Matt Huang

Bitcoin is a bubble, and the investors who keep saying so are, in Huang's phrase, "more right than they know."¹ He wrote that in a report prepared for institutional allocators, in the section where he was supposed to be selling them the asset.

Huang is the co-founder and managing partner of Paradigm, a crypto investment firm started in 2018 on the premise that "to invest at the frontier, you have to live on it."² Before that he was at Sequoia Capital, where he led the firm's crypto investing, and he thanks his former colleagues there for "their open-minded interest in Bitcoin circa 2014-2018" in the acknowledgments of the paper that made his name; his partner and co-founder at Paradigm is Fred Ehrsam.¹ The tension that organizes him is visible in that first sentence and never leaves: his method is to hand the skeptic the strongest version of the objection and then convert it into the mechanism of adoption. The bubble becomes the go-to-market strategy. The casino becomes the bootstrap. Sovereign hesitation becomes a race nobody can afford to sit out. The standing caveat runs in one direction, because every document in this record is a Paradigm publication written by the man whose fund benefits from the argument being believed, and each one is a recruiting document as much as an analysis.

The concession

The May 2020 paper begins where an institutional pitch normally ends, with the thing the reader already suspects. Huang defines a bubble asset as one overvalued relative to intrinsic value and then concedes the whole category: on that definition all monetary assets are bubble assets, because a store of value is demanded not for its direct utility but for its expected future value, which makes the value reflexive.¹ He gives the argument its most damaging witness, Robert Shiller on gold: "Gold is a bubble, but it's always been a bubble. It has some industrial uses, but basically it's like a fad that's lasted thousands of years."¹ The resolution is a phrase rather than a rebuttal. Money is a bubble that has not popped yet, and for a monetary asset "the utility is in the collective speculation itself."¹

What keeps this from being sophistry is the scoring that surrounds it. He treats money as a competitive market with network effects and grades Bitcoin against the classic monetary properties, scarcity, portability, fungibility, divisibility, durability, and broad acceptability, then names the one it fails: broad acceptability is "Bitcoin's primary weakness," and he measures it along two axes, the share of people who accept it and the share of wealth that does.¹ The reader is directed to the file on store of value preceding medium of exchange, which carries the sequencing law that makes the weakness a stage rather than a verdict. The paper's honest sentence is the one that follows: a better product is not enough, so the asset needs a go-to-market strategy.

Then he supplies one, and it is the bubbles. Four of them, dated and priced: roughly one dollar to thirty one and back to two in 2011; thirteen to two hundred sixty six and back to sixty five in 2013; sixty five to twelve hundred forty two and back to two hundred by January 2015; a thousand to nineteen thousand five hundred and back to thirty five hundred by December 2018.¹ Each runs the same script. High-conviction buyers accumulate while the asset is boring and unloved, price draws media, media draws shorter-horizon speculators, and the cycle repeats until demand exhausts. What survives the crash is the part he cares about: a wider base of long-term holders, more infrastructure built by exchanges and incumbents, and successively higher floors, roughly two dollars in 2011, two hundred in 2015, thirty five hundred in 2018.¹ The reader is directed to the file on the rising cycle floor, where the same pattern is stated six years later by CZ as technical analysis, arrived at independently and from the opposite end of the market.

The sizing is done against gold rather than against technology. Gold's aggregate value he puts at roughly nine trillion dollars in May 2020, split between central bank reserves, private investment, jewelry, and other forms, with Bitcoin itself then above eight thousand dollars a coin and one hundred fifty billion in aggregate; total international reserves, on the International Monetary Fund's count, reached thirteen trillion in 2019.¹ The forecast that follows is deliberately modest, which is what makes it checkable. Bitcoin was unlikely to challenge the dollar as a means of exchange or unit of account any time soon; it was likely to earn a place alongside gold in portfolios, and eventually in central bank holdings.

The reading list

The same month he published a second, stranger document: seven things to read about Bitcoin, aimed at the identical audience.³ It is a curator's move rather than an author's. Instead of arguing again, he hands the allocator seven other people's arguments, most of them from investors with reputations at stake: Wences Casares on the family in Patagonia that lost its savings three times, to devaluation, then hyperinflation, then outright confiscation of bank deposits; Paul Tudor Jones announcing a Bitcoin position inside a letter forecasting a Great Monetary Inflation; Bill Miller writing as a value investor; Murray Stahl on the erosion of purchasing power; John Pfeffer, formerly of KKR, concluding that most cryptoassets will not repay their risk and that the one exception is a dominant non-sovereign store of value.³

Two details in the selection are worth more than the list. The seventh entry, Vijay Boyapati's essay, supplies the four-stage law that the archive files separately, collectible to store of value to medium of exchange to unit of account, drawn from William Stanley Jevons's account of gold, and it is the intellectual spine of Huang's own position.³ The third entry is the one he disagrees with. Marc Andreessen's 2014 essay is the payments-first case, and the paper Huang published in the same month declines to endorse it, holding that Bitcoin was not positioned to become a means of exchange first.¹ He included it anyway. A man selling a thesis put the best-known argument against his own sequencing at number three on the list.

