A Wave of Acquisitions May Have Shielded Big Tech from Competition
The fraction of startups that are acquired instead of going public has skyrocketed over the last few decades, according to a new study co-authored by Florian Ederer at Yale SOM. This acquisition spree may have eliminated many potential competitors of big tech firms, the research suggests.
When Facebook announced in 2012 that it was acquiring Instagram for $1 billion, many critics questioned the decision to pay such a huge amount for a tiny startup with no revenue. Florian Ederer, an associate professor of economics at Yale SOM, remembers telling his MBA students, “This makes no sense.”
In retrospect, Facebook’s acquisition looks like a canny business move to strengthen its social media dominance—and in the last few years, lawmakers and regulators have alleged that the deal was intended to suppress competition. If Instagram were a standalone business today, it might pose a credible threat to Facebook (now called Meta). But at the time, people didn’t fully recognize Instagram’s potential to compete with the popular social media network.
Those companies that have been particularly voracious acquirers of VC-backed startups are also ones that have become increasingly insulated from product market competition.
In a new study, Ederer and his colleague Bruno Pellegrino of the University of Maryland investigated whether the huge wave of startup acquisitions in recent years has insulated big tech firms from competition. The researchers found that the percentage of venture capital-backed startups that were acquired—rather than going public—skyrocketed from roughly 10% to 90% over the last three decades.
While it had been clear acquisitions were growing, “I don’t think anybody has quite so viscerally documented how big this shift has been,” Ederer says.
Meanwhile, four big players in tech—Meta, Google, Apple, and Amazon—have also seen a decrease in what the researchers call “product market centrality.” This is a measure of how similar their products are to other firms’ products; the lower the score, the closer the company is to having monopoly power.
In other words, the number of small firms that can challenge big tech companies has dwindled—perhaps because most of them have been snapped up.
“Those companies that have been particularly voracious acquirers of VC-backed startups are also ones that have become increasingly insulated from product market competition,” Ederer says.
In theory, acquisitions can benefit consumers. In the best-case scenario, bringing a small company under the wing of a larger one enables the product to be deployed at a bigger scale. For instance, the product could be integrated into the big company’s platform and reach more users. Or the acquiring firm’s resources allow the product to be improved in ways that wouldn’t be possible if the startup had remained independent.
On the other extreme are “killer acquisitions,” as Ederer and his co-authors called them in a previous study: A large company buys a startup and simply lets the product die, eliminating a competitive threat.
Ederer and Pellegrino were interested in the gray area in the middle. What if big companies weren’t killing off startups outright, but acquiring as many as possible to reduce competition?
To investigate, the researchers obtained data from the National Venture Capital Association on VC-backed start-ups from 1985 to 2019. In the 1980s, roughly 90% of those small firms had gone public, and about 10% had been acquired.
By 2019, “this pattern has entirely reversed,” the team writes; about 90% of startups were acquired. While the number of VC-backed start-ups has grown dramatically over the last few decades, the number of IPOs hasn’t increased much; instead, “the really, really huge growth is in acquisitions,” Ederer says.
The reason for the turnaround is “a bit of a puzzle,” he says. Perhaps startup founders want the stability that comes with being part of a large company, rather than competing against an established firm. Founders also may have become more tempted to accept acquisition offers as the sums reached dizzying heights.
In addition, antitrust enforcement may have become more lax. Or perhaps the process of an IPO has become too laborious.
Finally, big tech firms might have become more aggressive about acquiring early-stage startups that weren’t even strong competitors at the time, but could pose a threat later.
To explore this possibility, the team used information from the Hoberg-Phillips Data Library about companies’ products from the firms’ SEC forms to determine how similar two firms’ products were.
Using a methodology first developed in previous work by Bruno Pellegrino, the researchers then computed a product market centrality score for each company. If centrality was high, that meant its products were very similar to those from other firms. In this scenario, the firm couldn’t charge much of a markup because consumers had many similar choices.
If centrality was low, the company could charge more because very few similar products were available. A centrality score of zero meant the firm held a monopoly.
The team then focused on five firms that had been very active in acquisitions: Apple, Amazon, Microsoft, Meta, and Google (and its parent company Alphabet). For all of the companies except Microsoft, the firms’ relative product market centrality in the tech industry decreased from 2001 to 2019. In other words, fewer and fewer other tech companies offered similar products.
One could argue that this pattern isn’t a reason for concern; it just means that these giant firms provide very unique products. But the trend also supports the hypothesis that “they’ve eliminated all the competition and potential competition,” Ederer says. (Microsoft may be an outlier because the company came under tremendous antitrust scrutiny in the 1990s and may have scaled back acquisitions.)
The researchers also examined the startups that did go public. For each company, they analyzed data such as the quantity and cost of products offered. The team could then estimate a measure called the entrant productivity premium, which captured how productive a firm had to be in order to go public.
From 1997 to 2019, the average entrant productivity premium jumped from around 10 to over 20%. In other words, firms had to be more productive in order to go public, perhaps because having this edge was necessary to challenge existing companies or because the process of an IPO had become more onerous.
The study signals that a shift in antitrust policy is needed, Ederer says. With big tech firms, “we have seen hundreds of acquisitions, and until recently not a single one has ever been stopped,” he says.
The pushback is already starting. Over the last year, the Federal Trade Commission has tried to block Meta’s acquisition of the virtual reality firm Within Unlimited (though a judge denied the agency’s attempt to stop the deal) and Microsoft’s acquisition of the video game company Activision Blizzard. “I think we’re going to see more of these types of challenges,” he says.