Corporations and the financing of innovation: the corporate venturing experience
Paul A. Gompers
Corporate internal investments in innovative activities, including research and development, have often been maligned for their ineffectiveness (Jensen 1993). Over the past forty years, corporations have attempted to capture the value from waves of technology and innovation. But during much of this time, corporations saw young, nimble start-ups capitalize on opportunities that the corporations saw first.
Why have corporations had difficulty bringing innovations to market? Many of the best ideas have languished, unused, whether because of internal resistance (for example, from managers who did not want to see a product launched that competed with one of their offerings) or an inability to execute on the initial insight. In other cases, defecting employees started new firms that turned those ideas into blockbuster commercial successes. The achievements of fast-growing technology firms such as Microsoft and Cisco Systems–many of which relied on acquisitions rather than on internal research and development (R&D) for the bulk of their new ideas–also made conventional approaches to innovation look lackluster by comparison. In response to these factors, many corporations entered the venture capital market in hopes of spurring their own innovative capacity.
Corporations have good reason to explore new ways of stimulating innovation. All too often, their investments in traditional R&D laboratories have generated paltry returns as researchers have focused on incremental product advances or on academic ideas with little relevance to the corporation (Henderson 1993; Henderson and Cockburn 1996). Worse, even when these corporate laboratories managed to come up with truly innovative ideas, other organizations–especially venture-backed start-ups–have sometimes seized the opportunity to commercialize them.
But how can companies best stimulate innovation in a corporate setting? The venture capital industry’s success may be difficult to replicate. Though total disbursements from the venture industry during 1975-2000 proved considerably less than the R&D spending of either IBM or General Motors alone, venture-backed firms have scored remarkable successes (Reinganum 1989; Lerner 1997).
This paper explores the history, structure, and performance of corporate venture programs in the United States. It chronicles the cyclical nature of the industry over the past forty years, a time during which corporate venture capital programs were often halted before the full fruits from the investment activity could be realized. This study shows that the corporate venture capital market in the United States has gone through three waves of activity that track the overall independent venture capital market.
The paper next explores the experience of corporate venture investment using a detailed microlevel data set. The analysis finds that such investments are increasingly made in related industries–that is, over time, the strategic fit between corporate venture capital investments and the parent corporation’s business has increased. In addition, contrary to previous assumptions, corporate venture capital investments have, on average, been more successful than independent venture capital investments. This success is exclusively associated with strategic corporate venture investments–that is, nonrelated investments have much lower success rates. This study concludes that corporations appear to be learning many of the best practices from the independent venture capital sector. The success of corporate venture investing has increased over time.
Types of Corporate Venturing
Large corporations have long been attracted to venture capital investing. Many of these efforts have been motivated by a desire to gain access to cutting-edge technologies for strategic reasons. Sometimes these strategic goals far outweighed any consideration of financial return for corporate investors, allowing financial investors to treat these corporations as later-stage, valuation-insensitive investors. This behavior led many independent venture capitalists to introduce early-stage technology companies to corporate investors only during later rounds of financing, when portfolio companies required large amounts of cash raised at extremely high valuations to preserve the venture capitalists’ percentage ownership. This practice created situations in which corporations invested in companies that were often significantly overvalued and made it difficult for corporate investors to achieve acceptable financial returns. As a result, many corporate investors reached the conclusion that it was not possible to achieve both financial and strategic goals in doing early-stage technology investing.
Corporations used several models to achieve their strategic and financial objectives for venture capital investments. Each of these models, however, created problems that ultimately caused corporations to fail to reach their goals.
Internal corporate venture group. Some corporations created internal corporate venture groups to analyze venture capital opportunities and make investments. Problems typically arose with this strategy because it limited deal flow to those companies that wanted to be associated with that particular corporation. Entrepreneurs were limited by this structure because, while they could receive excellent depth of assistance in the corporation’s area of expertise, they were forced to sacrifice breadth of available resources. In addition, early-stage entrepreneurs were often concerned about protecting their intellectual property and wanted to avoid alliances that could threaten their position. For example, a small high-technology company in a precarious financial situation might be reluctant to approach IBM or Sony directly for funding. Therefore, the very companies in which these corporations wanted to invest were usually the ones that never made it to their doorsteps.
Dedicated external fund. Other corporations placed investment capital in a dedicated fund that existed as a separate entity outside the corporation. This structure did not solve many entrepreneurs’ concerns because they still needed to feel comfortable with forming an alliance with the particular corporation sponsoring the fund. Since corporations were able to use only dedicated external funds to attract entrepreneurs that wanted to be aligned with them, the corporations were not able to allocate assets across industry areas besides their own even though diversification through pooled investments might have produced better risk management and higher financial returns. In addition, a dedicated external fund often frustrated a corporation’s desire to gain strategic leverage with start-up companies. The corporation’s relationship was too distant for the corporation’s employees to work closely with the entrepreneurs.
Passive limited partner in a venture fund. Existing venture funds gave corporations the opportunity to become passive limited partners and make diversified investments in entrepreneurial companies. (1) The venture capitalists managing these funds typically had little incentive to involve corporations in early investments. Instead, the venture capitalists would send corporate limited partners deals at later stages for passive investment at fairly high valuations. (2) In addition, this structure did not allow corporations to achieve strategic objectives since the corporations, as passive investors, did not have direct relationships with the entrepreneurs. The “information flow” to corporations depended on the venture capitalists’ goodwill.
The History of Corporate Venturing Investments
The first corporate venture funds emerged in the mid-1960s–about two decades after the initial institutional venture capital funds formed. (3) Since that time, corporate venturing has undergone three boom-and-bust cycles that closely track the independent venture capital sector. Corporations have typically entered the corporate venture capital market after the independent sector showed signs of success (Gompers and Lerner 1998a). All too often, however, corporations overbuilt capacity without carefully thinking out the implications (Block and Ornati 1987). This strategy invariably led to retrenchment.
The first wave, As traditional venture capital funds fueled the success of corporations such as Digital Equipment Corporation, Memorex, Raychem, and Scientific Data Systems, large companies took notice, reviewing these successes as new potential opportunities. Large companies began establishing divisions that emulated venture capitalists. During the late 1960s and early 1970s, more than 25 percent of the Fortune 500 firms set up such programs (Rind 1981).