The casino

By September 2023 the audience had changed and the concession had to get larger, because the objection was no longer that Bitcoin was a bubble but that crypto was a scam. "The Casino on Mars" opens with the metaphor that gives it its title: crypto as a new planet being settled, whose early arrivals are a mixed bunch of explorers, researchers, ordinary people, and "speculators, some rough and disreputable."⁴ He does not dispute the casino. He argues it is load-bearing. The 1849 gold rush turned a village into the port that eventually became the center of technology, and today's speculative frenzy is what attracts settlers and pays for infrastructure on barren ground.⁴

Underneath the metaphor is his sharpest single formulation: "speculation is the 'hello world' of digital property rights."⁴ Give people the ability to create scarce assets and they will trade them, the way children trade cards, because the entire point of a property-rights system is to record transfers and the obvious first thing to do with one is to test it. He pairs that with the argument that the killer application already exists and that comfortable readers cannot see it: "Like the fish who asks, 'What the hell is water,'" crypto is hard to appreciate where property rights and monetary stability are taken for granted, and it is adopted differentially in Argentina, Turkey, and Ukraine.⁴ The reader is directed to the file on crypto as a museum, where Justin Sun defends the same activity as an exhibition worth wandering and Marty Bent answers that the mission is too serious for one; Huang's frontier reading is the only one of the three that makes the speculation productive rather than merely tolerable.

This is also the one essay where the concession is not immediately converted. A section titled the casino's dark side grants that too much speculation and airdrop farming add noise to the price signal that should be directing capital, that short-term speculation is zero-sum and can burn newcomers permanently, and that the bad actors are real: "Imagine a roving gang of bandits who greet newcomers with a beatdown and a mugging," he writes, "welcome to San Francisco crypto!"⁴ The remedy he offers is one clause long, that some self-regulation or regulation may be needed, and he leaves it there. He also puts himself in the frame once, which he almost never does: to think in the early days that Bitcoin would reach the legitimacy it now has seemed crazy, "I remember because I was there."⁴

An appendix aimed at policymakers contains the line that has aged into the archive's file on reserve currency succession: counterintuitively, crypto might end up being a boon to the dollar, because dollar stablecoins are the most popular currency on the new planet by a wide margin.⁴ It is a genuine cost to his own case. The scarce asset he has spent five years arguing for gets a smaller, non-sovereign role, and the incumbent he was implicitly betting against gets extended reach.

Gunpowder, not the iPhone

The sharpest of the four essays is also the shortest. In November 2024 he published fewer than three hundred words, a fair share of them links, on why the game theory of sovereign adoption had changed.⁵ Paradigm's own 2018 thesis, formed when Bitcoin was around four thousand dollars a coin, had treated adoption by nation states as improbable; the whole essay is a single reclassification of that judgment, from improbable to underpriced.⁵

The evidence is stacked rather than argued: El Salvador's holdings above five hundred million dollars, Abu Dhabi's sovereign wealth fund engaged in mining, a president-elect discussing a strategic reserve on stage at a Bitcoin conference, Senator Cynthia Lummis's draft legislation, and a prediction market pricing roughly a thirty percent chance of a United States reserve inside the first hundred days.⁵ Then the turn, credited to Tyler Cowen's instruction to solve for the equilibrium: "Sovereigns can no longer afford to dismiss BTC. From a game theoretic perspective, BTC is like gunpowder, not the iPhone."⁵ Christopher Nolan can decline the iPhone and may even benefit from declining it, but once gunpowder existed no sovereign could opt out. He closes on one sentence: "The race to build BTC reserves is on."⁵

Note what the word "underpriced" is doing. It is not a claim that reserves will be built, it is a claim that the market is mispricing the odds that they will, which is the only claim an investor can be held to. The reader is directed to the file on the strategic Bitcoin reserve for what the scoreboard did next, and the result is mixed rather than a vindication.

