* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Moving from private to public ownership: Selling out to
Investment fund wikipedia , lookup
Financial economics wikipedia , lookup
Present value wikipedia , lookup
Systemic risk wikipedia , lookup
Investment management wikipedia , lookup
Stock valuation wikipedia , lookup
Private equity wikipedia , lookup
Financialization wikipedia , lookup
Private equity secondary market wikipedia , lookup
Public finance wikipedia , lookup
Corporate venture capital wikipedia , lookup
History of private equity and venture capital wikipedia , lookup
Business valuation wikipedia , lookup
Private equity in the 2000s wikipedia , lookup
Global saving glut wikipedia , lookup
Private equity in the 1980s wikipedia , lookup
Corporate finance wikipedia , lookup
Moving from private to public ownership: Selling out to public firms vs. initial public offerings* Annette Poulsena, ** and Mike Stegemollerb a Terry College of Business, University of Georgia, Athens, GA 30602, USA b Rawls College of Business, Texas Tech University, Lubbock, TX 79409, USA Abstract We study the movement of assets from private to public ownership through two alternative means: the acquisition of private companies by firms that are public (sell-outs) or by initial public share offerings (IPOs). We use a matched sample to control for many of the industry- and economy-wide variables that other researchers have found to be important in the decision to go public. Our results suggest that firms will move to public ownership through an IPO when they have greater growth opportunities. Sell-outs seem to be preferred when managers are liquidating more of the firm and when they face financial constraints. February, 2005 JEL Classification: G34, G32 Keywords: initial public offerings, sell-outs, going public, acquisitions * We thank Jeff Coles, Laura Field, Kathleen Fuller, Randy Heron, Jim Linck, Sandy Klasa, David Mustard, Lance Nail, Jeff Netter, and Bill Petty for their many helpful comments. In addition, we appreciate seminar comments at the University of Georgia, Texas Tech University, and Texas A&M University, and at the 2004 Corporate Governance Conference at the University of Texas and the 2004 Financial Management Association Meetings. ** Corresponding author. Tel: (706) 542 3645; fax: (706) 542 9434. E-mail address: [email protected] (Annette Poulsen) Moving from private to public ownership: Selling out to public firms vs. initial public offerings Abstract We study the movement of assets from private to public ownership through two alternative means: the acquisition of private companies by firms that are public (sell-outs) or by initial public share offerings (IPOs). We use a matched sample to control for many of the industry- and economy-wide variables that other researchers have found to be important in the decision to go public. Our results suggest that firms will move to public ownership through an IPO when they have greater growth opportunities. Sell-outs seem to be preferred when managers are liquidating more of the firm and when they face financial constraints. 1. Introduction Takeovers of private firms by publicly traded firms (sell-outs) and initial public offerings (IPOs) are two methods through which privately owned assets move to public ownership. These transactions are comparable since they represent significant shifts in ownership structure, a channel for raising capital, and a means of liquidation for owners. However, there are fundamental differences between the transactions. While IPOs result in a significant change in ownership dilution (see, e.g., Mikkelson, Partch and Shah, 1997), private takeovers represent an even more dramatic change with prior owners frequently relinquishing their entire ownership stake. We consider firm-specific characteristics that may help to determine why a firm may choose to go public through an IPO when a sell-out to a public firm is a viable alternative for accessing public equity markets. Our evidence suggests the importance of growth considerations, financial constraints, and owner preferences for liquidity in the decision. IPOs are subject to great interest because they represent a significant change in the capital structure and governance of the firm. There is an extensive literature on why firms decide to go public through an IPO. Using aggregate and industry-level data, Lowry (2003) suggests that IPO volume is related to firms’ demands for capital and investor sentiment. Pagano, Panetta, and Zingales (1998) look at firm-specific data such as firm size and market-to-book valuations and report that Italian IPOs are more likely to be larger firms and to have higher valuations. Most closely related to our study, Brau, Francis, and Kohers (2003) look specifically at the IPO versus 2 takeover decision, relying on aggregate and industry-level data. They report that IPOs are more likely under macroeconomic conditions such as high cost of debt and a “hotter” IPO market, while takeovers are more likely in high market-to-book industries. Our work adds to this previous research by considering firm-specific factors that help to determine the choice of the method of moving to public ownership. We consider several factors that might influence the choice between a sell-out and an IPO, including the owner’s desire for liquidity, growth opportunities, financial constraints, and asymmetric information in firm valuation. Overall, our results support the importance of growth opportunities in the firm’s decision to undertake an IPO rather than a sell-out. We also find that sell-out firms are characterized by lower insider ownership both prior to and after the transaction as compared to IPO firms, suggesting that owners in sell-out firms prefer liquidity to control. Firms that are more difficult to value are more likely to transition through sell-outs to public firms as are firms with more liquidity constraints, perhaps due to difficulties in selling to diffuse shareholders when the firm is in financial distress. Our analysis uses a unique sample construction that allows us to consider specific firm characteristics that might affect the choice of the transition method. We identify 366 sell-out firms with sufficient data for analysis and then create a comparison IPO sample, matched by the value of the firm, the industry and time. Thus, we explicitly control for several of the macroeconomic and industry variables found to be important by previous researchers – the cost of capital, industry market-to-book ratios and the hotness of the IPO market, for example. Rather than relying on aggregate or industry data, we collect public filings from the private firms and from the acquirers in the takeovers to gather firm-specific data for each firm in our sample. These data allow us to directly compare similar firms and consider how each firm determined its means of moving to public ownership. The two samples are very comparable in average, median and aggregate size. The average, median and total size of sample sell-outs is $196.2 million, $113.1 million and $72 billion, respectively. For IPOs, these measures are $211.4 3 million, $122.6 million and $77.4 billion. Thus, our research is focused on larger-size sell-outs compared to average-sized IPOs. This analysis is particularly interesting since it is in this area that the managers have the clearest opportunities to decide between the two means of transition. The remainder of the paper begins with a background discussion of the relevant literature on IPOs, sell-outs and venture capital in section 2. We develop our testable hypotheses in section 3. Section 4 discusses sample selection and descriptive statistics of transitioning firms. The results of our empirical tests are in section 5 and section 6 contains concluding remarks. 2. Background Sell-outs are transactions where a public company buys all of the outstanding shares of a privately held firm. The private firm sometimes retains the service of an investment banker in facilitating the transaction. These transactions often result in target management selling a significant portion, if not all, of their ownership of their firm. In IPOs, however, the private firm generally sells off only a portion of the outstanding equity, with the previous owners retaining significant ownership and control of the public corporation. Sell-outs differ from IPOs in several other ways. From a takeover perspective, sell-outs represent the possibility for synergy between the firm and the acquirer (e.g., see Bradley, Desai and Kim (1988) and Mulherin and Boone (2000)). Sell-outs do not involve the lengthy public disclosure, do not require the additional costs of regulation faced by IPOs, nor are they bound by regulations on trading after the transaction such as the quiet period for issuing owners in IPOs. As Chemmanur and Fulghieri (1999) show, information gathering is done by a large number of investors in IPOs in contrast to a limited number of investors in a sell-out. In addition, any underpricing effects are gained by the acquirer’s shareholders. When private firms move to public ownership, they must choose between the alternative methods to do so. A recent Wall Street Journal (Grimes, 2004) story begins: IPO or sale? Sale or IPO? These days many young companies that seemingly are ready to go public through stock offerings are instead surprising the market and 4 agreeing to be bought by other companies, making the potential IPO moot. It’s a “bird in hand” strategy that is spreading its wings. When Viewstar Corporation’s decided to forego an IPO in lieu of a sell-out, the firm issued the following statement by Kamran Kheirolomoom, the President and CEO of Viewstar, illustrating a similar perspective: Although Viewstar had planned an IPO of Viewstar Common Stock and considered it an attractive opportunity for the Viewstar shareholders, the Viewstar Board of Directors has concluded that the anticipated benefits of the proposed merger with Digital will provide a better opportunity for the shareholders to realize the full value of their investment. Although both of these transaction types provide many similar benefits to the firm and its managers, academic literature and the popular press focus primarily on IPOs. It is generally stated by entrepreneurs that an IPO is the most desired form of “harvest” (see, e.g., Kensinger, Martin and Petty (2000)). However, Sahlman (1990) documents that more venture-backed firms resulted in sell-outs than IPOs in the 1980s (709 sell-outs versus 555 IPOs). Our research helps to determine the characteristics of firms that lead to a preference for one mode of transition versus another. Beneficial characteristics of a sell-out that are shared with an IPO include access to public debt and equity markets (through the parent in the case of