At one end of the spectrum, large corporations financed new firms that were already receiving venture capital from independent venture capital organizations. Most of these efforts, such as General Electric’s Business Development Services, Inc., invested directly in start-ups. This strategy let managers tailor their firms’ portfolios to their particular technological or business needs. In other cases, the corporations simply provided funds to a separate venture capital firm. This separate firm would in turn invest the money in entrepreneurial organizations.
At the other end of the spectrum, projects such as DuPont Corporation’s Development Department and Ralston Purina’s New Venture Division sought to promote new ventures internally. These programs encouraged the companies’ own product engineers and scientists to forge ahead with their innovations–and provided financial, legal, and marketing support. In some cases, these units were separate legal entities, which at times also had outside equity investors. More typically, however, the corporate parent retained ownership of the program.
In 1973, the market for new public offerings–the primary avenue through which venture capitalists exit successful investments–dried up as small technology stocks experienced very poor returns. Returns of independent venture funds shrank, and commitments to the independent venture capital sector fell. Corporations, in light of the declining market, began scaling back their own venturing initiatives. The typical corporate venture program begun in the late 1960s was dissolved after just four years.
The second wave. The independent venture industry’s prospects brightened again in the late 1970s and early 1980s. Two regulatory changes had a dramatic impact on venture capital commitments (Gompers and Lerner 1998c). First, the top capital gains tax rate was reduced in 1978. Second, the Department of Labor eased pension investment restrictions in 1979, allowing pension managers to invest substantial amounts into venture capital funds. In addition, several new technological innovations, including personal computer hardware and software, provided an opportunity for new companies to exploit new markets. The flow of funding into the venture capital industry grew, and the number of active venture organizations proliferated.
Corporate venturing increased shortly thereafter. By 1986 corporate funds managed $2 billion, or nearly 12 percent of the total pool of venture capital. Whereas the earlier wave of corporate venturing had taken aim at a broad range of investment opportunities, now high-tech and pharmaceutical companies–such as Control Data, EG&G, Eli Lilly, and Monsanto–led the charge.
The boom of the early 1980s, however, was soon followed by another retrenchment. In 1987, the stock market crashed and the market for new public offerings again deflated. As in the past, returns and fund-raising by independent partnerships shrank as well. Chart 1 provides a profile of this relationship. This time, corporations scaled back their commitment to venture investing even more dramatically. By 1992 the number of corporate venture programs had fallen by one-third, and their capital under management represented only 5 percent of a much smaller venture pool.
[GRAPHIC OMITTED]
The third wave. The venture capital industry expanded once again in the late 1990s, fueled in large part by the highly visible successes of telecommunications and Internet-related companies. As rates of return on venture capital investments rose, corporations once again became attracted to the opportunity of corporate venturing. Chart 2 graphs the pace of venture capital fund-raising through 2001 and median rates of return on venture capital investments through 2000. The graph shows the dramatic expansion in returns along with the unprecedented rise in fund-raising. During this period, many corporations had decided to reevaluate the innovation process itself. For much of the century, large corporations had typically relied on central R&D laboratories to crank out new product ideas. Now, these organizations began exploring other ways to access new ideas–including joint ventures, acquisitions, and university-based collaborations. Corporate venture programs gave corporations the opportunity to capitalize on these relationships.
[GRAPHIC OMITTED]
The rapid diffusion of the Internet and its power to either enhance or cannibalize “bricks-and-mortar” businesses intensified this interest. Corporations everywhere realized that e-commerce presented both an opportunity and a threat. However, many organizations lacked the internal resources to explore these new opportunities. Corporate venturing provided one solution. For example, the Tribune Company, the Sony Corporation, and United Parcel Service all instituted efforts to invest in on-line businesses.
Finally, numerous venture capital groups, looking for strategic-partnering opportunities, expressed interest in collaborating with corporations. In earlier years, traditional venture investors had approached corporate investors with a mix of caution and skepticism. The waxing and waning of corporate interest–which historically had fluctuated more wildly than cycles in the venture industry had–made many venture capitalists nervous.
But as the venture capital sector grew increasingly crowded in the late 1990s, the venture community adopted a different attitude. Venture capitalists increasingly saw corporate investments as a potential strategic advantage. And a new focus on revolutionary business strategies–such as customer-relationship management–woke up venture groups to their own limitations. A corporate partner, some venture firms surmised, might provide the knowledge and experience that venture organizations needed to improve their own skills and professionalism. Such groups forged partnerships with corporations, not only accepting money from them as investors but also structuring unique collaborations that sought to draw upon the expertise of the large organization.
Corporate Venture Activity
Corporate venture capital activity is difficult to measure, but Chart 3 provides some measure of the level of activity. Chart 3 graphs the number of corporate venture capital programs announced publicly by Fortune 100 companies. The three historical “waves” show up prominently in the graph. The number of programs established during the 1962-98 period totals well above 100. Though not all corporations established venturing programs during these decades, many that did often set up more than one. In addition, a single company might abandon and revive a series of such programs.
[GRAPHIC OMITTED]
Another indicator of the size of the corporate venturing effort can be seen in Table 1, which shows the fifteen largest corporate venture capital programs in 2000 and their capital under management. The table shows that the types of firms engaged in corporate venturing come from a diverse set of industries. Many are high-technology leaders in their fields, such as Intel and Siemens. Others are relatively low-technology or financial companies, including Comdisco, Time Warner, and Visa International.
The overall scope of corporate venture activity over recent years is shown in Table 2, which compiles the number and (in latter years) the size of venture investments made directly by corporations. These numbers do not include cases in which companies committed capital to independent venture groups, which then invested the funds. Nor do they reflect instances in which a financial services organization or a subsidiary of an operating corporation (for instance, Goldman Sachs or GE Capital) made an investment. The table demonstrates the tremendous growth of corporate venturing during the third wave. The number of corporate venture investments increased nearly twenty-fold over sixteen years, and the amount of corporate venture investments that could be tracked amounted to nearly $8 billion in 1999. (4)
Empirical Analysis
Data description. VentureOne, established in 1987, collects data on firms that have obtained venture capital financing. The VentureOne database used in this analysis includes firms that have received early-stage equity financing from venture capital organizations, corporate venture capital programs, and other organizations.
