Supporters of more robust antitrust policy have pointed to the subsequent success of Figma after authorities blocked Adobe’s acquisition of it. Skeptics, including venture capitalists, have argued that the one case reveals nothing systematic about the benefits of stronger merger review. Venture capitalists happen to be the one party with the data and resources to fund the studies to show any systematic correlation one way or the other. They should do so, writes Shishene Jing.
When Figma priced its initial public offering at a valuation north of $19 billion in the summer of 2025—with shares briefly pushing the company’s market capitalization toward $68 billion on the first day of trading—it revived a fight that had lain dormant since Adobe abandoned its $20 billion bid for the company in December 2023. Lina Khan, the former chair of the Federal Trade Commission whose scrutiny during her tenure (alongside those from European and British regulators) helped kill that deal, took a victory lap, framing the offering as proof of the value of letting startups grow into independently successful businesses rather than folding them into incumbents. Vinod Khosla, the Sun Microsystems co-founder and a venture capitalist unaffiliated with Figma, was not persuaded. He shot back on X that Khan was in no position to override the judgment of founders, employees, and investors. He also accused her of falling into “retrospective predictability“: the claim of foresight for an outcome that was only obvious after the fact.
The debate extends beyond the unconsummated merger between Figma and Adobe to an empirical question neither side has the data to answer with confidence: whether blocking acquisitions of fast-growing, well-funded targets systematically produces more value—for the market, for consumers, for the target itself—than letting the deal close. Khan has one data point. Khosla has a theory about who should be trusted to weigh it. Neither has anything close to a systematic answer, because the field that would generate one—merger retrospective research—is chronically underfunded, data-starved, and structurally biased toward the easiest cases to study. That’s a problem Khosla and venture capitalists are unusually well-positioned to fix and has every self-interested reason to do so.
Figma’s story
Adobe agreed to acquire Figma in 2022 for $20 billion, or roughly 50 times Figma’s revenue at the time. Regulators concluded there was no clear path to approval because of competition concerns and the companies mutually terminated the deal in December 2023. Adobe paid Figma a $1 billion reverse termination fee, and Figma used the capital to keep growing. Over the next 18 months, revenue climbed roughly 40% year over year, the company turned profitable, and it built out new AI-powered products before filing to go public in mid-2025. The IPO priced at a valuation below what Adobe had originally offered, but the stock’s first-day pop pushed Figma’s market value well past the abandoned deal price, delivering strong returns to its venture fund backers: Sequoia, Andreessen Horowitz, Kleiner Perkins, Index, Thrive, and others.
That is a genuinely good outcome for Figma’s shareholders and, on its face, a defensible data point for the proposition that blocking the deal didn’t destroy value. But a single successful IPO following a single blocked acquisition tells you almost nothing about the general base rate of outcomes across blocked deals. Data for this comparison is lacking. We don’t know how many would-be Figmas were quietly absorbed by incumbents in adjacent markets during the same window, what happened to competition in those markets afterward, or how many companies facing the same choice architecture—sell to a short list of natural acquirers, or bet on an independent path—made the opposite choice and are worse off for it. Khosla’s “retrospective predictability” jab is, on this narrow point, not wrong: knowing that this particular blocked deal worked out doesn’t establish that blocking deals generally works out. But the same limitation applies with equal force to using any one case to argue merger enforcement was too aggressive.
The evidence that already exists
What complicates Khosla’s framing is that to the extent to which we do have systematic evidence, it doesn’t support the idea that antitrust agencies have been overzealous or are wrong to think stronger enforcement generally benefits competitive outcomes. Economist John Kwoka’s meta-analysis of merger retrospective studies—compiled in his 2015 book Mergers, Merger Control, and Remedies and updated in subsequent work—pulled together dozens of academic studies that measured actual post-merger price effects using pre- and post-merger data. His finding, discussed in ProMarket in 2017, was that a clear majority of the mergers studied showed that mergers led to meaningful price increases, averaging somewhere in the high single digits to low double digits after controlling for other factors. The effect was considerably larger for mergers cleared with behavioral remedies like requirements for access to proprietary technology, as opposed to structural remedies like divestitures. That is the opposite of what you’d expect if agencies were routinely blocking or burdening deals that posed no competitive risk. If anything, the existing retrospective literature suggests enforcers have historically been too permissive, not too aggressive, which is precisely the argument around which Khan built her tenure as FTC Chair.