the sell-out), liquidity of ownership previously tied up in an illiquid firm, and the possibility of linking management and employee compensation to traded securities. Firms that sell their assets to public firms are also able to access public markets with significantly lower regulator costs. Sell-outs may be less beneficial than IPOs in other ways, though. Management loses its ability to set firm policy due to dilution of ownership. In addition, it may be more difficult to raise capital for the firm’s projects since it would be competing with other projects of the acquiring firm in internal capital markets. Previous research suggests that both sell-out firms and IPO firms are profitable prior to going public, outperforming similar firms. Matsusaka (1993) studies sell-outs in the late 1960s to mid 1970s and finds that private firms undergoing a takeover are more profitable than comparison 5 public firms. He suggests that the transactions result from synergy considerations, not corporate control issues. Camerlynck and Ooghe (2000) examine a sample of private Belgium firm takeovers from 1992-1994. They also find that private firms involved in sell-outs are, on average, more profitable than their industry cohort and industry- and size-matched counterparts. Additionally, they show that these firms are highly liquid, have low leverage, and are less likely to experience financial distress than median firms within their industry. Similar analysis of operating performance has been performed for IPOs. Mikkelson, Partch and Shah (1997) and Jain and Kini (1994) report the operating performance of private companies before and after the IPO. Both studies find that IPO firms outperform their industry counterparts and that firms that go public are doing so when they are doing relatively well. Pagano, Panetta, and Zingales (1998) suggest that the high valuation may reflect market timing by firms when they go public and find that firm valuation drops quickly after the IPO. Multiples are often used as the basis by which firm value is assessed in the sell-out and IPO process. Koeplin, Sarin and Shapiro (2000) analyze a set of sell-outs and public takeovers from 1984 to 1998. They find that sell-outs are valued at a 20-30% discount to similar public takeover deals. However, the magnitude of the discount only holds for multiples of earnings and disappears when multiples of revenues are used for evaluation. Kim and Ritter (1999) analyze IPO multiples for 1992 and 1993. A comparison of the multiples from Koeplin et al. and Kim and Ritter suggests that IPOs are valued somewhat higher than the sell-out firms: the mean (median) market-to-book multiple for an IPO is 3.5 (3.0) vs. 2.4 (1.9) for a sell-out, and the IPO price-tosales multiple is 2.7 (2.1) vs. 1.4 (1.1) for a sell-out. Lerner (1993) finds that the return to investments in private firms that go public via an IPO is more than four times that of sell-outs for venture-backed private firms, seemingly justifying any higher valuation placed on IPOs and paid by investors. Thus, although the difference in pre-transaction performance is small there is evidence that firms that choose to go public via an IPO are valued more highly than firms in a sell-out. 6 For an IPO, the initial offering return is referred to as underpricing and represents a transfer to the new public owners of the firm. In his analysis of IPO underpricing, Ritter (1987) documents that the average stock-price return on the day of an IPO is 14.8% for firm commitment offers and 47.8% for best efforts for 1028 firms taken public from 1977 to 1982. Fuller, Netter and Stegemoller (2002) report the returns to bidding firms from acquisitions of private targets in the 1990s. They find that private acquisitions result in a 2% average return to bidders for the five days surrounding the announcement of the acquisition. When the sell-out dollar returns to the bidder are viewed as a percentage of the target's value, the announcement day returns for the bidder averages 49.7% of the value of the sellout firm. This return to the acquirer is similar to Ritter's finding on underpricing in best effort IPOs. Thus, returns to purchasers of private firms may consist of both synergy gains and underpricing effects. The market assessment of the value transferred from the old owners to new owners seems to be similar in sell-out and IPO transactions. 3. Determinants of Method of Transition from Private to Public Assets We hypothesize that firm-level characteristics, such as the owner’s desire for liquidity, the firm’s growth opportunities, the firm’s liquidity constraints, and difficulties in valuation of the firm will influence private owners’ decision between undergoing an IPO versus a sell-out. While earlier work by Brau, Francis and Kohers (2003) suggests the importance of macroeconomic and industry factors in the sell-out vs. IPO decision, our matching procedure holds these economywide considerations constant. Instead, our inspection of the public filings of our sample firms allows us to better understand the motivating, firm-specific factors for each firm’s choice. In this section, we provide additional discussion on the influence of firm characteristics on the choice of the method of transition made by the entrepreneur and other owners of the firm. 3.1. Liquidity and ownership considerations Perhaps the most significant consequence of a going public transaction is the resulting dilution of ownership. Pre-transaction owners must consider the amount of control they will lose 7 and the resulting trade-offs. In an IPO, owners sell off a portion of their stake in the private firm. The dilution effects of an IPO are relatively minimal since the average IPO in our sample leaves the original owners with approximately half of the post-transaction ownership of the firm. However, a sell-out entails selling the whole entity to a public company that is most often comprised of diverse shareholders. Unless the acquiring firm pays with stock, the owners of the private company retain no ownership of the newly merged company. In the event that stock is used as a method of payment, the private company is typically small enough that the original owners acquire a small block of the merged firm’s stock. Ownership, or the level of control rights, is negotiated throughout the life of a private firm. As the firm has need for capital infusions, insiders must decide among alternative types of capital investment that have differing effects on control rights. In the years or months leading up to the transition from private to public ownership, owners have had numerous opportunities to relinquish ownership and control and perhaps even fully exit their initial investment altogether. “Angel” investors, venture capitalists, banks, large corporations and other institutional investors offer entrepreneurs capital in exchange for an ownership stake in the private firm. Many of these entities invest in private firms simply to recognize a capital gain when the firm is sold to the public, to another firm, or via management buyout. Those firms where insiders have a smaller stake in the firm may value their control rights significantly less than those firms where insiders have struggled to retain control over the decisions of the firm. This viewpoint suggests that firms with low insider ownership will be more likely to sell out to a public corporation via a takeover, whereas firms with a high percentage of insider ownership will be more likely to undertake an IPO. An alternative perspective suggests that private firms with low insider ownership have convinced outside investors that the firm and its management team are of high enough quality that the firm is worthy of an investment. When such a firm needs to raise additional capital, it might prefer to use an IPO and to not force high quality management to relinquish control. It 8 follows that low insider ownership might imply a higher value attached to management retention and control and would thus be associated with a higher probability of choosing to move to public ownership through an IPO. If venture capitalists are better at identifying higher quality management, then this scenario might be especially true when they have made significant investments in the firm. 3.2. Growth and the need for capital We expect the growth characteristics of the private firm to influence the choice of transition method for private-firm owners. A firm may range from being a capital-starved firm with many growth opportunities, to a mature firm producing a great deal of cash flow but having few positive net present value projects in which to invest. In an IPO, the private firm raises public capital and allocates it to projects that management deems most important. The ability to raise public capital is also relevant for sell-outs but in a constrained framework. After the sell-out is completed, the investment opportunities of the sell-out firm must compete with other subsidiary operations for scarce resources within the merged firm in the internal capital market. Stein (1997) suggests that the internal capital market may enhance the value of the overall firm as managers are able to allocate funding to winners, known as winner picking. However, Stein also shows that these same firms may be likely to participate in “loser sticking” – allocating funds to poorly performing projects on the basis of it being a “favorite” project. Since sell-out firms are competing in the internal capital market for funds, high-growth firms might avoid a sell-out due to the constraints imposed by that market. By undertaking an IPO, the firm may have greater flexibility in accessing resources, especially through its new access to the equity markets. Thus, we hypothesize that firms with greater growth potential will choose to go public through an IPO rather than through a sell-out. Myers (1984) suggests that there is a strong link between a firm’s growth options and its capital structure. Smith and Watts (1992) and Gaver and Gaver (1993) find empirical support for this premise. In particular, they find that public firms with more growth opportunities are more 9 likely to use equity financing than those firms with fewer growth opportunities. In the case of sell-outs and IPOs, capital structure may be the constraint that spurs a firm to seek public financing. A firm with positive investment projects but constrained by a large amount of debt may raise sufficient funds for its projects through an issuance of public equity. Similarly, a firm with few investment opportunities may seek to be purchased by a firm with a larger capacity for debt, thereby reducing free cash flows and agency costs. Using measures of overall growth in the economy, Lowry (2003) finds a positive relation between the economy-wide demand for capital and IPO