All-In Podcast: Coatue's Laffont Sees $4 Trillion AI IPO Wave as the Ecosystem's Great Reckoning
Thomas Laffont of Coatue Management presents at the All-In Liquidity Summit 2026, June 4, 2026
Thomas Laffont, co-founder of Coatue Management, made his podcast debut on the All-In show this week, and he did not show up empty-handed. Armed with two weeks of proprietary data work, the manager of one of technology investing's most respected $55 billion platforms laid out a sweeping case that the private market ecosystem is not just healing — it is on the verge of a liquidity event unlike anything the industry has seen in a generation. The central thesis: SpaceX, OpenAI, and Anthropic are about to return more capital to the ecosystem in a matter of months than the prior decade of exits combined.
The Unicorn Economy Has Quietly Rebalanced
Laffont opened with a data point that reframes the narrative around private market health. The unicorn economy, on average, is up 70% since September 2024, roughly mirroring the public market's move over the same period. But the more telling structural shift is in the composition of funding. The so-called "unicorn factory" that peaked in the ZIRP era of 2021 — which produced 479 new unicorns in a single cohort — has normalized sharply. The result, mathematically, is that funding per unicorn has increased fivefold since 2021. Fewer companies are being minted, but each one is commanding dramatically more capital.
That concentration matters because it cuts both ways. Coatue's data shows that less than 20% of the 2021 unicorn cohort had either raised a new round or exited within 20 quarters of achieving unicorn status, compared to roughly 80% of the pre-ZIRP cohort at the same point in their lifecycle. The 2024 AI cohort is the open question, and Laffont is clearly watching which historical pattern it tracks.
Three IPOs That Change the Math for an Entire Decade
The most striking data point of Laffont's presentation was the exit pipeline. With SpaceX weeks away from going public and Anthropic having filed its S-1 confidentially as of the day of the summit, Coatue estimates that just those three companies — SpaceX, OpenAI, and Anthropic — will return more capital than the prior ten years of venture exits combined. That is not rhetorical flourish; Laffont presented it as a direct quantitative comparison. For an ecosystem that has been chronically imbalanced between cash consumed and cash returned, this is a structural correction, not a cyclical one.
The revenue trajectory underpinning those valuations is what makes the IPO math credible rather than speculative. Coatue's chart, which Laffont noted starts only in January 2025, shows OpenAI and Anthropic surpassing Workday, then ServiceNow, then Adobe, then Salesforce in revenue scale within months of each other. As of the summit, both companies had surpassed Google Cloud and Azure. Coatue estimates one of them could exceed AWS by year-end and potentially surpass all of Microsoft's revenue by 2028. Laffont was measured in his framing — "this is just based on some assumptions and some forecasts" — but the directional message was unambiguous.
The SpaceX Framework: Why Valuation Per Launch Is Rising, Not Falling
Laffont spent meaningful time on SpaceX, offering Coatue's internal investment framework, which he called their "code framework," to explain what has been a counterintuitive market signal. The number one driver correlated to SpaceX's valuation is launch cadence — which is intuitive. But the ratio of valuation to number of launches has also been rising, which should not be the case if this were simply a commoditized launch services business.
Coatue's explanation is that the quality of SpaceX's business model structurally improves the more it launches. In the early phase, launches serve a small number of government customers in a one-time, unpredictable revenue model. As SpaceX built out Starlink, it entered a recurring revenue constellation business. Now, Laffont argues, SpaceX is becoming a platform — one where governments, militaries, and enterprises will seek to operate their own constellations, with SpaceX as the enabling infrastructure. Optionality from space data centers, lunar and Martian applications would layer on top of that.
On the IPO price specifically, Laffont was candid about the limits of his visibility: "I don't know whether $1.75 is the right price for the IPO and frankly I have no clue." What he does know is that the global telco and broadband profit pool is between $200 and $400 billion, and Starlink is addressing it with a product that, in his words, "works all the time, no radio towers." His framing was direct: "I believe that within a few years Starlink will power a device which will actually enable you to make a phone call anywhere in the world, and we think that's a solved problem."