The companies are initially identified from a wide variety of sources, including trade publications, company Web pages, and telephone contacts with venture investors. VentureOne then collects information about the businesses through interviews with venture capitalists and entrepreneurs. Among the data collected are the names of the investors, the amount and valuation of the venture financings, and the industry, history, and current status of the firm. Data on the firms are updated and validated through monthly contacts with investors and firms. (5) VentureOne then markets the database to venture funds and corporate business development groups (see Gompers and Lerner 2000 for a detailed discussion of the database).
For this analysis, the VentureOne data were supplemented when necessary. Some firms in the VentureOne sample were missing information, such as an assignment to one of the 103 VentureOne industry classes or information on the firm’s start date. To determine this information, a variety of reference sources were consulted, including Corporate Technology Information Service’s Corporate Technology Directory (1996), Dun’s Marketing Services’ Million Dollar Directory (1996), Gale Research’s Ward’s Business Directory of U.S. Private and Public Companies (1996), National Register Publishing Company’s Directory of Leading Private Companies (1996), and a considerable number of state and industry business directories in the collections of Harvard Business School’s Baker Library and the Boston Public Library. Several electronic databases were also employed: the Company Intelligence and Database America compilations available through LEXIS’s COMPANY/USPRIV library and the American Business Disk CD-ROM directory.
The investors in the VentureOne database were diverse. They included individuals, institutional investors such as pension funds, traditional independent venture funds (such as Kleiner, Perkins, Caufield & Byers), and funds sponsored by corporations, financial institutions, and government bodies. In order to understand the impact of organizational structure, many of the analyses below concentrate on two types of funds: independent venture partnerships and corporate funds. As discussed above, other hybrid venture funds, such as those affiliated with commercial and investment banks, were eliminated because many of these closely resembled traditional venture organizations.
To identify independent and corporate venture capital organizations, the analysis used an unpublished database of venture organizations assembled by Venture Economics’ Investors Services Group. Venture Economics is a unit of Securities Data Company and tracks the venture capital industry. The organization was known as Capital Publishing when it was established in 1961 to prepare a newsletter on federally chartered small business investment companies (SBICs). Since 1977 the company has maintained a database on venture partnerships, which includes over 2,000 venture capital funds, SBICs, and related organizations. The Investors Services Group database is used in preparation of directories, such as the Venture Economics annual volume Venture Capital Performance. The database is compiled from information provided by venture capitalists and institutional investors. This analysis excluded from either classification a variety of other organizations that make private equity investments, including individual investors, SBICs, funds sponsored by banks and other financial institutions, and funds associated with financial subsidiaries of nonfinancial corporations (such as General Electric Capital). To determine whether a company was a nonfinancial corporation, the firm directories noted above were consulted to determine the main lines of business in the year of the investment and thus draw as sharp a contrast as possible between corporate and independent funds.
In some cases, it was difficult to ascertain if an investor was a corporate venture organization. Some U.S. and several European companies invest in companies through traditional venture capital partnerships. For example, Eastman Kodak not only makes direct equity investments but also invests through a partnership called Aperture Partners, in which it is the sole limited partner. While many of these cases were identified for this analysis, some affiliations may have been missed. In other cases, independent venture organizations also cater to corporate investors. A prominent example is Advent, a Boston-based company that organizes comingled funds for financial investors and other funds for single corporate limited partners. From the VentureOne database, it is usually difficult to determine whether the private equity group is investing from its traditional partnerships or from one of its corporate funds.
Finally, the corporate venture capital investments were characterized by the degree of fit between the corporation and the portfolio firm. From information in the corporate annual reports for the 1983, 1989, and 1994 fiscal years, investments were classified as to whether there was a direct fit between one of the corporation’s lines of business during the period and the portfolio firm, whether there was an indirect relationship, or whether there was no apparent relationship. In the analyses below, investments are denoted as having a strategic fit only if there was a direct relationship between a line of business of the corporate parent and the portfolio firm. The results are robust to expanding the definition to include indirectly related transactions as well: for example, when a corporate fund invests in a firm that is a potential supplier to or customer of the corporate parent. Not all investments were classified. In some cases, the relationship could not be determined. In others, only the proximate annual reports could be obtained; in particular, it was difficult to obtain the 1983 and 1989 annual reports for many of the foreign firms.
The analysis was limited to investments in privately held firms between 1983 and 1994. While VentureOne has sought to “back-fill” its database with information on earlier venture investments, its coverage of the 1970s and early 1980s is poor. Furthermore, there was concern that VentureOne’s methodology may have introduced selection biases. While the database does not include all venture investments between 1983 and 1994, it provides a reasonable view of the activity in the industry during this period. (6) Investments made after 1994 were not included because I wish to assess the outcomes of the investments: it may take several years until the fate of venture-backed firms is clear. I also eliminated a variety of investments outside the scope of this analysis, such as purchases of shares of publicly traded firms and other financings.
Summary statistics. After presenting an overview of the sample, I undertake empirical analyses of the ultimate success of corporate and other venture investments.
Table 3 provides an overview of the sample by year. After the deletions noted above, the sample consists of 32,364 investments. Investments by independent venture funds represent over one-half of the total transactions in the sample. Corporate venture investments represent a much smaller share, about 6 percent. Chart 4 presents the fraction of investments that are corporate venture capital investments by year. Because on average about four investors participate in each financing round, the number of rounds, 8,506, is significantly smaller. In the analyses below, patterns are analyzed on both the investment and round level. (7)
Table 4 provides a comparison of four categories of investments: the total sample, investments by corporate and independent venture capital organizations, and corporate investments in which there was a strategic fit between the parent and the portfolio firm. In general, the corporate investments closely resemble those of the other funds:
* Status at the time of investment. Corporate funds tend to invest slightly less frequently in start-up and mature private firms. Instead, they are disproportionately represented among companies in the middle stages, such as “development” or “beta.” (8)
* Location of the firm. The sample disproportionately includes investments in firms based in California. This idiosyncrasy reflects VentureOne’s greater coverage of this region, particularly in the early years (see Gompers and Lerner 2000 for a discussion). While corporate venture investments as a whole are slightly more common in California than other venture investments, corporate invest-ments with a strong strategic fit are more frequent elsewhere.