volume. Pagano, Panetta, and Zingales (1998), however, consider growth measures for a sample of Italian firms that choose to go public and find little support for the importance of the firm’s recent growth. In addition to considering the impact of potential growth, firms that are considering moving to public status either though an IPO or a sell-out to a public company may do so because they have moved beyond their optimal amount of debt financing or because their liquidity is insufficient to fund the obligatory fixed payments accompanying debt. An IPO or sell-out provides the opportunity for alternative sources of equity financing. Since IPO firms may be less constrained in their ability to raise equity relative to sell-out firms, we expect that firms that are more capital constrained would choose an IPO over a sell out. Looking at economy-wide factors, Brau, Francis, and Kohers (2004) find some evidence that transitioning firms are more likely to go public via an IPO when Treasury bill rates are higher though their results considering other measures of the demand for funds are insignificantly different from zero. We look at leverage and cash constraints of the firms in our sample to see if firms with greater cash constraints are more likely to choose an IPO over a sell-out to a public firm. 3.3. Asymmetric information The buyer’s ability to gather and properly assess information about the firm seeking transition may play an important role in determining the method by which a firm moves to public ownership. Ellingson and Rydquist (1997) argue that firms with assets that are not easily valued 10 by diverse public shareholders are more likely to choose a direct sale to another firm or individual. Subscribers to an IPO are, in general, institutional managers that do not have particular expertise in the operational intricacies of the private firm. These managers then offer the shares to a dispersed group of even more uninformed investors. While investment bankers and money managers are more informed than the general investor, it is still difficult for them to value a set of assets that have unique qualities. In contrast, another company operating in a similar environment to the private firm would be better able to accurately value these firm-specific assets. Additionally, firm-specific information may retain its value only when the information is not accessible by outside competitors, as suggested by Chemmanur and Fulghieri (1999). By keeping information undisclosed, the private firm is able to hold a competitive advantage over other firms in the same industry, both public and private. Undertaking an IPO exposes firmspecific information much more than if the firm was acquired by another public company. We expect that firms that are more difficult to value will be more likely to use a sell-out to transition to public status.1 Firms with assets that are more easily valued by dispersed public shareholders are more likely to choose an initial public offering. However, we note that this hypothesis may be difficult to measure empirically. For example, high-growth firms are generally considered to be firms where it is difficult to identify the future prospects of the firm. Since we expect high-growth firms to be more likely to use equity markets to finance that growth, as noted in the previous section, the two hypotheses suggest opposite empirical predictions. In addition to growth, however, we offer several alternative measures of information availability including the age and the profitability of the firm. We expect that there will be less information for more profitable private firms and firms that invest less in R&D since individual investors and money 1 From a different perspective, Zingales (1995) argues that IPOs can give managers of private firms a means to establish a market value of the company before liquidating their position. Field and Mulherin (2003) show that IPOs are followed by a higher rate of takeover in the few years following the transaction than other publicly traded firms. This pattern is not confirmed in our data; by the end of 2003, 120 of the acquirers of the sell-out firms and 120 of the IPOs had been acquired or merged into another firm. 11 managers may use historical returns as an indicator of future profitability or investments in R&D may be difficult to value. 4. Sample Information and Empirical Results We select the sell-out and IPO samples from Securities Data Corporation (SDC) databases on U.S. Mergers & Acquisitions and U.S. Global New Issues, respectively. Dates are restricted to 1995 – 1999 for the announcement date of sell-outs and the issue date of IPOs. This period has several advantages. These five years represent an especially active IPO and takeover market. There were many transactions in many different industries, allowing us to examine the breadth of the market in our analysis. In addition, the Securities and Exchange Commission’s EDGAR database began keeping electronic filings in 1995 for sell-outs and in 1996 for IPO prospectuses, increasing data availability. Table 1 presents descriptive statistics of the frequency and size of IPOs and sell-outs on SDC. We examine all listings on SDC, and require some basic information to be included in this summary table. IPOs must be of US firms, must be listed on NYSE, AMEX or Nasdaq, must be an original IPO, must not be a spinoff and must be an offering of common stock. Sell-outs must have a disclosed dollar value, must be a private US firm, the acquirer must be traded on the NYSE, AMEX, or Nasdaq, and the acquirer must acquire 100% of the target. In addition, financial and utility firms are eliminated from both samples. The number and value of sell-outs is understated since acquiring firms are not required to report, nor publicly announce, insignificant acquisitions. The following statement from a February 4, 2002 Wall Street Journal article (Maremont, 2002) illustrates this point: Tyco International Ltd. said it spent about $8 billion in its past three fiscal years on more than 700 acquisitions that were never announced to the public. … [Tyco’s chief financial officer] said the company doesn’t disclose details on its numerous smaller deals because they aren’t “material” given Tyco’s huge size. Panel A of Table 1 reports the number and value of sell-outs and IPOs as reported on SDC, given the above restrictions. The median deal value for the 4,455 sell-outs reported on SDC 12 is $14.5 million. In contrast, the median market value of the 1,677 IPOs is $109.8 million. (Market value of IPOs is measured as the number of shares outstanding multiplied by the midpoint of the share price on opening day. We do not use the number of shares sold in the IPO since most IPO transactions are for less than 50% of firm ownership.) Average values vary similarly – the average sell-out is valued at $51.3 million and the average market value of an IPO firm is valued at $243.3 million. The total value of firms involved in IPOs dominates the total value of firms undergoing a sell-out; from 1995 to 1999, $408 billion transitioned from private to public ownership via IPO versus $228 billion via sell-out. The mean and median deal size increases over the period analyzed, also. Sell-outs more than double in the amount paid per deal from an average (median) deal value of $31.5 ($10.3) million in 1995 to $83.1 ($20.0) million in 1999. This trend is even more distinct for IPOs; the average (median) deal value in 1995 is $120.4 ($72.8) million and is $464.9 ($250.6) million in 1999. The frequency of IPOs relative to sell-outs also changes over the five-year period, reflecting the cyclical nature of “hot” IPO markets. For IPOs, 1997 and 1998 were the two lowest years in terms of number of deals. The same years were the highest two years for sell-outs. We place several restrictions on our sample for the rest of our analysis. We first identify all sell-out firms that match our criteria and then select matching IPO firms. We outline below our sample selection criteria for sell-out firms. Starting with the U.S. Mergers & Acquisitions database, we identify all targets with a disclosed dollar value and limit the sample further as indicated below: 1. The deal value must be for at least $50 million. SDC defines deal value as the total value of consideration paid by the acquirer, excluding fees and expenses. This step leaves 22,485 firms. 2. The target must be a private firm based in the United States, leaving 7,471 firms. 3. The acquirer must be a publicly traded U.S. firm traded on the AMEX, NASDAQ or NYSE, leaving 1,149. 13 4. Financials and utilities are removed, leaving 813 firms. 5. The deal value must be at least 10% of the acquirer’s net assets. Ten percent is the level of materiality as defined by the Securities and Exchange Commission. Significantly more data are available from acquirer filings when the target crosses this threshold. This step leaves 608 firms. 6. The firm is not a spin-off, roll-up, nor a subsidiary. This leaves 577 firms. 7. There must be financial data for the private target in the form of SEC filings such as S4, 8K, S3, Proxy, or S1 filings. This leaves 549 private takeover transactions. 8. Of the 549 possible transactions, we are able to find 366 transactions with at least one year of historical financial statements. All of the data used in the analysis are collected using Disclosure Global Access and the Securities and Exchange Commission’s EDGAR database. Data for the sell-out sample are generally available through filings made by the acquiring firm. Securities regulation S-X states that “if securities are being registered to be offered to the security holders of the business to be acquired, the financial statements…shall be furnished for the business to be acquired....” Acquisitions that are less than 10% of the acquirer’s total assets are exempt from this regulation. Firms that issue equity in conjunction with a business combination file S4s, while significant transactions are also detailed 8K, S1, S3 and proxy filings. The S4 filings provide historical financial statements, and other information such as records of fees paid by the acquired firm to the investment banker employed, a timeline of events leading to the purchase, reasons for the transaction, and the ownership structure of the firm prior to and after the acquisition or IPO. Matsusaka (1993) only studied sell-outs that were fully or partly financed with an equity issuance, while approximately 30% of sell-outs in our sample are cash deals that have no partial equity financing. Our data are also limited in that many sell-out firms do not have a full 3 years of historical financials. This is due, in part, to the large number of development stage corporations 14 that are acquired. In addition, there are requirements found in regulation S-X that relate the number of years of historical statement reporting required to the relative size of the target to the acquirer.2 From the sample of private takeovers, a matched set of IPO firms is created, matching by time of transaction, industry and size. Industry classifications are based on Fama and French (1997). We could not identify Fama-French matches for 22 firms, so they are matched by 2-digit (8 firms) or 1-digit (14 firms) SIC codes. For IPOs, size is defined as the market value of all shares following initial trading (not just those sold in the offering); for sell-outs it is defined as the deal value as reported by SDC. The full IPO sample is collected from SDC given the following constraints: 1. The IPO firm must be based in the United States and traded on the NYSE, AMEX, or Nasdaq. 