The Centacorn Paradox: Bigger Companies Compound Faster
One of the more analytically surprising findings Laffont presented was what he called the 10x paradox. Coatue looked at three cohorts of companies across both public and private markets and asked: what are the historical odds of achieving a 10x return from each valuation tier? The data showed that unicorns — companies over $1 billion — have roughly an 8% chance of ever becoming decacorns. Decacorns, in turn, have only a marginally better 8-13% chance of reaching $100 billion. But centacorns — companies already valued at $100 billion or more — have a 31% probability of achieving a subsequent 10x. Scale, it turns out, breeds further compounding rather than mean reversion, at least for the cohort that makes it that far.
The Q&A pushed this logic one step further, with Laffont and the hosts debating whether the odds for trillion-dollar companies reaching $10 trillion might be even higher than 31%. Laffont's framework for why this holds is straightforward: "For every step, you have a filter that says, do you have a compounding advantage? Do you have a stronger durability of earnings? And if so, you're going to accelerate to the next phase." The caveat he acknowledged — government intervention, as with the AT&T breakup — is the variable that no compounding model fully prices.
Where the AI Revenue Is Actually Coming From
Laffont addressed directly the question that has hovered over AI investing for two years: where is the revenue? Coatue estimates the AI ecosystem at approximately $140 billion today, scaling to $300 billion this year and doubling again in 2027. The breakdown across three pillars is instructive for investors trying to size exposure.
Consumer subscriptions represent the most visible revenue stream, but Laffont flagged advertising as the underappreciated one. Coatue estimates that approximately a quarter of ads currently served by Meta and Google are AI-enabled, and the firm believes that penetration will eventually reach 100%, representing a $150 billion revenue opportunity within existing advertising infrastructure. Enterprise software — cloud coding tools and similar applications — rounds out the third pillar. On memory specifically, Coatue estimates that memory per user could quintuple as AI systems require deeper contextual knowledge to deliver personalized services, which helps explain the strong performance of memory semiconductor names.
The Power Law Risk and the Price War Question
Laffont did not avoid the uncomfortable structural questions. The power law dynamic he documented — where an increasingly small number of companies are capturing an increasingly large share of value — raises a real question for the broader venture ecosystem. He acknowledged that the number of centacorns has stagnated, and that if no new ones emerge over the next decade, it would be a warning sign for capital allocators. His honest assessment of what the data implies for LP strategy was notably blunt: a rational LP, looking at the compounding data, might simply wait for a company to hit $100 billion and concentrate there, because "it's the most sure thing, it's the least brittle, it's the least amount of effort, and it's the quickest return."
On valuations, Laffont pushed back on bubble comparisons with a distinction he considers important: "These are not fake companies. These are companies generating substantial revenue at scale that are growing faster than anything we've ever seen." He noted that Anthropic reportedly had a profitable month, and that the trillion-dollar companies that reached that threshold most recently trade at some of the lowest earnings multiples in the S&P 500 — suggesting the move reflected pent-up energy releasing rather than speculative excess.
The most candid moment of the conversation came when Laffont raised the possibility of an AI price war. Drawing the analogy to ride-sharing and food delivery, where excess capital eventually triggered destructive pricing competition, he asked openly whether OpenAI and Anthropic might eventually pull a pricing lever against each other. "Rationally, they should," he said. The counterargument is that both companies are consuming capital on infrastructure at a pace that limits their ability to fund a prolonged price war, but Laffont's point stands as a genuine risk that the current consensus does not appear to have fully priced.
Public Markets as the Final Test
Laffont closed on a theme he clearly believes in: the IPO process itself as a disciplining mechanism. "The public market is the great test equalizer," he said, adding that he welcomes the scrutiny of short sellers, analysts, and the broader market apparatus. His one tactical observation was that the traditional "day one" antiseptic of public market listing is being delayed by the mechanical reality of passive fund flows, suggesting that a realistic window for genuine price discovery on these names may be six months or more post-listing rather than immediately upon debut.
For an ecosystem that has been debating its own health for the better part of three years, Coatue's data makes a credible case that the balance is shifting. Whether the incoming liquidity wave recycles into a new generation of centacorns or simply concentrates further into the names already at the top of the stack is, Laffont acknowledged, the question that will define the next decade of technology investing.