The venture capital funding model that dominates the tech industry is focused on the “exit strategy”— the ways funders and founders can cash out their investment. While in common lore the exit strategy is an initial public offering (IPO), in practice IPOs are increasingly rare – they now account for fewer than 1 in 10 exits for start-ups, and happen later in a company’s life than they used to.
Instead, most companies that succeed exit the market by merging with an existing firm. Innovative start-ups are especially likely to be acquired by the dominant firm in the market, particularly when they are venture funded, for a variety of reasons – because the dominant firms value the target’s technology, because they have lots and lots of money, or to eliminate a potential competitor who might leapfrog them in Schumpeterian competition. This paper argues that this focus on exit, particularly exit by acquisition, is pathological and one of the main reasons for the collapse of Schumpeterian competition in the tech industry. The structure of the VC market leads to concentration in the tech industry, reinforcing the power of dominant firms. It short-circuits the development of truly disruptive new technologies. Crucially, it means that the public loses access to many of the most promising new technologies.
To address this, this paper – which is available here – advances a number of measures to reduce the incentive to sell successful start-ups to incumbents, such as making it easier to take a company public, and making it harder for VCs to sell to incumbents, such as a targeted ban on acquisitions of thriving companies by dominant firms in the same or adjacent markets. Alternatively, those mergers should at least be subject to extra scrutiny by antitrust authorities, who should presumptively block them if an alternative acquirer is available or if an alternative deal would keep the company afloat.
Part I describes how incumbent acquisition became the dominant exit strategy for VC-backed firms.
Venture capital fills a gap in funding risky enterprises. While VCs are funding more firms than ever, which should promote Schumpeterian disruption and competition, the share of U.S. start-ups that sell to incumbents rather than compete as public companies or private concerns has shot up in the last twenty years. This is a consequence of the VC business model.
People who put time or money into companies – funders or founders – eventually want to reap their rewards. There are two main ways VCs can cash out their equity: by selling it to the public through an IPO, or by selling it to another firm. The increase of VC exits in recent years increased from about 380 per year to about 530. However, while in the late 1990s about 200 VC-backed companies went public per year on average, in recent years only about 75 have done so—a more than 60% decrease. While one in two exits were by IPO as recently as the 1990s, only about 1 in 10 exits are by IPO today. At the same time, the number of VC-backed firms acquired has jumped from 190 per year in the 1990s, to 450 per year recently —a 140% increase. And even those firms that do exit through public offerings are increasingly using offerings not to continue to run the business, but as a step towards being acquired. A high percentage of VC-backed firms that IPO go private by acquisition shortly thereafter. Some VCs anticipate this ultimate acquisition; rather than cash out upon the IPO, they “IPO-and-hold” – a strategy quite common in areas such as biotech.
Many of these acquisitions have involved a dominant incumbent buying a start up in the same or related market. Looking at the pattern of incumbents’ acquisitions shows their focus on acquiring nascent rivals and on controlling strategic complements that could otherwise destabilise their core business. Facebook, for instance, has acquired 90 start-ups—some of which were clear rivals, while others were unpredictable complements. These incumbent acquisitions have consequences: companies like Google and Facebook have remained dominant for decades. They stay on top by buying out the companies that might otherwise displace them, and they do so on a scale never before seen, staving off Schumpeterian competition.
Part II explain how this state of affairs came to be.
Several factors explain why companies increasingly exit through acquisitions instead of public offerings, including the chance to share in the incumbent’s scale economies and monopoly rents. The authors argue that the VC business model potentiates a number of these factors. First, the rewards to scale are greater than ever. Without complementary assets and scale, small firms suffer. This may encourage exits through acquisition instead of public offering. As the time and cost to replicate complementary assets and reach scale increases, cooperation with incumbents increases. At the same time, while large firms may not want to take the risk that a small start-up succeeds, small innovators may not want to run the risk of failing. Against this background, the VC model – and the need for large and quick returns – strongly favours investing in entrants that will exit early by selling to incumbents with scale and scope.
Second, the VC model positions start-up founders to recognise powerful incumbents, and to profit not by competing with, but by preserving and sharing in that dominance. The bigger the threat the start-up poses, the bigger the premium (i.e. acquisition price) the incumbent is willing to pay. Even companies ready to “IPO” sometimes opt to be acquired instead—using the IPO as leverage. VCs intensify this driver by selecting firms that pursue this strategy and then advising them on how to achieve it.
Reaching an exit by incumbent acquisition is not equally likely for all firms; it is far more likely for firms that work with VCs who have completed a past acquisition. Asymmetric information and incentives may also drive start-ups to exit through incumbent acquisitions rather than public offerings. Incumbents may understand a start-up’s market opportunity better than public traders understand, and be willing to pay the firm’s full value. Venture capitalists appear to solve information asymmetries and align estimates of value better than founders can do alone, leading VC-backed companies to more frequently achieve acquisition exits than others do.
Part III explains why this state of affairs is problematic.
There are several reasons to be concerned that start-ups tend to be acquired by incumbents rather than go public or merge with another maverick – and that VCs intensify this phenomenon.
- First, concentration in the tech industry is a large and growing problem. The companies that dominate the digital economy are all more than 20 years old and have dominated their market categories for more than a decade. Today’s tech monopolists have almost certainly held onto and even broadened their monopolies by acquiring firms that in another era would have displaced them. At the very least, these acquisitions have reduced the likelihood of disruptive innovation that would challenge the power of those monopolies.