2. Financials and utilities are excluded. 3. The market value of the company performing the IPO is more than $50 million. These restrictions result in 1,290 IPOs. From these firms, we choose a matched sample of IPOs for the sell-out sample. The average (median) absolute difference in firm size is $59.8 million ($7.9 million) and the average (median) difference in time of the offering is .5 years (0 years). Panel B of Table 1 provides the similar data as in Panel A, but the sample is limited to the 366 sell-outs in our sample and the 366 matching IPOs. Note that we match the 1995 sell-outs to IPOs from 1996 since we were not able to identify IPO prospectuses on either EDGAR or Compact Disclosure for 1995. The two samples are very comparable in average, median and aggregate size. The average, median and total size of sample sell-outs is $196.2 million, $113.1 million and $72 billion, respectively. For IPOs, these measures are $211.4 million, $122.6 million 2 Additional reading on these requirements can be found in Regulation S-X, Rule 3-05 – Financial Statements of Businesses Acquired or to Be Acquired, available at the SEC web site, http://www.sec.gov. 15 and $77.4 billion. Relative to the full sample, our research is focused on larger-size sell-outs compared to average-sized IPOs. Our focus is particularly interesting since it is in this firm size range that the managers have the clearest opportunities to decide between the two means of transition. In Table 3, we report the reasons for transition from private to public ownership as stated in SEC filings. We were able to collect reasons for transition from 78 sell-out documents, concentrated in S4 statements, and in all 366 IPO prospectuses. The discrepancy in number of filings is due to the uniform filing requirements for IPOs and the more scarce and non-mandatory nature of the corresponding information for sell-outs. We group the rationale for transition into several broad categories and then further categorize the reasons offered within those groups. Our broad categories include reasons related to access to capital and growth, debt, payouts, marketing and personnel, and other reasons. For IPO firms, the reasons given for transition focus mainly on the capacity of the firm to grow. More than 80% of the IPO firms mentioned the need for capital access or the desire for funding to achieve growth through, for example, acquisitions, research and development, or capital expenditures. IPO firms also specifically cite the ability to raise working capital (58%), and the ability to fund future acquisitions (35%), capital expenditures (29%), research and development (18%), and expansion (15%). In addition, 55% of the IPO firms specifically mentioned the desire to reduce debt. All of these reasons suggest that the IPO firms are more growth oriented than sell-out firms. In contrast, sell-out firms place greater emphasis on liquidity and the ability of the owners to “harvest” their initial investment. Only 68% of the sell-out firms mention access to capital or growth in some way as compared to 80% of the IPO firms. However, 88.5% of the sellout firms mention future payouts of the firm as the key determinant for the transition, compared to 18.9% of the IPO firms. More than 78% explicitly state they are looking for liquidity, another 5.2% say they plan to pay a distribution or repurchase stock and 35.9% indicate that a sell-out 16 would have favorable tax consequences with respect to payouts. Of the 18.9% of the IPO firms mentioning payouts to shareholders, 7.4% of those cases reflect the repurchase of preferred stock, which is probably held by venture capitalists. Sell-out firms are also more likely to mention reasons that reflect strategic considerations, such as marketing abilities or other synergistic benefits from merging into another firm. Table 4 provides median values for accounting, ownership and other firm-specific variables for our samples of sell-out and IPO firms. In general, we identified the balance sheet and income statement items for almost all of the firms in both samples. However, ownership data were more difficult to identify for the sell-out firms. As can be seen in these data, sell-out firms are generally significantly larger than IPOs in terms of revenues, total assets, and several income measures in terms of both medians and mean differences for each matched pair, even though the samples were matched on the basis of the value of the transaction. Sell-out firms also tend to be older than IPO firms, and have more employees. IPOs generally have greater insider ownership prior to the transaction and maintain higher insider ownership after it. Thus, these data suggest that sell-out firms tend to be larger, older and more established than the sell-out firms. We refer back to these descriptive statistics in the following analysis of our empirical results. 5. Empirical results We first report summary data and univariate tests on our variables of interest and follow these results with our logit regression analysis of the choice between the methods of going public. While the logit analysis provides a more comprehensive view of the impact of the explanatory variables, it requires complete data for every observation. With the scarcity of data for the sell-out sample, the univariate analysis allows us to provide summary information about differences between the samples to the extent possible with available data. 5.1. Insider ownership While sell-out and IPO transactions are both means to change the firm from private to public ownership, a very basic difference between the two transactions is the post-transaction 17 ownership of insiders. In sell-out transactions, insiders maintain ownership only if they accept payment in the form of the acquirer’s shares and, in that case, their control rights are diluted substantially. In IPOs, however, the insiders can establish how much of the company they want to keep and what fraction they wish to sell in the IPO.3 Thus, we expect IPOs to be more likely in those cases where insiders wish to maintain control. Table 4 provides evidence consistent with this hypothesis though we only have insider ownership data for 73 of the sell-outs prior to the transaction and 155 of the sell-outs after the transaction, including 105 of the sell-outs that were for all cash where we assume the posttransaction ownership stake is 0%. The median insider ownership prior to the transaction of firms purchased by public companies is 62.7%, which is significantly different from the 71.8% insider ownership of firms that chose to undertake an IPO. The disparity is also evident in the difference in the first quartile ownership (35.9% for sell-outs vs. 51.0% for IPOs) and the third quartile ownership (78.1% for sell-outs vs. 90.8% for IPOs). Thus, even prior to the transaction, insiders in sell-out firms have significantly lower holdings than in IPO firms. More distinct than the difference in pre-transaction ownership, however, is the difference in post-transaction ownership of insiders. The median ownership of insiders after the transaction for sell-outs is 0% vs. 50.8% for IPOs. For the 50 sell-outs where the selling managers maintain an ownership position in the acquiring company, the median ownership is 19.5%.4 Pre-transaction owners lose control of the firm in almost all sell-outs. In IPOs, at least 50% of the pre-transaction owners keep control of the firm. The difference in the change in ownership is statistically significant and is consistent with sell-outs being more of a liquidity event for the owners rather than a means to raise capital for the firm. 3 Downes and Heinkel (1982) and Ritter (1984) provide evidence suggesting that prices in an IPO are higher for those transactions in which insiders retain a larger proportion of shares. 4 The percentage change in insider ownership in each firm is also reported in Table 4. The median change in insider ownership is 45.9% for sell-outs and only 17.3% for IPOs. The change in ownership at the lowest quartile is 27% for sell-outs and 10.6% for IPOs and at the largest quartile it is 65.6% for sell-outs and 26.6% for IPOs. 18 5.2. Growth and capital needs Table 5 presents univariate statistics on different growth measures for sell-out and IPO firms to help assess the growth prospects of the firm. We find that prior to the “going public” transaction, IPO median sales growth is 39.0% for the year prior to the transaction and 32.2% for the year before that, as reported in Panel A. These growth rates are significantly greater than the median growth in sales experienced by sell-out firms of 22.1% and 19.1% for the year prior and the two years prior to the transaction, respectively. The median growth in sales over the full twoyear period is 89.4% for the IPO firms and 43.0% for the sell-out firms, with the difference again significantly different from zero. We also measure the aggregate growth of the firms by growth in total assets and growth in capital expenditures. Median asset growth for IPOs in the year prior to the transaction is 27.0%, which is significantly higher than the median growth in assets for sell-outs of 14.1%. (Balance sheet data are only available for two years prior to the transaction.) We find similar results for capital expenditures. Growth in capital expenditures is significantly higher in the year preceding the transaction in the IPO firms (median equals 56.5%) relative to the sell-out firms (median equals 16.3%). This evidence is consistent with our hypothesis that private firms with greater growth potential will prefer to transition to public status through an IPO as opposed to a sell-out. We differentiate the sample on the basis of whether the private firm had received venturecapitalist backing. Gompers (1995) suggests that venture capitalists are more likely to invest in high growth firms and firms with more uncertainty about their future earnings. In our sample, 119 of the sell-out firms are backed by VCs, while 177 of the IPOs are backed