The turn

For five years his entire case rested on the same load-bearing assertion: that the unserious surface of crypto was paying for a serious system underneath that nobody could point to yet. In September 2025 he went to build the part nobody could point to.⁶ Tempo is a payments-first blockchain incubated jointly by Stripe and Paradigm, and his diagnosis of why it was needed is an admission about the ground he had been defending: today's crypto stack caters, explicitly or implicitly, to trading, and is comparatively underoptimized for payments.⁶ The casino built the wrong thing well.

What Tempo is for is conspicuously unglamorous next to a decade of monetary theory: global payments and payroll, remittances, tokenized deposits for round-the-clock settlement, embedded financial accounts, microtransactions, agentic payments.⁶ The design input he names is the tell about who the customer is, since the list is almost entirely incumbents and large platforms rather than crypto-native builders: Anthropic, Coupang, Deutsche Bank, DoorDash, Lead Bank, Mercury, Nubank, OpenAI, Revolut, Shopify, Standard Chartered, and Visa.⁶ The reader is directed to the file on the payment rails leapfrog, where the same move is made by a central bank instead of a firm, which is the comparison that shows what a private settlement rail is actually competing with.

The staffing decision is the part that makes this a turn rather than another investment. He took the operating job himself, saying he would lead Tempo while continuing to lead Paradigm alongside a colleague he names only as Alana, and he flags the incubation as an exception rather than a model, since Paradigm expects such opportunities to be rare.⁶ The essayist who spent five years arguing that speculation was subsidizing infrastructure went and put his own hours into the infrastructure.

Where things stand

In July 2026 Paradigm announced a fourth fund of 1.2 billion dollars and a mandate visibly wider than the one it started with: still crypto, but now also AI, robotics, and other frontiers, with drone delivery, rapid manufacturing, orbital defense, open-weight AI research, prediction markets, and Tempo named as the shape of the portfolio.² The method is stated as a claim about proximity rather than about selection, researching and building alongside founders rather than only writing checks, which is why the firm ships open-source tooling as well as capital.² The register has changed with the mandate, from monetary history to something closer to a forecast: "Sufficiently steep exponentials are indistinguishable from magic," and the era favors those willing to throw out existing playbooks and recompute their view of reality frequently.²

The objection to the whole body of work is the one his own method invites. A writer who consistently concedes the strongest charge and then shows why the charge is the mechanism has built an argument that cannot lose: the bubble is adoption, the casino is bootstrapping, the hesitation is a race. The single place he stops short is the casino's dark side, where the noise in the price signal and the newcomers burned permanently get named and then get no conversion at all, only a sentence allowing that some regulation may be needed. That paragraph is the best evidence that the rest is analysis rather than rhetoric, and it is one paragraph in five years of writing.

The classification here is provisional. Huang is filed under Finance and Technology Builder because the plate's definition describes the two halves of his record rather than his sector: he builds the capital layer and the infrastructure layer beneath products other people will ship, and he takes his position inside a paradigm before the thing that would prove it has arrived, which is what a 2018 fund raised against a four thousand dollar asset and a 2025 settlement chain built for the enterprises he says have not come onchain yet both are. The plate is contradicted in one place worth naming. Its anchor buys roads and waits, whereas Huang's distinctive instrument is an argument: the essays are the product, and the fund's reputation, deal flow, and recruiting run through them. Whether persuasion at that scale is a variety of road-building or a different plate entirely is open, and the archive does not settle it here.

Key facts

• Co-founder and managing partner of Paradigm, started in 2018, and previously at Sequoia Capital, where he led crypto investing.
• Wrote four essays between 2020 and 2024 that became the institutional case for Bitcoin as a store of value rather than as a payments network or a technology platform.
• Concedes in the 2020 paper that all monetary assets are bubble assets, and argues that Bitcoin's bubbles are its go-to-market strategy rather than a defect.
• Grades Bitcoin against the classic monetary properties and names broad acceptability as its only real weakness.
• Documented four Bitcoin bubbles and the successively higher floors they left behind, roughly two dollars in 2011, two hundred in 2015, and thirty five hundred in 2018.
• Published a seven-item reading list in the same month, including Marc Andreessen's payments-first essay, which his own companion paper declines to endorse.
• Argues in 2023 that speculation is "the 'hello world' of digital property rights," while conceding that excess speculation adds noise to the price signal and that short-term trading is zero-sum.
• Reclassified sovereign Bitcoin adoption in November 2024 from improbable, Paradigm's 2018 view, to underpriced, on the argument that Bitcoin is "like gunpowder, not the iPhone."
• Took the operating job at Tempo, the payments-first blockchain incubated by Stripe and Paradigm, in September 2025, while continuing to lead Paradigm.
• Announced a 1.2 billion dollar fourth fund in July 2026 with a mandate widened from crypto alone to AI, robotics, and other frontiers.