The catch is that this literature has a structural blind spot, and it’s the same spot that concerns Figma. The FTC’s own account of its Merger Retrospective Program—which it revamped and expanded in 2020—acknowledges that most retrospectives study horizontal mergers between established competitors and focus on short-run price effects, because that’s what the available data supports. Retrospectives are harder to do on deals involving nascent or potential competitors, non-price effects like innovation and quality, or—most relevant here—deals that never closed in the first place. There is limited systematic literature on what happens after a proposed acquisition of a fast-growing startup is abandoned: whether the target thrives independently, stagnates, gets acquired by someone else, or quietly fails.
Why retrospectives are scarce
The reasons why merger retrospectives are in general so limited is because they require pre- and post-merger data on prices, output, quality, or innovation that is usually proprietary, litigation-sensitive, or simply expensive to assemble. These include market capitalization tables, customer contracts, internal pricing data, and competitor responses. Government agencies have some access to this through compulsory process, but the FTC’s Bureau of Economics has been candid that its retrospective work is resource-constrained and dependent on data that happens to become available through litigation or voluntary cooperation, which skews the sample toward completed, contested deals rather than the abandoned ones.
Academic researchers face the same problem accessing data, even more so without subpoena power, and they operate under publication incentives that favor clean, tractable identification strategies over the harder, more policy-relevant question of counterfactual outcomes for deals that didn’t happen. No one in academia or the government is well positioned to build the natural-experiment panel that would actually resolve the Khan-Khosla argument: a set of proposed acquisitions of well-funded, fast-growing targets, some blocked and some consummated, tracked over five- and ten-year horizons for competitive effects, innovation, pricing, and target-company outcomes.
The case for venture capital
However, there is one party well-placed to access the data to create this study: venture capital. VCs have capitalization tables, funding histories, exit outcomes, and visibility into deals. VCs could track data on acquisition offers, acceptances, and ultimate results for investors across deal types. They could then create a dataset showing how the value and products of the acquired startup compared to those startups that went unacquired in terms of prices, quality, and innovation, focusing especially on evaluating companies that went public and comparing them to those that were acquired. They could maintain data on other relevant characteristics of the companies that insiders deem relevant. A research vehicle capitalized by a consortium of funds, structured with the kind of data-sharing and firewall arrangements universities already use to protect competitively sensitive information, could build exactly the part of the dataset that’s currently missing: blocked-versus-consummated acquisitions of similarly positioned targets, tracked over time, by people with access to the numbers that would make the analysis rigorous rather than anecdotal.
Not only do VCs have the data, but they have the incentive, too. A VC fund that actually understood the base rate—how often blocked or abandoned deals lead to strong independent outcomes versus stagnation, how enforcement patterns have shifted the realistic exit menu, which sectors show the clearest evidence of acquisition suppressing a nascent competitor’s independent trajectory—would be better able to price risk and structure portfolios.
The obvious objection is that VCs funding this research would bias it: a Khosla-funded retrospective program would simply produce papers confirming Khosla’s priors. Yet, plenty of research areas run on privately funded, academically housed work with disclosure rules, blind data protocols, and independent peer review that keep the funder’s thumb off the scale. The structure that might work here is an endowed program housed at an independent research center— the kind of institution that already publishes retrospective work, like the Stigler Center or another university-based center with a track record of methodological rigor—funded through a pooled vehicle so no single fund or corporate entity controls the findings, with data-sharing terms that protect portfolio companies while giving researchers what they need to do real pre/post comparisons.
There’s also a sharper way to put the case to VCs: if you actually believe what Khosla tweeted—that founders, employees, and investors know better than regulators whether an acquisition or an independent path serves competition and the company—then you should want the retrospective evidence to exist, because on the numbers we do have, the opposite belief is better supported.
Conclusion
Figma will keep getting cited by both sides of this argument for years, because it’s vivid, recent, and has a stock price attached. It shouldn’t have to do that work alone. The empirical question underneath the Khan-Khosla exchange is answerable with better data and more studies. Venture capital has the balance sheets, the data access, and, if the industry means what it says about trusting founders and markets over regulators, the self-interest to fund the research that would finally put a real sample size behind an argument that has been running on tweets and ideology.
Author’s Disclosure: The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