by venture capitalists, a significant difference. For those firms with VC backing, the involvement of the VCs is similar. Given that a VC has backed the firm, the median number of years invested in the firm is 3 in both samples, the median number of rounds of VC financing is 3 for sell-outs and 4 for IPOs, and the 19 median number of VC funds investing in the firm is 4 for sell-outs and 5 for IPOs, with the latter two measures significantly higher for IPOs. In Panel B of Table 5, we report support for Gompers’ suggestion that venture capitalists are more likely to invest in higher growth firms, whether the firm eventually goes public through a sell-out or an IPO. Growth in sales is significantly higher for VC-backed firms for both sell-outs and IPOs, and the same relation is observed for growth in assets and capital expenditures, though the difference is not always significant. Valuation multiples provide an additional measure of the growth opportunities of the firms in our sample. In general, higher valuation multiples of either assets or earnings is suggestive of future growth. The dollar amount that investors pay for a private firm serves as the only observable market value that is comparable across IPOs and sell-outs. In Table 6, we document the dollar amount paid for the firm as a multiple of total assets and sales. We do not consider multiples of earnings or cash flows due to problematic nature of negative values for these accounting items. We measure the market value of sell-outs as the deal value reported by SDC. For IPOs, we use the product of opening market price and total shares outstanding. The median market value to book value of assets for IPOs is 6.1, significantly greater than the 2.9 ratio for sell-outs, suggesting that IPOs have more growth opportunities and therefore are valued more highly than a comparable sell-out firm. Confirming this observation, the market value to sales ratio is also significantly higher for the IPO sample as compared to the sell-out sample. In Panel B of Table 6, we consider the impact of VCs on the valuation multiples. Whether backed by VCs or not, we find that both the market value to assets ratio and the market value to sales ratio is higher for the IPO firms. In addition to the relation between growth and capital structure and the choice of the method of transitioning to public ownership, we consider the impact of leverage, liquidity and financial distress on that choice. Panel A of Table 7 reports differences in the debt structure of the firms. In general, the two samples are very similar in their debt. Total debt to assets for both sets 20 of firms is 71% in year -1 and close to 70% in year -2. There is no significant difference in interest payments to total debt between the sell-out firms and IPOs. We do find significant differences between the two samples in measures of liquidity and financial distress. In Panel B, we report the cash to total asset ratio and the current asset to current liability ratio as measures of the liquidity of the firms. Median scaled cash holdings for sell-out firms for years -1 and -2 are 6.4% and 5.9%, respectively. Scaled cash holdings for IPOs are significantly higher than that of sell-outs at 8.6% and 9.0% for years -1 and -2, respectively. There is no significant difference between the current ratio for the firms. The greater cash holdings of IPO firms contradict the hypothesis that these firms sell public equity because they are unable to meet their short-term liquidity needs. Panel C considers additional measures of potential financial distress. We report here the percentage of firms with interest payments greater than free cash flow, as measured with EBITDA and the percentage of firms with interest payments greater than cash holdings. Surprisingly, these two measures are contradictory. We find that in the year prior to the transaction 32.5% of IPOs have interest expenses that exceed their EBITDA, significantly greater than the 20.3% of sell-outs where this is true. However, when we consider interest expenses relative to cash holdings of the firm, significantly more (35.3% vs. 28.7%) of the sell-out firms have interest expenses that exceed cash holdings. We combine these measures in an interactive variable in the regression analysis to further consider their impact on choice of transition method. Our results seem to suggest that IPO firms have lower cash flows but not lower cash holdings relative to sell-out firms. The low and often negative cash flows of the IPO firms may reflect the relatively early development stage of growing firms. 5.3 Asymmetric information The ability of a firm to produce sustainable profits may have a significant impact on the marketability of a firm to the general public. The earnings of a firm may serve as the single best proxy for the probability of future positive performance. Conversely, poor profitability may result 21 from a firm being in the development stage, or it may be the result of poor management or inefficiencies due to small scale. We measure pre-transaction performance with two accounting measures: operating income scaled by sales and operating income scaled by total assets, as reported in Table 8. Both return measures indicate that sell-out firms are more profitable than IPO firms before the transaction occurs. Median returns on sales for sell-outs range from 8.7% in year –1 to 6.1% in year –3; these numbers for IPO firms range from 3.2% in year –1 to 2.5% in year – 3. Similar results are found when operating income is scaled by total assets. Median return on assets for sell-outs is 8.7% in year –1 and 8.0% in year –2. Again, the return measures are lower for IPO firms; they are 3.2% in year –1 and 2.9% in year –2. The initial profitability results are not consistent with our hypothesis that, if profitability is a proxy for less asymmetric information, more profitable firms will undertake an IPO. We also consider returns after adjusting for industry performance, as reported in Table 8. We calculate the abnormal operating return by subtracting an industry return from the private firms’ return measure. Our measure for the industry return is the return on an equal-weighted portfolio of publicly traded firms in the same two-digit SIC code. Overall, we find the performance of private firms significantly below that of publicly traded firms in the same industry. The median industry-adjusted return on sales for sell-outs ranges from -2.8% in year -1 to -3.8% in year -2. For IPOs, the industry-adjusted measures range from -5.6% to -7.7%. Most of the industry-adjusted measures for both sets of firms are significantly less than zero and the IPO measures of return are significantly less than the comparable measures for the sell-out firms. The industry-adjusted results confirm the findings from the non-industry-adjusted measures that sellout firms do not seem to be subject to more asymmetric information. The age of the firm is an alternative proxy for possible asymmetric information. Cumming and MacIntosh (2001), in a study on venture capital exits in Canada and the United States, suggest that older firms tend to have a more established market and management team, and a longer operating history. These characteristics reduce informational asymmetries between 22 the owners and potential investors. In Table 4, we provide results on differences in firm age. We are able to determine 191 founding dates for sell-outs and 343 for IPOs. We find a significant difference in the median age of IPOs versus sell-outs, though again the results are counter to the asymmetric information hypothesis. At the 25th percentile, the firms are 4 years old for both sets of firms. At the 50th and 75th percentile operating history is 8 years and 15 years for IPOs and 10 years and 23 years for sell-outs. Additionally, the average ages of IPOs and sell-outs are 14 and 19 years, respectively. These results suggest that the reduced profitability of IPOs versus sell-outs may be the result of being in the early stages of development for IPO firms and not their overall quality. In Table 9, we report three additional measures of asymmetric information in the private firms at the time of, and leading up to, the transition. The first is research and development expense scaled by firm sales. The value of research and development is generally difficult to value by the general market, but is perhaps more easily valued by acquiring firms with which there are synergies to be gained. We find that the dollar amount invested in research and development, whether scaled by sales or assets, is significantly greater for IPOs than sell-outs, as measured by the Wilcoxon test statistic. However, the median level of research and development is zero percent of either assets or sales for both sets of firms. We also measure information asymmetries with intangible assets scaled by total assets, and find no significant differences in these ratios. Thus, these alternative measures of asymmetric information issues again do not support the importance of information asymmetry in determining the transition method for private firms. 5.4. Logistic Regressions We use logistic regression analysis to provide more comprehensive analysis of the various univariate tests provided above. While univariate tests can offer insights into factors that might be important in determining the choice of transition method, multivariate analysis allows better understanding of the various interactions between the variables. We present results from 23 our logistic regression analysis of factors influencing transition method for private firms that accessed public equity markets via a sell out to a public firm or an IPO in Tables 11 and 12. Table 11 presents results based on the largest sample possible. Table 12 provides additional results where the sample size is allowed to fall as we introduce additional explanatory variables that are not available for all firms. Table 10 reports the correlation matrix for the variables used in Table 11. The explanatory variables represent the broad categories of hypotheses investigated. In the initial regression presented, we proxy for growth of the private firms through scaled capital expenditures and whether the private firm was backed by venture capitalists. Scaled capital expenditures indicate whether the firm is actively expanding its asset base. This measure is available for all but one firm since it relies on only one year of data. Backing by venture capitalists is generally associated positively with growth in the firms, as reported in Tables 5 and 6. The venture backing variable is available for all firms in the sample while we lose a significant number of observations when we use the change in assets or sales directly in the regressions. We include the alternative measures of growth in robustness