* Industry of the firm. Venture capital investments tend to focus on a few high-technology industries. This pattern is even stronger for corporate venture investments with a strategic focus.
* Maturity of the firm and investment characteristics. Corporate venture funds tend to invest in later and larger financing rounds and in slightly older firms than other venture funds do.
Trends and determinants of investment relatedness. This section explores the trends and determinants of whether corporate venture capital investments are made in related industries or not. As the previous discussion made clear, many corporate venture capital efforts have failed when they made investments in companies in totally unrelated markets. It is often believed that large existing players in an existing market can add value to new entrants. Understanding when and how corporate venture groups choose to invest in related companies is critical to determining whether corporate investments can add value.
Chart 5 shows the fraction of the corporate venture capital investments that are made in related industries. One surprising observation is that a large fraction of investment is in related industries. In each year of investment, at least 68 percent of investments made by corporate venture capital groups are in companies in a related industry. It also appears that the fraction of the investments made in related industries has increased over the sample period. By the end of the sample, between 76 percent and 77 percent of the investments were being made in related industries.
[GRAPHIC OMITTED]
Table 5 undertakes a regression analysis to understand the determinants of investment relatedness. The dependent variable is a dummy that equals one if the corporate investment is in a related industry. The independent variables include the age of the firm, a time trend to understand whether the rate of related investments have increased over time, the stage of development of the company, and a dummy variable that equals one if the company is headquartered in Massachusetts or California.
Not surprisingly, the probability of a corporate investment being made in a related industry increases over time. In fact, each year the probability of a related investment increases by 3.5 percent. It therefore appears that corporations were learning about the value of related investments during the decade.
It also appears that investments in early-stage companies are more likely to be in a related industry than investments in later-stage companies. Firms in the development stage or the beta stage are significantly more likely to be in related industries. This tendency is also encouraging. Existing players in an industry can provide significant value to young, entrepreneurial firms. Large corporations are also likely to get the most value from investing in the younger start-ups.
Finally, there are interesting geographical differences in the rate of related investments. Corporate venture investments in Massachusetts are far more likely to be in related industries than are investments in California or in the rest of the country. The rate of related investments in California is no different than in the rest of the country. Perhaps the types of venture capital firms in Massachusetts create an environment that is more accepting of corporate investments by industry leaders.
Success of venture investments. Even though these complex motives–and benefits–make it hard to compare the success of corporate versus independent venturing, a pattern does emerge if one examines the data. In fact, in making a comparison, let’s look only at corporate venture investments made between 1983 and 1994 to ensure that those efforts had time to ripen. (9)
The status of the firms in the spring of 1998 is determined from the VentureOne database. Table 6 presents the outcomes for four classes of investors as well as tests of the statistical significance of the differences between them. Firms backed by corporate venture groups are significantly more likely to have gone public than those financed by other organizations and are less likely to have been liquidated. These differences are particularly strong for those investments in which there was a strategic tie between the corporate parent and the portfolio firm. These comparisons may be influenced, however, by differences between the firms backed by corporate and other venture investors.
The evidence is striking: In more than 30,000 investments into entrepreneurial firms by venture capital organizations of all types, corporate efforts appear to be at least as successful as those backed by independent venture organizations (using such criteria as the probability of a portfolio firm’s going public). As Table 6 shows, 35 percent of the investments by corporate funds went to companies that had gone public by the end of the sample period as opposed to 31 percent for independent funds. The differences persist even when different criteria for success are used: for instance, firms that went public or were acquired at a valuation that was at least three times that of the original investment.
It might be thought that these results are just consequences of the fact that corporate groups often invest in later financing rounds. By this point in many investee firms’ development, uncertainties have cleared up and prospects have brightened. As it turns out, the same results ensued even when controls for a portfolio firm’s age and profitability at the time of the original investment are added.
The success is not uniform, as the final column of the table reveals. The success of a venturing effort varies with the “tightness” of fit between the corporation and the portfolio firm–that is, whether the corporate parent and the investee are in the same line of business. To assess this fit, one can examine corporate annual reports and classified investments. The success of a corporate program depends on the presence of a direct, strategic overlap between corporate parent and investee. As just one illustration, the probability of going public by the end of the sample period is 39 percent for companies that had this kind of alignment compared with much lower percentages for nonaligned firms.
To address this concern, I examine these patterns in a regression framework in Table 7. I estimate logit regressions, alternatively using each investment and each financing round as observations. I seek to explain the probability that the investment had gone public by the spring of 1998 or the probability that the firm had gone public, filed a registration with the U.S. Securities and Exchange Commission (a preliminary step before going public), or been acquired for a valuation of at least twice the postmoney valuation of the financing. (10) As independent variables, I use the age of the firm at the time of the investment and the ordinal rank of the investment round. I also employ dummy variables denoting investments by corporate and independent venture capital funds, corporate venture investments in which there was a strategic fit with the portfolio finn, firms based in California and Massachusetts, the status of the firm at the time of the investment, the year of the investment, the industry of the firm, and a constant.
The results are consistent with the univariate comparisons above. Corporate venture investments are significantly more successful than other investments. (In most of the regressions, independent venture investments are also more successful though the effect is smaller in magnitude and statistical significance.) When the dummy variable denoting corporate venture investments with a strategic fit is added to the regressions, the corporate venture dummy variable becomes insignificant (and frequently negative). Corporate venture investments in general do not perform better–only those with a strategic fit. These results seem consistent with the complementarities hypothesis above.
A Clinical Look at the Corporate Venture Evidence
In addition to strategic fit, market knowledge, and resources, the way a corporation approaches its venture program influences its chances of success (Siegel, Siegel, and MacMillan 1988; Sykes 1990). In companies whose venture programs do not succeed, managers have made two fatal mistakes:
* They never created consensus inside the organization about the program’s objectives and its potential benefits to the company.
* They failed to build relationships and establish credibility outside the corporation. (In many instances, they assumed that the corporation’s name alone would ensure success.)
