implications for the resource-based view of the firm. Research Policy, 27: 507-523. Neter, J., Wasserman, W., & Kutner, M. H. 1985. Applied linear statistical models (Second Edition). Homewood, IL: Irwin. Pagano, M., Panetta, F., & Zingales, L. 1998. Why do companies go public? An empirical

• analysis. Journal of Finance, 53: 27-64.

Ravenscraft, D. J., & Scherer, F. M. 1991. Divisional sell-off: A hazard function analysis. Managerial and Decision Economics, 12: 429-438. Rangan, S. 2000a. Search and deliberation in international exchange: Microfoundations to some macro patterns. Journal of International Business Studies, 31: 205-222. 28

Rangan, S. 2000b. The problem of search and deliberation in economic action: When social networks really matter. Academy of Management Review, 25: 813-838. Reuer, J. J., & Koza, M. P. 2000. Asymmetric information and joint venture performance: Theory and evidence for domestic and international joint ventures. Strategic Management Journal, 21: 81-88. Ritter, J. R. 1984. The hot issue market of 1980. Journal of Business, 57: 215-240. Ritter, J. R. 1987. The costs of going public. Journal of Financial Economics, 19: 269-282. Rock, R. H. 1994. Economic drivers of M&A. In Rock, M. L., Rock, R. H., & Sikora, M. (Eds.) The mergers and acquisitions handbook (Second Edition). New York: McGraw-Hill. Schipper, K., & Smith, A. 1986. A comparison of equity carve-outs and seasoned equity

• fferings: Share price effects and corporate restructuring. Journal of Financial

Economics, 15: 153-186. Shaver, J. M. 1998. Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect FDI survival? Management Science, 44: 571-585. Simon, C. J. 1989. The effect of the 1933 Securities Act on investor information and the performance of new issues. American Economic Review, 79: 295-318. Slovin, M. B., Sushka, M. E., & Ferraro, S. R. 1995. A comparison of the information conveyed by equity carve-outs, spin-offs, and asset sell-offs. Journal of Financial Economics, 37: 89-104. Smith, C. W. 1986. Raising capital: Theory and evidence. Midland Corporate Finance Journal, 4: 6-22. Spence, A. M. 1974. Market signaling: Informational transfer in hiring and related screening

• processes. Cambridge, MA: Harvard University Press.

Stigler, G. J. 1961. The economics of information. Journal of Political Economy, 69: 213-225. Stuart, T. E., Hoang, H., & Hybels, R. C. 1999. Interorganizational endorsements and the performance of entrepreneurial ventures. Administrative Science Quarterly, 44: 315-349. Teece, D. J. 1992. Competition, cooperation, and innovation: Organizational arrangements for regimes of rapid technological progress. Journal of Economic Behavior and Organization, 18: 1-25. Welch, I. 1989. Seasoned offerings, imitation costs, and the underpricing of initial public

• fferings. Journal of Finance, 44: 421-449.

Williamson, O. E. 1991. Comparative economic organization: The analysis of discrete structural

• alternatives. Administrative Science Quarterly, 36: 269-296.

29

Yip, G. S. 1982. Diversification entry: Internal development versus acquisition. Strategic Management Journal, 3: 331-345. Zingales, L. 1995. Insider ownership and the decision to go public. Review of Economic Studies, 62: 425-448. 30

TABLE 1 Sectoral Distribution of Divestitures Preceded by Initial Public Offeringsa

I Industry (SIC) II IPOs (%) III Divestiture (%) IV Percent IPO V Total (%) Food and Kindred Products (20) 11 (6.1) 269 (7.5) 3.9 280 (7.4) Tobacco Products (21) 2 (1.1) 6 (.2) 25.0 8 (.2) Textile Mill Products (22) 1 (.6) 44 (1.2) 2.2 45 (1.2) Apparel (23) 3 (1.7) 75 (2.1) 3.8 78 (2.1) Lumber and Wood Products (24) 2 (1.1) 77 (2.1) 2.5 79 (2.1) Furniture and Fixtures (25) 6 (3.3) 70 (1.9) 7.9 76 (2.0) Paper and Allied Products (26) 6 (3.3) 77 (2.1) 7.2 83 (2.2) Printing and Publishing (27) 0 (0) 0 (0) 0 (0) Chemicals and Allied Products (28) 34 (18.9) 329 (9.1) 9.4 363 (9.6) Petroleum Refining (29) 0 (0) 0 (0) 0 (0) Rubber and Plastics (30) 2 (1.1) 173 (4.8) 1.1 175 (4.6) Leather (31) 0 (0) 12 (.3) 12 (.3) Stone, Clay, Glass, and Concrete (32) 1 (.6) 86 (2.4) 1.1 87 (2.3) Primary Metals (33) 5 (2.8) 153 (4.2) 3.2 158 (4.2) Fabricated Metal Products (34) 2 (1.1) 263 (7.3) .8 265 (7.0) Industrial and Commercial Machinery and Computer Equipment (35) 25 (13.9) 559 (15.5) 4.3 584 (15.4 Electronic and Electrical Equipment (36) 33 (18.3) 577 (16.0) 5.4 610 (16.1) Transportation Equipment (37) 9 (5) 202 (5.6) 4.3 211 (5.6) Measuring, Analyzing, and Controlling Instruments (38) 30 (16.7) 460 (12.7) 6.1 490 (13.0) Miscellaneous Manufacturing (39) 8 (4.4) 171 (4.7) 4.4 179 (4.7) Total 180 (100) 3608 (100) 4.8 3783 (100)