- Second, incumbent acquisition has contributed to increasing concentration of technological capacity. Technology is diffusing from leaders to followers more slowly than it used to.
- Third, and perhaps most problematic, tech giants often buy up promising start-ups only to shut them down. Sometimes this is intentional. Economists have documented cases of “killer acquisitions”—companies that buy incipient competitors in order to eliminate the threat they pose. While especially prominent in biotech, the practice is also prominent among big tech firms: Oracle, Google, and Facebook have all bought and shut down competing firms, sometimes in the same day. Start-ups acquired by incumbents fall into three basic categories: direct competitors, companies that offer complementary products and companies that might change the nature of the market altogether. Purposeful killer acquisitions seem most likely of direct competitors and perhaps companies that threaten the business model altogether.
On the other hand, there are at least three reasons why any particular merger may be beneficial. First, some technologies might work well only at scale. Second, the incumbent might get the innovation into the hands of more people, simply because it has more customers. Third, the market leaders may be best positioned to put complementary technologies to work. There are also other reasons we might want to put up with anticompetitive consequences in order to encourage investment in start-ups However, while these may be valid points in particular cases, they do not disprove nor help solve the problems of concentration caused by the selling of start-ups to incumbents.
Part IV asks what is to be done.
The authors offer a combination of carrots and sticks designed to better align start-up funding and exits with social welfare.
One carrot would be to make IPOs easier. This could involve steps such as increasing the information available to companies seeking to go public, reducing the regulatory requirements to bring an IPO, and changing the tax treatment of acquisitions compared to public offerings. A second carrot would be to make it more attractive to stay in business as an independent company rather than sell – not as a flashy unicorn, but as a business that continues to make a modest profit. In terms of incentives for VCs, one could facilitate pre-IPO secondary markets to allow VCs and employees to sell their shares without selling the company as a whole, and so reduce the pressure on the company to merge. Along the same lines, late-stage funders could be encouraged to cash out early-stage funders, rather than just add to the investment in the company. A third carrot would consist of encouraging alternatives to VC funding, e.g. by favouring debt over equity as a funding mechanism for start-ups.
As regards sticks, one could develop disincentives to use venture capital or, more plausibly, disincentives to sell to incumbents. One way to discourage mergers and encourage companies to continue to operate is to vary the tax treatment of those two options. We should tax mergers to force companies to internalise the cost the merger imposes on society. A firm that sells out does not bring the same benefits to society as a firm that continues to compete. We might consider not only changes to tax law designed to entice companies to continue to operate, but others to discourage companies from merging with incumbents (e.g. by taxing goodwill created by the acquisition, i.e. the difference to fair market value). To negate the advantages for VCs of being able to monetise their investments quickly by selling when compared to an IPO, we could require a lockup period for stock sales. A stronger version would require that the merged firm meet competitively important goals before paying out. To ensure this is targeted at acquisitions by incumbents, this lock up could be tied to the merger review period – either ex ante or ex post.
A more direct approach would be to prohibit or restrict some incumbent mergers altogether. Unfortunately, the existence of antitrust laws regulating mergers has not prevented exit strategies from leading to unprecedented concentration in technology markets, nor has it prevented killer acquisitions in other innovative fields like biotechnology. Indeed, there is some evidence that acquirers structure their transactions in part to avoid detailed antitrust scrutiny. Further, the nature of high-tech markets makes merger analysis more difficult. Another fact that complicates merger analysis is that many start-up acquisitions are not of direct competitors but conglomerate mergers of companies that may potentially become competitors in the future.
Yet, the law can and should do more to limit the sale of innovative start-ups to incumbents. One (rather extreme) possibility is just to ban mergers altogether. However, this would go too far and, given the small number of IPOs, it is reasonable to worry that a flat ban on mergers would discourage venture investment too much. Nonetheless, there is room for antitrust to regulate acquisitions of start-ups more than it currently does. First, agencies should pay particular attention to acquisitions by incumbent monopolists even if they do not present as direct competitors. Acquisitions of adjacent firms are likely to increase concentration and prevent the development of fundamentally new sources of competition. Second, antitrust agencies should presumptively block acquisitions of directly competitive start-ups by dominant firms. When start-ups are not direct competitors, a presumption against mergers that involve adjacent or potentially market-disrupting technologies may also be appropriate. This presumption should be rebuttable by sufficient proof of efficiencies from the merger, or by strong evidence that the start-up’s technology is uniquely complementary to the incumbent’s so that it is unlikely to be profitably deployed by anyone other than the incumbent.
As they say in California, this is good stuff. Perhaps there is a bit too much of it – even when they are easy to read, I fail to understand why American legal articles must all be as long as a monograph – , and it covers topics going well beyond competition law. Regardless, this piece does a really good job of explaining why VCs matter for the tech industry, what problems concerning the digital sphere are exacerbated by those VCs’ business model, and of proposing solutions for those problems.
I would expect some to argue that these solutions do not go far enough. However, if one considers solely the impact that funding models have on start-ups, the proposals strike me as trying to achieve quite a bit.
I particularly enjoyed the recognition that the problem is not so much individual acquisitions – which may be individually beneficial – but the pattern of systematic acquisitions that needs to be broken up. This is a point overlooked in many of the papers concerning acquisitions by digital titans – such as those I reviewed during these weeks. It is striking – to me at least – that recent agency reports and legislative proposals go far beyond what is proposed in the literature, which may be a result of those proposals purposefully seeking to go beyond the existing framework. This paper can likewise be commended for thinking outside the existing legal toolbox.