Solidifying internal cohesion. Many corporations plunge into corporate venturing without realizing that how they design the program matters. As a result, conflict can arise over the program’s objectives–and can even force the dissolution of the effort. For instance, as discussed earlier, departments that feel threatened by or otherwise uncomfortable with the program might push to have it terminated. Or the venture unit’s interests and the corporation’s goals may be unaligned–for example, venture personnel are rewarded solely on financial return whereas the corporation makes strategic goals a priority.
Exxon Enterprises, whose venture capital effort ranks among the most spectacular failures in the field, suffered the consequences of internal dissension. (11) The oil giant (called Esso at the time), seeking to diversify its product line, launched its venture program in 1964. The program began with a mandate to exploit technology in Exxon’s corporate laboratories: for example, making building materials out of petroleum derivatives.
In the late 1960s, however, the fund managers decided to make minority investments in a wide variety of industries, from advanced materials to air-pollution-control equipment to medical devices. In the late 1970s, the strategy changed yet again–the program now focused solely on systems for office use. Finally, in 1985, Exxon abandoned the venture effort entirely. Each shift in corporate strategy had brought on waves of costly write-downs. The information-systems effort alone generated an estimated $2 billion in losses for the corporation.
What explains this disaster? In part, the corporate venture team came to the project with scant investment experience and made numerous poor decisions. But equally important, senior managers at Exxon could not agree on the program’s overarching purpose. Moreover, various divisions at Exxon insisted on detailed reviews of the program. These reviews consumed so much time that they distracted the fund managers’ attention away from the selection and oversight of investments. Meanwhile, various organizations within the corporation had a hand in structuring the program. For instance, Exxon’s human resources staff complained that the venture firms’ compensation schemes did not mirror those of the overall corporation. In the late 1970s, human resources succeeded in replacing the venture staff’s separate stock-option schemes with a standard salary-plus-bonus plan. An exodus of fund managers soon followed.
Internal consensus is particularly important in venture programs with strong strategic objectives. The $100 million Java Fund, launched in 1996 by Kleiner, Perkins, Caufield & Byers, is one example of a fund that gave a number of corporations a chance to invest primarily for strategic reasons. (12) The fund specifically invested in companies that used Java, a programming language developed by Sun Microsystems that runs on a wide variety of operating systems and challenged Microsoft Windows. In addition to raising capital from traditional limited partners (such as the Harvard, Stanford, and Yale University endowments), the fund also tapped firms such as Cisco, IBM, Netscape, Oracle, and, of course, Sun. Even though these firms competed intensely with each other, they all wanted to see this programming language take root because it would level the playing field with their formidable competitor Microsoft.
Cultivating external relationships. Good relationships with independent venture firms are also essential to the success of corporate programs. Why? Particularly today, the venture capital business is highly competitive. Identifying and gaining access to attractive opportunities can be difficult for new players. Meanwhile, investors have to make decisions quickly, often with scant information about an opportunity. Close ties between corporate venture efforts and traditional venture firms can
* bring promising opportunities to the corporate fund’s attention,
* bring early-stage transactions–which often have lower valuations and more strategic potential–to the corporate fund’s attention,
* ensure that venture capitalists deal with corporate capitalists professionally and respectfully, and
* let corporate groups tap into independent groups’ knowledge.
Despite all these potential benefits, relations between corporate and independent venture groups continue to be somewhat strained. The venture capital community is close-knit; many leading firms have syndicated transactions with each other for decades. Though these firms’ skepticism about corporate venture funds has abated somewhat, a residual amount remains. Furthermore, unscrupulous venture groups have been known to exploit naive corporate investors, offering them overpriced investments or withholding bad news about potential investees.
To make relationship building even more difficult, it takes time for corporations to build credibility in the eyes of independent venture capitalists. Many corporations launch venture programs assuming that their names alone will earn them instant respect. They then discover that their venture program is not going anywhere without “road shows” with venture groups, conference presentations, and press releases to publicize the company’s activities.
There are several important lessons to be learned from these accounts:
* Form an appropriately sized fund. Too small a fund suggests a limited commitment by the corporation to the program; too substantial an effort leads to speculation that the corporation does not understand the dangers associated with growing too quickly.
* Recruit one or more of the fund’s investment professionals from the venture capital community.
* Articulate a clear investment strategy.
* Simultaneously invest in venture capital partnerships specializing in similar technologies.
* Consider joint ventures (1) with a specific venture capitalist firm (for instance, Softbank and K-Mart formed a collaboration called BlueLight (13)), (2) with several other corporations and a venture capitalist firm (such as Kleiner Perkins’ Java Fund), and (3) with a number of venture capitalist firms. (For instance, Sutter Hill Ventures, Technology Crossover Ventures, and buyout fund Bain Capital joined in mid-2000 with the consulting firm eLoyalty to establish the eLoyalty Ventures Fund. (14))
Conclusions
This paper has explored the experience of corporations’ investments in young, entrepreneurial firms. Historically, the media and academics have maligned corporate investments in venture capital and highlighted visible failures. This study, however, finds quite a different result. Corporate venture investments have waxed and waned in tandem with the independent venture capital industry. Many of today’s leading technology corporations are extremely active in the sector. In addition, corporate venture capital groups have been increasingly willing to invest in start-ups in related industries. The probability of making investments in related companies increases with early-stage companies. Finally, the paper shows that corporate investments are at least as successful as independent venture capital investments. In addition, the probability of success is substantially higher for corporate venture investments in related industries.
While corporate venture investing suffers from many of the same problems that have affected fads in venture capital investing as a whole, corporate venture investments have a successfull track record. The experience of recent corporate programs, many of which have been initiated by companies that can trace their own history to venture capital investments, bodes well for the future of corporate venturing.
APPENDIX
Definitions of Firm Categorizations
Investment Stages
Start-up: Company with a skeletal business plan, product, or service development in preliminary stages.
Development: Stage at which product or service development is under way but the company is not generating revenues from sales.
Beta: For companies specializing in information technology, the phase at which the product is being tested by a limited number of customers but not available for broad sale. For life sciences companies, beta is synonymous with a drug in human clinical trials or a device being tested.
Shipping: The stage at which the product or service is being sold to customers and the company is deriving revenues from those sales but expenses still exceed revenues.
Profitable: The stage at which the company is selling products or services and the sales revenue yields a positive net income.