aColumn II represents private firms that went public and were acquired within three years, and column III represents private firms that

were sold outright rather than undergoing a staged divestiture via an IPO. Percentages may not sum to 100 due to rounding. 31

TABLE 2 Descriptive Statistics and Correlation Matrixb Variable Mean S.D. (1) (2) (3) (4) (5) (6) (1) Initial public offering 0.20 0.40

• (2) Spatial dispersion

0.95 0.01

• 0.03
• (3) R&D intensity

3.86 2.58 0.16***

• 0.25***
• (4) Strategic alliances

0.08 0.45 0.25***

• 0.06†

0.10**

• (5) Firm size

0.50 2.34 0.20***

• 0.00
• 0.05

0.01

• (6) Industry incumbents

22.04 16.02

• 0.09*

0.14*** 0.04

• 0.01

0.05

• (7) Industry uncertainty

0.03 0.02 0.06

• 0.02
• 0.04
• 0.02

0.04 0.06

bN=820. † p<0.10, * p<0.05, ** p<0.01, *** p<0.001.

32

TABLE 3 Determinants of Staged Divestiture via IPOsc Variable Model I Model II Model III Intercept

• 1.48***

(0.20)

• 23.20*

(9.49)

• 31.19**

(10.81) Firm size 0.56*** (0.12) 0.71*** (0.14) 0.67*** (0.14) Industry incumbents

• 0.01*

(0.01)

• 0.02*

(0.01)

• 0.02**

(0.01) Industry uncertainty 3.81 (4.52) 6.74 (4.83) 7.98† (4.85) Strategic alliances

• 0.66***

(0.14) 0.96*** (0.22) Spatial dispersion

• 22.72*

(10.00) 31.21** (11.39) R&D intensity

• 0.62***

(0.12) 0.62*** (0.12) R&D intensity · Strategic alliances

• 0.48**

(0.18) χ2 41.64*** 110.21*** 125.49*** Log Likelihood, L(βk)

• 377.84
• 341.33
• 337.16
• 2[L(βI)-L(βk)]
• 73.02***

81.36***

• 2[L(βII)-L(βIII)]
• 8.34**

bN=820. Positive coefficients indicate that increases in the variable tend to increase the

likelihood the firm will use an IPO prior to divestiture. † p<0.10, * p<0.05, ** p<0.01,

*** p<0.001.

33

34 TABLE 4 Determinants of Staged Divestiture via IPOsd Variable Model I Model II Model III Model IV Intercept

• 15.72†

(0.20)

• 17.16*

(8.36)

• 54.21***

(11.25)

• 60.85***

(12.23) Firm size 0.62*** (0.14) 0.63*** (0.13) 0.99*** (0.17) 0.95*** (0.17) Industry incumbents

• 0.01*

(0.01)

• 0.01*

(0.01)

• 0.03***

(0.01)

• 0.03***

(0.01) Industry uncertainty 11.34* (4.84) 11.25* (4.88) 2.10 (5.34) 3.18 (5.30) Strategic alliances 0.98*** (0.21) 1.03*** (0.21) 1.02*** (0.20) 1.55*** (0.41) Spatial dispersion 14.90† (8.80) 16.43† (8.81) 54.69*** (11.80) 61.69*** (12.83) Tobin’s Q 1.10*** (0.20) 1.06*** (0.19)

• High-tech
• 2.38***

(0.27) 2.40*** (0.28) Tobin’s Q · Strategic alliances

• 0.59*

(0.26)

• High-tech · Strategic alliances
• 0.77†

(0.46) χ2 72.45*** 80.78*** 112.81*** 116.12***

bN=820. Positive coefficients indicate that increases in the variable tend to increase the

likelihood the firm will use an IPO prior to divestiture. † p<0.10, * p<0.05, ** p<0.01,

*** p<0.001.