testing reported in Table 12. We measure the importance of capital constraints with the leverage of the firm (total debt scaled by total assets) and also a variable indicating whether interest expense is greater than the cash holdings reported on the balance sheet. In the last three regressions in Table 11, we add an indicator variable for whether interest expense is greater than EBITDA and also an interactive variable that measures whether EBITDA and cash are both less than interest expense. We consider the latter a measure of financial distress since the firm does not have sufficient resources to pay interest expenses from either source. The third hypothesis we investigate is whether firms that are more difficult to value are more likely to transition via a sell out rather than an IPO. It may be easier to market and sell a difficult-to-value firm to a closely related entity or to a small group of buyers. We use several measures to proxy for valuation uncertainty. First, since firms that have greater accounting profits 24 may be easier to value, we use industry-adjusted return on assets as a measure of the profitability of the firms. R&D (scaled by assets) is an alternative measure of difficulty in valuation, with firms having more R&D being more difficult to value. We use the log of sales in the year preceding the transaction as a third measure of difficulty in valuation, which proxies for the size of the firm with the expectation that larger firms should be easier to value. We find support for our hypothesis concerning growth of the firm. Those firms that have higher capital expenditures are significantly more likely to perform an IPO as are firms that receive VC backing. In both our univariate analysis and the regression analysis, our results suggest that those firms that are growing rapidly have a greater need to access equity markets through an IPO rather than through the more constrained environment of being part of another firm. We also find support for the importance of other capital structure considerations. While there was not a significant difference in leverage between our two samples in the univariate tests, in the regression analysis we find that firms with higher leverage, as measured by total debt scaled by total assets are more likely to choose an IPO. Since an IPO results in a lower debt to asset ratio and since evidence suggests that higher growth firms choose to be less dependent on debt, this result is consistent with our evidence that higher growth firms choose to go public through an IPO. The coefficient on the indicator variable showing whether interest expenses are greater than the firms’ cash holdings is insignificantly different from zero (as is the coefficient on whether interest expenses are greater than EBITDA in unreported regressions), downplaying the importance of liquidity constraints in the transition decision. However, when we include the two indicator variables as an interaction term – i.e., a value of one indicates that the firm has insufficient cash holdings or EBITDA to cover interest expenses – we find that the firm is more likely to choose a sell-out. This result suggests that those firms that are most liquidity-constrained (or in financial distress) find accessing the equity market through an IPO less attractive or are less 25 able to access those markets directly. Thus, this result combined with our earlier growth results emphasize the importance of growth, rather than liquidity constraints, in leading firms to enter into an IPO. We also consider whether asymmetric information and difficulty in valuing the firm has an impact in the choice of transition method. We use several measures to proxy for difficulty in firm valuation. The industry-adjusted return on assets, a measure of profitability, and R&D expenses have a significant correlation coefficient of -.75 and when they are entered in the regression together, both are insignificantly different from zero suggesting a multicollinearity issue. When entered separately, they are each significantly different from zero. Firms that are more profitable are significantly more likely to go public via an IPO, as are firms with less R&D. These results confirm the expectation that firms that are more difficult to value are more likely to access public markets by marketing and selling the firm to a small set of buyers that can more readily value the firm. Our univariate results suggested that more profitable firms would choose a sell out. Here, once factors such as growth and capital constraints are controlled for, we confirm our expectation that more profitable firms choose to go public via an IPO. Our third measure for difficulty in valuation is the log of sales in the year preceding the transaction, with the expectation that firms with more sales would be less difficult to value and therefore be more likely to choose an IPO, analogous to our finding with respect to profitability. However, we find the opposite – firms with more sales are significantly less likely to access the public markets through an IPO. This finding is consistent, however, with the stylized fact that many firms that went public through an IPO, especially in the late 1990s, were in the initial early stage of development and expecting to grow rapidly. In many cases, these relatively young firms had very low sales. Fourteen of the firms in our IPO sample had sales less than $50 million; none of the sell-out firms had sales below $50 million. In general, it seems that the expectation of rapid growth dominated the information asymmetry effect in our sample. 26 In Table 12, we present several alternative specifications of the regressions in Table 11. In the first three regressions, we add alternative measures of growth for the firm – sales growth in regression one, asset growth in regression two, and capital expenditure growth in regression three. Because these variables require two years of data, they are not available for the full sample. However, even with the reduced sample size, we find that the latter two measures of growth are associated with significantly higher probability of accessing public equity markets through an IPO. In the fourth regression of Table 12, we include insider ownership after the transaction. We lose more than half of the asset sale sample when we include this variable but we do confirm the univariate tests showing that insiders are likely to keep higher control when they undergo an IPO. Most of the other variables, however, become insignificantly different from zero, presumably because of the reduced sample size. 6. Conclusions There are two primary means through which privately held stock is transferred to public owners – acquisitions of private firms by public corporations and initial public stock offerings. These transactions are similar since they are both channels for accessing the public equity market, reflect significant shifts in ownership, and are a means of liquidation for owners. However, they have many differences in terms of the dilution of ownership, access to public equity markets, liquidity of owner investment and structure of post-transaction management. We compare these transactions for insights into the decision-making process of entrepreneurs and institutional owners with respect to the one of the largest transactions a firm may undergo. We use a unique sample of 366 sell-outs and 366 IPOs that are matched by market value of the firm, industry and year that allows us to directly compare similar firms that could choose either method of transition. We gather firm-specific information from public filings of the firms themselves or of the acquiring firms, while most previous research in this area has relied on industry- or economy-wide data. In addition, by focusing on firms that have an average market 27 value of about $200 million ($120 million in median value), we are directly comparing firms that could viably choose either transition method. The evidence suggests that firm characteristics contribute significantly to the decision of whether a firm accesses public equity markets through an IPO or a sell-out. Most importantly, firms that are characterized by high pre-transaction growth and higher valuations are more likely to transition with an IPO. In addition, we find that insiders of sell-out firms dramatically reduce their control of the firm while insiders of IPOs maintain significant stock ownership and that firms that are more difficult to value are more likely to transition through sell outs to public firms. Our results also suggest that firms with more liquidity constraints may be more likely to choose a sell-out, perhaps due to difficulties in selling to diffuse shareholders when the firm is in financial distress. Our results add to our understanding of this very important decision made by firms looking to move to public ownership. In addition, we contribute to the general understanding of why firms choose to do an IPO. Through our matched sample, we control for many of the industry considerations and macroeconomic factors such as cost of capital and periods of market overvaluation that have been suggested by previous authors as explanations for the decision to undertake an IPO. By contrasting the IPO and sell-out decisions with extensive firm-specific information from the various filings related to the transaction, we are able to confirm the importance of ownership preferences, growth opportunities, liquidity constraints, and difficulties in firm valuation in determining the appropriate firm organization. 28 References Bradley, Michael, Anand Desai, and E. Han Kim, 1988, Synergistic gains from corporate acquisitions and their division between the stockholders of target and acquiring firms, Journal of Financial Economics 21, 3-40. Brau, James, Bill Francis and Ninon Kohers, 2003, The choice of IPO versus takeover: empirical evidence, Journal of Business 76, 583-612. Camerlynck, Jan and Hubert Ooghe, 2003, Pre-acquisition profile of privately held companies involved in takeover: an empirical study, working paper, University of Ghent. Chemmanur, Thomas and Paolo Fulghieri, 1999, A theory of the going-public decision, The Review of Financial Studies 12, 249-279. Cumming, Douglas and Jeffrey MacIntosh, 2003, Venture capital exits in Canada and the United States, University of Toronto Law Journal 53, 101-200. Downes, David, and Robert Heinkel, 1982, Signaling and the valuation of unseasoned new issues, Journal of Finance 37, 1-10. Ellingsen, Tore and Kristian Rydqvist, 1997, The stock market as a screening device and the decision to go public, working paper, Stockholm School of Economics and Norwegian School of Management. Fama, Eugene, and Kenneth French, 1997, Industry costs of equity, Journal of Financial Economics 43, 153-193. Field, Laura and Harold Mulherin, 2003, Newly-public firms as acquisition targets: a comparision with established target firms, working paper, Pennsylvania State University. Fuller, Kathleen, Jeffry Netter, and Mike Stegemoller, 2002, What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions, Journal of Finance 57, 1763-1793. Gaver, J. and K. Gaver, 1993, Additional Evidence on the Association Between the Investment Opportunity Set and Corporate Financing, Dividend, and Compensation Policies, Journal of Accounting and Economics 16, 1993, 125-160. Gompers, Paul, 1995, Optimal investment, monitoring, and the staging of venture capital, Journal of Finance 50, 1461-1489. Grimes, Ann, 2004, Many start-ups opt for ‘dual tracking’ IPO or outright sale, Wall Street Journal July 23, http://online.wsj.com/article_print/0,,SB109054612006071960,00.html Jain, Bharat and Omesh Kini, 1994, The post-issue operating performance of IPO firms, Journal of Finance 49, 1699-1726. Kensinger, John, John D. Martin, and J. William Petty, 2000, Harvesting value from entrepreneurial success, Journal of Applied Corporate Finance 12, 81-93. 