Restart: The stage at which the firm is recapitalized at a reduced valuation, accompanied by a substantial shift in the product or marketing focus.
Industry Groups
Computer hardware: Firms whose primary lines of business are personal computing, mini-computers or workstations, mainframe computers, CAD/CAM/CAE systems, data storage, computer peripherals, memory systems, office automation, source data collection, multimedia devices, and computer networking devices.
Computer software: Firms whose primary lines of business are compilers, assemblers, and systems, applications, CAD/CAM/CAE/CASE systems, recreational and home, artificial intelligence, educational, multimedia software, and online services.
Communications: Firms whose primary lines of business include modems, computer networking, fiber optics, microwave and satellite communications, telephone equipment, pocket paging, cellular phones, radar and defense systems, television equipment, teleconferencing, and television and radio broadcasting.
Medical: Firms whose primary lines of business include biotechnology, pharmaceuticals, diagnostic imaging, patient monitoring, medical devices, medical lab instruments, hospital equipment, medical supplies, retail medicine, hospital management, medical data processing, and medical lab services.
Regions
Eastern United States: Firms whose headquarters are located in Connecticut, Delaware, the District of Columbia, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, and West Virginia.
Western United States: Firms whose headquarters are located in Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
Source: Adapted from VentureOne (1998)
TABLE 1
Corporate Venture Capital Fund
Capital under
Corporate sponsor management
Electronic Data Systems 1,500
General Electric 1,500
Andersen Consulting 1,000
Comdisco 500
Time Warner 500
Times Mirror 500
Visa International 500
Intel Corporation 450
AT&T 348
Hikari Tsushin 332
News Corporation 300
ValueVision International 300
Comcast 250
PECO Energy 225
Siemens 210
Note: The estimated capital under management is shown in millions of
current dollars in 2000. If the corporation organizes multiple
programs, these are consolidated. Some corporations do
not make formal commitments in advance to their venture programs
or do not disclose the size of these commitments. These
firms are not included on the list. Among the largest corporate
venture capital programs falling into these categories are those
of Cisco, Dell, Johnson & Johnson, and Microsoft.
Source: Asset Alternatives (2000)
TABLE 2
Number of Corporate Venture
Capital Investments
Number of Dollar volume
rounds of rounds
1983 53
1984 91
1985 139
1986 129
1987 152
1988 179
1989 202
1990 233
1991 249
1992 214
1993 198
1994 193
1995 65 193
1996 101 369
1997 229 708
1998 391 1,449
1999 936 7,968
Note: The series reporting number of investments before 1995
and in and after 1995 may not be strictly comparable. For 1995
and after, the dollar volume of these investments (in millions of
2000 dollars) is also reported.
Source: Asset Alternatives (2000)
TABLE 3
Distribution of the Sample
Number of investments Number of Dollar
Total Corporate VC Independent VC rounds amount
1983 1,841 53 1,013 436 2,219
1984 2,249 91 1,206 550 2,905
1985 2,593 139 1,382 625 2,910
1986 2,557 129 1,381 592 2,394
1987 2,675 152 1,397 642 3,065
1988 2,599 179 1,385 611 2,687
1989 2,866 202 1,490 720 3,069
1990 2,826 233 1,455 784 3,640
1991 2,890 249 1,472 757 3,207
1992 3,166 214 1,699 911 3,891
1993 3,118 198 1,586 931 4,532
1994 2,984 193 1,601 947 4,973
Total 32,364 2,032 17,067 8,506 39,492
Note: The table depicts the number of venture capital investments in
the VentureOne sample by year between 1983 and 1994 as well as the
number of financing rounds (a round may consist of several investments
by different investors) and the aggregate amount of funding disbursed
(in millions of 1994 dollars). Similar tabulations of the number of
investments are presented for corporate and independent venture funds.
Source: VentureOne
Characteristics of Firms at the Time of Investment
Entire Corporate Corporate VC and Independent
sample VC only strategic fit VC only
Status at time
of investment
Start-up 9.8 7.1 6.4 10.4
Development 30.5 33.6 35.9 31.2
Beta 4.1 5.5 6.4 4.1
Shipping 45.5 44.4 42.9 44.8
Profitable 7.6 6.9 5.6 7.3
Restart 2.4 2.5 2.8 2.3
Location of firm
All western 59.7 63.7 59.6 60.8
United States
California 51.6 53.7 51.3 52.7
All eastern 24.1 25.2 29.1 23.4
United States
Massachusetts 12.8 14.0 16.5 12.6
Industry of firm
Medical 25.5 25.9 24.2 24.2
Computer hardware 16.7 17.0 16.2 16.8
Communications 14.5 14.2 22.1 15.5
Computer software 15.1 15.1 14.0 16.2
/on-line
services
Other 28.1 27.9 23.5 27.3
Round of investment
Mean 2.4 2.8 2.9 2.4
Median 2 3 3 2
Age of firm at time
of investment
Mean 3.9 4.0 4.2 3.8
Median 3.0 3.3 3.4 2.8
Amount invested in
venture round
Mean 6.1 6.2 6.0 5.7
Median 4.3 4.5 4.7 4.2
Note: The sample consists of 32,364 investments in privately held
venture-backed firms between 1983 and 1994. The table presents the
stage of the firm’s development at the time of the investment, the
geographic location of the firm, the industry of the firm, the ordinal
rank of the venture round, the age of the firm at the time of the
investment (in years), and the amount of the investment in the
financing round (in millions of 1994 dollars). Separate tabulations are
presented for investments by corporate venture firms, corporate funds
where there was a strategic fit between the parent and portfolio firms,
and independent venture funds.
Source: VentureOne
TABLE 5
Logit Regression Analyses of Strategic Fit of Corporate Investments
Was the corporate investment
in a related industry?
Age of firm at time of financing 0.0105 [1.16] 0.0116 [1.28]
Time trend 0.0351 [2.38] 0.0358 [2.42]
Firm is in development stage? 0.524 [2.96] 0.506 [2.86]
Firm is in beta stage? 0.697 [3.02] 0.6860 [2.97]
Firm is in shipping stage? 0.184 [1.02] 0.179 [0.99]
Firm is in profitable stage? -0.180 [-0.72] -0.176 [-0.70]
Firm is in restart stage? 0.456 [1.54] 0.463 [1.56]
Firm based in California? 0.098 [1.05]
Firm based in Massachusetts? 0.362 [2.92]
Log likelihood -2,878.1 -2,880.8
[chi square]-statistic 67.71 70.01
p-value 0.000 0.000
Number of observations 2,032 2,032
Note: The sample in the regressions consists of 2,032 corporate
investments in privately held, venture-backed firms between 1983 and
1994. The dependent variable is a dummy that takes the value of one if
the firm is in an industry that is related to the parent of the
corporation sponsoring the venture investment. Independent variables
include the age of the firm at the time of the investment, a time
trend, firms based in California and Massachusetts, the status of the
firm at the time of the investment, the industry of firm (not
reported), and a constant (not reported). All dummy variables take on
the value of one if the answer to the posed question is in the
affirmative. Absolute t-statistics are reported in brackets.
Source: VentureOne
TABLE 6
Status of Corporate and Independent Venture Investments
Entire Corporate Independent
Status at end of analysis sample VC only VC only
Initial public offering completed 31.1 35.1 30.6
Registration statement filed 0.7 0.2 0.7
Acquired 29.0 29.0 30.3
Still privately held 20.6 21.1 19.7
Liquidated 18.7 14.6 18.7
Corporate VC
Status at end of analysis and strategic fit
Initial public offering completed 39.3
Registration statement filed 0.3
Acquired 27.5
Still privately held 18.3
Liquidated 14.7
Note: The sample consists of 32,364 investments in privately held
venture-backed firms between 1983 and 1994. The table presents the
eventual outcome of the firms as a percentage of the sample. Separate
tabulations are presented for investments by corporate venture firms,
corporate funds where there was a strategic fit between the parent and
portfolio firms, and independent venture funds.
Source: VentureOne
TABLE 7
Logit Regression Analyses of Firms in Spring 1998
Observations are investments
Did firm go public?
Age of firm at time of financing -0.02 [5.52] -0.02 [0.50]
Round number 0.13 [11.39] 0.13 [11.18]
Corporate venture investment? 0.15 [2.54] -0.19 [1.31]
Independent venture investment? -0.003 [0.09] -0.002 [0.07]
Corporate investment and 0.52 [3.15]
strategic fit?
Firm based in California? 0.30 [9.29] 0.29 [8.96]
Firm based in Massachusetts? 0.36 [7.83] 0.36 [7.75]
Firm is in development stage? 0.44 [7.73] 0.42 [7.27]
Firm is in beta stage? 0.25 [2.83] 0.22 [2.50]
Firm is in shipping stage? 0.38 [6.28] 0.36 [5.95]
Firm is in profitable stage? 1.32 [17.08] 1.30 [16.61]
Firm is in restart stage? -0.56 [4.20] -0.56 [4.19]
Log likelihood -14,743.6 -14,252.0
[chi-sqaure]-statistic 2,409.9 2,362.4
p-value 0.000 0.000
Number of observations 24,515 23,740
Observations are investments
Did firm go public, register,
or have favorable acquisition?
Age of firm at time of financing -0.02 [6.17] -0.02 [6.13]
Round number 0.13 [11.48] 0.13 [11.29]
Corporate venture investment? 0.12 [2.15] -0.23 [1.64]
Independent venture investment? 0.07 [2.54] 0.07 [2.56]
Corporate investment and 0.57 [3.55]
strategic fit?
Firm based in California? 0.23 [7.44] 0.22 [6.98]
Firm based in Massachusetts? 0.24 [5.26] 0.23 [5.04]
Firm is in development stage? 0.38 [6.99] 0.35 [6.41]
Firm is in beta stage? 0.14 [1.60] 0.11 [1.24]
Firm is in shipping stage? 0.30 [5.20] 0.28 [4.82]
Firm is in profitable stage? 1.10 [14.77] 1.08 [14.27]
Firm is in restart stage? -0.43 [3.64] -0.45 [3.71]
Log likelihood -15,477.4 -14,973.7
[chi-sqaure]-statistic 2,065.5 2,025.7
p-value 0.000 0.000
Number of observations 24,515 23,740
Note: The sample in the first four regressions consists of 32,364
investments in privately held, venture-backed firms between 1983 and
1994; in the fifth and sixth regressions, the sample consists of 8,506
financing rounds of privately held, venture-backed firms between 1983
and 1994. The dependent variable in the first, second, fifth, and sixth
regressions is a dummy variable that takes on the value of one if the
firm had gone public by the spring of 1998. In the third and fourth
regressions, the dummy takes the value of one if the firm had gone
public, filed a registration statement, or been acquired at twice the
postmoney valuation (in inflation-adjusted dollars) at the time of the
investment by the spring of 1998. Independent variables include the age
of the firm at the time of the investment, the ordinal rank of the
investment round, and dummy variables denoting investments by corporate
and independent venture capital funds, corporate venture investments
where there was a strategic fit with the portfolio firm, firms based in
California and Massachusetts, the status of the firm at the time of the
investment, the year of the investment (not reported), the industry of
firm (not reported), and a constant (not reported). All dummy variables
take on the value of one if the answer to the posed question is in the
affirmative. Absolute t-statistics are reported in brackets.
Source: VentureOne
(1.) For a discussion of typical independent venture capital fund structures, see Gompers and Lerner (1996, 1999).
(2.) For a discussion of staging and its implications for investors, see Gompers (1995).
(3.) This history of corporate venture capital is based on Fast (1978); Hardymon, DeNino, and Salter (1983); Venture Economics (1986); and assorted press accounts. It is largely based on the history of corporate venture capital presented in Gompers and Lerner (1998b).
(4.) Because many corporations do not report their private investments in entrepreneurial firms, these figures should be regarded as conservative estimates of the level of corporate venture capital activity. The true level would be higher.
(5.) Information about the financing of private firms is typically not revealed in public documents, and investors and entrepreneurs may consider this to be sensitive information. VentureOne seeks to overcome this reluctance by emphasizing that its database also helps firms obtain financing. In particular, firms can alert investors whether they intend to seek further private financing or intend to go public in upcoming months.