29 Kim, Moonchul and Jay Ritter, 1999, Valuing IPOs, Journal of Financial Economics 53, 409737. Koeplin, John, Atulya Sarin, and Alan Shapiro, 2000, The private company discount, Journal of Applied Corporate Finance 12, 94-101. Lerner, Joshua, 1993, Venture capitalists and the decision to go public, Journal of Financial Economics 35, 293-316. Lowry, Michelle, 2003, Why does IPO volume fluctuate so much?, Journal of Financial Economics 1, 3-40. Maremont, Mark, 2002, Tyco made $8 billion of acquisitions over 3 years but didn’t disclose them, Wall Street Journal February 4, A3. Matsusaka, John, 1993, Target profits and managerial discipline during the conglomerate merger wave, Journal of Industrial Economics 41, 179-189. Mikkelson, Wayne, Megan Partch and K. Shah, 1997, Ownership and operating performance of companies that go public, Journal of Financial Economics 44, 281-307. Myers, Stewart, 1984, The capital structure puzzle, Journal of Finance 39, 575-592. Mulherin, J. Harold, and Audra Boone, 2000, Comparing acquisitions and divestitures, Journal of Corporate Finance 6, 117-140. Pagano, Marcos, F. Panetta and Luigi Zingales, 1998, Why do companies go public? An empirical analysis, Journal of Finance 53, 27-64. Ritter, Jay, 1984, Signaling and the valuation of unseasoned new issues: a comment, Journal of Finance 39, 1231-1237. Ritter, Jay, 1987, The costs of going public, Journal of Financial Economics 19, 269-281. Sahlman, William, 1990, The structure and governance of venture capital organizations, Journal of Financial Economics 27, 473-524. Smith, Cliff and Ross Watts, 1992, The investment opportunity set and corporate financing, dividend and compensation policies, Journal of Financial Economics 32, 263-292. Stein, Jeremy, 1997, Internal capital markets and the competition for corporate resources, Journal of Finance 52, 111-113. Zingales, Luigi, 1995, Insider ownership and the decision to go public, Review of Economic Studies 62, 425-448. 30 Table 1 Values of Sell-outs and IPOs from 1995 through 1999 The following table describes the size and number of deals in a given year. The total row provides the average deal value and also sums the deal values and number of sell-outs and IPOs. Deal value is defined as the total value of consideration paid by the acquirer, excluding fees and expenses. Market value is defined as the midpoint of the price on the opening day multiplied by the number of shares outstanding. The average, median and total columns are reported in millions of dollars. Sell-out Deal Values IPO Market Values Average Median Total N Average Median Total Panel A: All transactions 1995 1996 1997 1998 1999 Total 31.5 41.2 48.9 46.7 83.1 51.3 10.3 13.7 13.0 16.0 20.0 14.5 1995 1996 1997 1998 1999 Total 163.4 193.1 169.7 183.1 244.3 196.2 112.0 114.5 94.0 116.9 130.7 113.1 19,029 34,027 54,574 48,601 72,210 228,442 603 826 1,116 1,040 869 4,455 120.4 172.2 206.4 289.9 464.9 243.3 N 72.8 94.7 93.6 127.1 250.6 109.8 39,616 83,696 66,677 53,928 164,094 408,010 329 486 323 186 353 1,677 111.1 93.8 134.1 167.3 122.6 20,163 17,459 14,370 25,373 77,365 105 113 64 84 366 Panel B: Sample transactions 4,084 16,411 12,898 16,659 21,744 71,798 25 85 76 91 89 366 192.0 154.5 224.5 302.1 211.4 Table 2 The Goodness-of-fit for the Matching Procedure of IPOs and Sell-outs Descriptive statistics for the matched sample of IPOs and sell-outs is reported in this table. The first column represents the difference in size, as measured by the deal value of sell-outs and the initial market value of IPOs, between the two samples. Column two shows the number of firms that we are able to match using Fama and French (1997) industries. It also reports firms that are matched by two-digit SIC and the remainder of firms that are either matched by one digit SIC or matched without using industry as criteria. The last column displays the difference in the transaction year between IPOs and sell-outs. Difference in size Industry differences Difference in years Mean = $59.8 Median = $7.9 Standard deviation = $174.9 Exact matches (Fama-French) = 344 Mismatched but same 2-digit SIC = 8 Mismatched in different 2-digit SIC = 14 Mean = 0.5 Median = 0 Standard deviation = 0.87 Table 3 Reasons Given for the Transition from Private to Public Ownership The following table represents reasons found in S.E.C. filings for private firms either being acquired by a public company or undertaking an initial public offering. Sell-outs are represented in columns two and three and IPOs are represented in the last two columns. There are more observations for IPOs since the information gathered from prospectuses for IPOs is uniform, while the corresponding information for sell-outs is scarcer and not always mandatory. In addition to the reason reported by the private firm, the frequency with which the reason is reported is recorded in both absolute and percentage terms. Reason Sell-outs (N=78) # % Access to Capital and Growth Acquisitions 3 3.9% Capital Expenditures 0 0.0% General Access to Capital 34 43.6% Growth 33 42.3% R&D 20 25.6% Working Capital 2 2.6% Total 53 68.0% Debt Debt Reduction 5 6.4% Payouts Favorable Tax Consequences 28 35.9% Liquidity 61 78.2% Pay a Distribution 2 2.6% Repurchase Stock 2 2.6% Repurchase Preferred Stock 0 0.0% Total 69 88.5% Marketing and Personnel Issues Related to Personnel 16 20.5% Marketing Activities 32 41.0% Total 39 50.0% Other Create Value 11 14.1% Efficiencies / Scale / Synergies 43 55.1% Industry Conditions / Competition 18 23.1% Reorganization & Alliances 0 0.0% Risk Reduction 27 34.6% Timing 3 3.9% IPOs (N=366) # % 130 107 204 54 67 217 294 35.5% 29.2% 55.7% 14.8% 18.3% 59.3% 80.3% 202 55.2% 0 0 35 7 27 69 0.0% 0.0% 9.6% 1.9% 7.4% 18.9% 5 17 17 1.4% 4.6% 4.6% 0 0 0 4 0 0 0.0% 0.0% 0.0% 1.1% 0.0% 0.0% Table 4 Descriptive Statistics for Sell-out and IPO Firms for Transactions from 1995 to 1999 The following table provides the median observation for sell-outs and IPOs from the sample. Wilcoxon test statistics for differences in distribution are reported in the description rows and are represented by asterisks. The p value for the mean difference in variables equal to zero is in the last column. Dollars are recorded in millions. Sell-out IPO p value Revenues prior to transaction*** $47.7 366 $26.9 366 .0066 Total assets prior to transaction*** $35.0 366 $23.0 366 .0381 Retained earnings*** $1.3 337 -$2.7 364 .0061 Operating income prior to transaction*** $3.9 366 $0.8 366 .0180 Undistributed free cash flow*** $6.1 365 $1.7 366 .0109 Net income*** $1.6 366 $0.1 366 .0579 Capital expenditures $2.0 365 $1.7 366 .9544 $113.1 366 $122.6 366 .1155 Years of operating history*** 10 268 8 355 .0006 Number of employees*** 345 112 218 366 .1055 47 36 47 364 .6113 Insider ownership prior to transaction*** 62.7% 73 71.8% 364 .1014 Insider ownership after the transaction*** 0.0% 155 50.8% 364 .0001 51.7% 47 17.3% 364 .0001 2.1% 186 2.7% 366 .1353 Dollar value of transaction Age of CEO Change in insider ownership*** Transaction expenses as % of deal value*** *** ** * significantly different at 1% level significantly different at 5% level significantly different at 10% level Table 5 Growth variables using Accounting Numbers in Sell-outs and IPOs from 1995 to 1999 This table presents the median growth rates for sales, total assets, and capital expenditures. Three years of pretransaction data is available for income and cash flow statement variables. Balance sheet data is available for the two years prior to the transaction. Year 0 represents the year in which the transaction occurred. Panel A shows growth rates in the year before, the year prior to the year before, and the two years prior to the transaction. Wilcoxon test statistics for differences in distribution are reported in the sell-out rows, and are represented by asterisks. Panel B partitions the growth rates by whether the firm received venture backing or not. Growth rates in this panel are represented by the change in sales, assets and capital expenditures from year -2 to year -1. The last column represents the Wilcoxon twosample test probabilities. The number of observations appears below the median growth rate for both panels. Years -2 to -1 Panel A Years -3 to -2 Sales Years -3 to -1 Sell-outs 22.1%*** 294 19.9%** 202 43.0%*** 202 IPOs 39.0% 331 32.2% 279 89.4% 279 Sell-outs 14.1%*** 300 - - IPOs 27.0% 337 - - Sell-outs 16.3%*** 294 29.6% 187 43.6%*** 187 IPOs 56.5% 342 56.8% 288 156% 288 Total assets Capital expenditures Panel B Non-VC VC backed backed Sales Pr > [Z] Sell-outs 28.5% 106 19.4% 188 .087* IPOs 66.7% 165 28.2% 166 .000*** Sell-outs 17.5% 111 16.1% 189 .862 IPOs 53.5% 169 30.6% 168 .078* Sell-outs 23.4% 107 16.1% 187 .654 IPOs 90.6% 171 42.3% 171 .011** Total assets Capital expenditures *** ** * significantly different at 1% level significantly different at 5% level significantly different at 10% level Table 6 Valuation Multiples of IPOs and Sell-outs from 1995 to 1999 This table presents valuation multiples for IPO and sell-out firms. For every multiple the numerator is market value for an IPO and deal value for a sell-out. The market value / book value of assets uses the value of assets from the balance sheet. The sales multiple is value scaled by total revenues. In Panel B, we distinguish by whether the transitioning firm was backed by venture capitalists or not. Wilcoxon test statistics for differences in distribution are reported in the sellout rows, and are represented by asterisks. Panel B reports the differences in the multiples of VC and non-VC firms in the second set of asterisks for sell-outs and also in the IPO column. Sell-outs Panel A: All firms MV / book value of assets MV / sales 2.9*** 2.4*** IPOs 6.1 4.2 Panel B: By