(6.) See Gompers and Lerner (2000) for an analysis of the comprehensiveness of the VentureOne database over time. Concerns about selection biases are addressed by repeating the analyses below using observations from only 1988 to 1994, when VentureOne’s coverage of the industry was much more comprehensive. The results are little changed.
(7.) The reader may note that the dollar amounts reported here are greater in some years than the cumulative disbursements from venture capital funds reported elsewhere (for example, Kortum and Lerner 1998). This disparity reflects the fact that the VentureOne data represent total financings from all sources for privately held venture-backed firms rather than just funds from venture capital organizations.
(8.) See the appendix for definitions of stages, regions, and industries.
(9.) For a discussion of return in the venture capital industry, see Venture Economics (1988, 1997).
(10.) The postmoney valuation is defined as the product of the price paid per share in the financing round and the shares outstanding after the financing round. In calculating the valuations, VentureOne converts all preferred shares into common stock at the conversion ratios specified in the agreements. Warrants and options outstanding are included in the total as long as their exercise price is below the price per share being paid in the financing round.
The results are also robust to the use of a third dependent variable, the probability that the firm has not been liquidated by the spring of 1998.
(11.) The discussion of Exxon Enterprises is based on the Venture Economics study cited above.
(12.) The account of the Java Fund relies on Fisher (1996) and on .
(13.) The BlueLight Fund is discussed in Chesbrough and Rotelli (2000).
(14.) The eLoyalty Fund is documented in Leibowitz (2000).
REFERENCES
Asset Alternatives. 2000. The corporate venturing directory and yearbook. Wellesley, Mass.: Asset Alternatives.
Block, Zenas, and Oscar A. Ornati. 1987. Compensating corporate venture managers. Journal of Business Venturing 2 (1987): 41-52.
Chesbrough, Henry W., and Mary Teichert Rotelli. 2000. Hotbank: Softbank’s new business model for early stage venture incubation. Harvard Business School Case No. 9-600-100.
Corporate Technology Information Services. 1996 (and earlier years). Corporate technology directory. Woburn, Mass.: Corporate Technology Information Services.
Dun’s Marketing Services. 1996 (and earlier years). Million dollar directory. Parsippany, N.J.: Dun’s Marketing Services.
Fast, Norman D. 1978. The rise and fall of corporate new venture divisions. Ann Arbor, Mich.: UMI Research Press.
Fisher, Lawrence M. 1996. $100 million fund will finance Java-based ventures. New York Times, August 22, D4.
Gale Research. 1996 (and earlier years). Ward’s business directory of U.S. private and public companies. Detroit: Gale Research.
Gee, Robert E. 1994. Finding and commercializing new businesses. Research/Technology Management 37 (January/February): 49-56.
Gompers, Paul A. 1995. Optimal investment, monitoring, and the staging of venture capital. Journal of Finance 50:1461-89.
Gompers, Paul A., and Josh Lerner. 1996. The use of covenants: An analysis of venture partnership agreements. Journal of Law and Economics 39:463-98.
–. 1998a. Venture capital distributions: Short- and long-run reactions. Journal of Finance 53:2161-184.
–. 1998b. The determinants of corporate venture capital success: Organizational structure, incentives, and complementarities. In National Bureau of Economic Research Conference Volume on Concentrated Ownership. Cambridge, Mass.: NBER.
–. 1998c.What drives venture capital fundraising? Brookings papers on economic activity: Microeconomics 1998: 149-92.
–. 1999. An analysis of compensation in the U.S. venture capital partnership. Journal of Financial Economics 51:3-44.
–. 2000. Money chasing deals? The Impact of fund inflows on private equity valuations. Journal of Financial Economics 55:281-325.
Hardymon, G. Felda, Mark J. DeNino, and Malcolm S. Salter. 1983 When corporate venture capital doesn’t work. Harvard Business Review 61 (May/June): 114-20.
Henderson, Rebecca 1993. Underinvestment and incompetence as responses to radical innovation: Evidence from the photolithographic alignment equipment industry. Rand Journal of Economics 24:248-70.
Henderson, Rebecca, and Iain Cockburn. 1996. Scale, scope, and spillovers: The determinants of research productivity in drug discovery. Rand Journal of Economics 27:32-59.
Jensen, Michael C. 1993. Presidential address: The modern Industrial Revolution, exit, and the failure of internal control systems. Journal of Finance 48 (July): 831-80.
Kortum, Samuel, and Josh Lerner. 1998. Does venture capital spur innovation? Boston University and Harvard University. Unpublished working paper.
Leibowitz, Alissa. 2000. Bain, Sutter, and TCV to invest in eLoyalty Fund. Venture Capital Journal 40 (September): 16-18.
Lerner, Josh. 1997. An empirical examination of a technology race. Rand Journal of Economics 28:228-47.
National Register Publishing Company. 1996 (and earlier years). Directory of leading private companies, including corporate affiliations. New Providence, Wilmette, N.J.: National Register Publishing Company.
Reinganum, Jennifer R. 1989. The timing of innovation: Research, development, and diffusion. In The Handbook of Industrial Organization, edited by R. Schmalensee and R.D. Willig. New York: North-Holland.
Rind, Kenneth W. 1981. The role of venture capital in corporate development. Strategic Management Journal 2:169-80.
Siegel, Robin, Eric Siegel, and Ian C. MacMillan. 1988. Corporate venture capitalists: Autonomy, obstacles, and performance. Journal of Business Venturing 3:233-47.
Sykes, Hollister B. 1990. Corporate venture capital: Strategies for success. Journal of Business Venturing 5:37-47.
Venture Economics. 1986. Corporate venture capital study. Unpublished manuscript.
–. 1988. Exiting venture capital investments. Needham, Mass.: Venture Economics.
–. 1997. Investment benchmark reports-venture capital. New York: Venture Economics.
VentureOne. 1998. VentureOne 1997 Annual Report. San Francisco: VentureOne.
PAUL A. GOMPERS The author is a professor of business administration at Harvard Business School and a research associate at the National Bureau of Economic Research. He thanks VentureOne for making this project possible through generous access to its database of venture financings. He thanks Harvard Business School’s Division of Research for financial support.
COPYRIGHT 2002 Federal Reserve Bank of Atlanta
COPYRIGHT 2004 Gale Group