venture backing VC backed MV / book value of assets MV / sales 5.1*** 4.7***,b 9.2a 6.6a Non-VC backed MV / book value of assets MV / sales *** ** * a b c Sell-out/IPO different at 1% level Sell-out/IPO different at 5% level Sell-out/IPO different at 10% level VC/non-VC different at 1% level VC/non-VC different at 5% level VC/non-VC different at 10% level 2.7*** 1.8*** 4.1 2.8 Table 7 Accounting Ratios Measuring the Capital Structure of IPOs and Sell-outs from 1995 to 1999 The following table presents median accounting ratios for sell-outs and IPOs. Panel A presents ratios measuring the debt structure of the firm. Firm liquidity is measured in panel B. Panel C shows the degree to which cash flows are used to pay off debt. EBITDA is measured as earnings before taxes plus amortization and depreciation; interest is interest expense reported in the income statement; cash is cash on the balance sheet. These ratios are measured for the two years prior to IPO or sell-out. Wilcoxon test statistics for differences in distribution are reported in the sell-out rows, and are represented by asterisks. Year –1 Panel A: Debt structure Sell-outs Year -2 Long-term debt / total debt 24.0% 366 32.7%* 300 Total debt / total assets 71.0% 366 70.5% 300 Interest / total debt 3.3% 366 2.7% 296 Long-term debt / total debt 24.2% 365 23.3% 336 Total debt / total assets 71.1% 366 69.6% 337 Interest / total debt 2.5% 366 2.8% 335 IPOs Panel B: Liquidity Sell-outs Cash / total assets 6.4%** 366 5.9%** 300 Current assets / current liabilities 1.35 366 1.55 300 Cash / total assets 8.6% 366 9.0% 337 Current assets / current liabilities 1.39 366 1.44 337 IPOs Panel C: Financial distress Sell-outs % of firms where interest > EBITDA 20.3%*** 365 17.8%*** 297 % of firms where interest > cash 35.3%* 365 34.3% 297 % of firms where interest > EBITDA 32.5% 366 33.0% 342 % of firms where interest > cash 28.7% 366 30.4% 342 IPOs *** ** * significantly different at 1% level significantly different at 5% level significantly different at 10% level Table 8 Median Operating Performance Prior to IPO or Sell-out The following table reports the median income of sell-outs and IPOs scaled by sales and total assets. Year 0 represents the year in which the transaction occurred. Raw returns are simply the median operating return. Industry adjusted returns are calculated by subtracting the median operating return of the industry (two digit SIC), taken from public companies, from the median return of the sample firm. Wilcoxon test statistics for differences in distribution are reported in the sell-out rows, and are represented by asterisks. Year -1 Year -2 Sell-outs Operating income / sales ROS Industry-adjusted ROS N 5.5% -2.8%***,a 366 Year -3 4.9% -3.4%***,a 294 6.1% -3.5%***,c 202 Operating income / total assets ROA Industry-adjusted ROA N 8.5% -1.4%***,b 366 7.7% -1.3%***,a 296 IPOs Operating income / sales ROS Industry-adjusted ROS N 2.9% -5.7%a 352 1.9% -7.7%a 331 2.5% -7.1%b 279 Operating income / total assets ROA Industry-adjusted ROA N *** ** * 2.9% -7.0%a 366 significantly different at 1% level significantly different at 5% level significantly different at 10% level 2.8% -7.5%a 336 a significant at 1% level significant at 5% level significant at 10% level b c Table 9 Asymmetric Information and Agency Costs using Accounting Ratios in Sell-outs and IPOs from 1995 to 1999 This table presents research and development scaled by sales and assets and intangible asset scaled by total assets. Medians are reported with the 75th percentile in brackets. Year 0 represents the year in which the transaction occurred. Wilcoxon test statistics for differences in distribution are reported in the sell-out rows and are represented by asterisks. Year –1 Year -2 Sell-outs Year -3 R&D / sales 0% [6.3%]*** 366 0% [8.4%]*** 294 0% [2.4%] 202 R&D / total assets 0% [9.5%]*** 366 0% [12.8%]*** 296 - Intangibles / total assets 0% [5.0%] 366 0% [5.5%] 299 - R&D / sales 0% [16.5%] 352 0% [19.0%] 331 0% [13.6%] 279 R&D / total assets 0% [23.1%] 366 0% [22.2%] 337 - Intangibles / total assets 0% [7.7%] 365 0% [1.9%] 337 - IPOs *** ** * significantly different at 1% level significantly different at 5% level significantly different at 10% level Table 10 Correlations Across Regressors This table displays the correlation coefficient matrix for regressors included in Table 11. All of the variables are measured in year -1. Scaled capital expenditures is capital expenditures scaled by total assets. Venture backed is a dummy variable where 1 denotes the presence of venture financing. Leverage is total debt scaled by total assets. Interest expense > EBITDA is a dummy variable equal to one if the level of interest payments exceeds that of operating income plus depreciation and amortization. Interest expense > Cash expense is a dummy variable equal to one if the level of interest payments exceeds that of cash reported on the balance sheet. Financial distress is an interaction term of the dummy variables Interest expense > EBITDA and Interest expense > Cash. Abnormal return on assets is earnings before interest and taxes (EBIT) divided by total assets net of the same return for the median publicly traded firm in the same two-digit SIC. Scaled research and development is research and development scaled by total assets. Log of sales is the natural log of revenues. Scaled capex Venture backed Leverage Interest expense > EBITDA Interest expense > Cash Financial distress Abnormal return on assets Scaled R&D Scaled capital expenditures 1.00 Venture backed 0.11 1.00 Leverage 0.14 -0.02 1.00 Interest expense > EBITDA 0.16 0.29 0.17 1.00 Interest expense > Cash 0.03 -0.09 0.21 -0.12 1.00 Financial distress 0.02 0.05 0.29 0.42 0.37 1.00 Abnormal return on assets -0.17 -0.29 -0.37 -0.46 0.11 -0.23 1.00 Scaled R&D 0.10 0.29 0.26 0.32 -0.16 0.12 -0.75 1.00 Log of sales -0.14 -0.24 -0.06 -0.38 0.28 -0.06 0.57 -0.54 Log of sales 1.00 Table 11 Logistic Regression Analysis of Factors Influencing Transition Choice for Private Firms Being Acquired or Going Public from 1995 to 1999 The dependent variable is a dummy variable taking the value 0 if the private firm was acquired by a publicly traded company and 1 if the company conducted an IPO. All of the independent variables are measured in year -1. Scaled capital expenditures is capital expenditures scaled by total assets. Venture backed is a dummy variable where 1 denotes the presence of venture financing. Leverage is total debt scaled by total assets. Interest expense > EBITDA is a dummy variable equal to one if the level of interest payments exceeds that of operating income plus depreciation and amortization. Interest expense > Cash expense is a dummy variable equal to one if the level of interest payments exceeds that of cash reported on the balance sheet. Financial distress is an interaction term of the dummy variables Interest expense > EBITDA and Interest expense > Cash. Abnormal return on assets is earnings before interest and taxes (EBIT) divided by total assets net of the same return for the median publicly traded firm in the same two-digit SIC. Scaled research and development is research and development scaled by total assets. Log of sales is the natural log of revenues. P-values are reported in brackets next to the parameter estimates. Variables (1) (2) (3) (4) Intercept 1.10 [.030] 0.90 [.084] 0.80 [.110] 0.80 [.116] Scaled capital expenditures 1.50 [.008] 1.36 [.017] 1.37 [.015] 1.32 [.020] Venture backed 0.59 [.001] 0.53 [.002] 0.52 [.002] 0.52 [.002] Leverage 0.26 [.075] 0.29 [.047] 0.28 [.050] 0.24 [.080] - 0.50 [.030] 0.52 [.025] 0.45 [.047] -0.20 [.250] -0.02 [.930] -0.01 [.970] -0.02 [.925] - -0.89 [.032] -0.89 [.033] -0.91 [.028] 0.15 [.386] 0.20 [.269] 0.28 [.061] - Scaled R&D -0.34 [.387] -0.30 [.452] - -0.55 [.089] Log of sales -0.16 [.008] -0.15 [.002] -0.14 [.003] -0.13 [.003] 47.6 [.000] 53.7 [.000] 53.6 [.000] 52.9 [.000] 365/366 365/366 365/366 365/366 Interest expense > EBITDA Interest expense > Cash Financial distress Abnormal return on assets Likelihood ratio test statistic Number of observations Table 12 Logistic Regression Analysis of Factors Influencing Transition Choice for Private Firms Being Acquired or Going Public from 1995 to 1999 The dependent variable is a dummy variable taking the value 0 if the private firm was acquired by a publicly traded company and 1 if the company conducted an IPO. All of the independent variables are measured in year -1. Scaled capital expenditures is capital expenditures scaled by total assets. Venture backed is a dummy variable where 1 denotes the presence of venture financing. Leverage is total debt scaled by total assets. Interest expense > EBITDA is a dummy variable equal to one if the level of interest payments exceeds that of operating income plus depreciation and amortization. Interest expense > Cash expense is a dummy variable equal to one if the level of interest payments exceeds that of cash reported on the balance sheet. Financial distress is an interaction term of the dummy variables Interest expense > EBITDA and Interest expense > Cash. Abnormal return on assets is earnings before interest and taxes (EBIT) divided by total assets net of the same return for the median publicly traded firm in the same two-digit SIC. Scaled research and development is research and development scaled by total assets. Log of sales is the natural log of revenues. Sales growth is the difference in year -1 sales and year -2 sales scaled by year -2 sales. Asset growth is the difference in year -1 total assets and year -2 total assets scaled by year -2 total assets. Capital expenditures growth is the difference in year -1 capital expenditures and year -2 capital expenditures scaled by year -2 capital expenditures. Insider ownership after is the amount of ownership firm insiders have in the new firm after the sell-out or IPO. P-values are reported in brackets next to the parameter estimates. Variables Intercept (1) (2) (3) (4) 0.79 [.162] 0.42 [.477] 0.98 [.092] -3.30 [.002] - - - 2.31 [.104] Venture backed 0.41 [.020] 0.38 [.035] 0.43 [.018] -0.13 [.746] Leverage 0.43 [.009] 0.45 [.007] 0.41 [.012] 0.42 [.335] Interest expense > EBITDA 0.34 [.171] 0.34 [.181] 0.29 [.250] 1.10 [.061] Interest expense > Cash 0.02 [.935] 0.01 [.949] 0.04 [.865] 0.68 [.142] -0.94 [.031] -0.98 [.028] -0.92 [.036] -1.68 [.126] 0.16 [.313] 0.16 [.345] 0.16 [.325] -0.24 [.569] Log of sales -0.12 [.024] -0.09 [.083] -0.14 [.011] 0.02 [.851] Sales growth 0.03 [.593] - - - Asset growth - 0.18 [.014] - - Capital expenditures growth - - 0.04 [.073] - Inside ownership after - - - 14.77 [.000] 31.6 [.000] 39.3 [.000] 37.3 [.000] 426.7 [.000] 298/342 299/337 294/342 154/364 Scaled capital expenditures Financial distress Abnormal return on assets Likelihood ratio test statistic Number of observations