Stablecoins Are a $3 Trillion Treasury Trade Wearing a Crypto Costume

Stablecoins are a $3 trillion Treasury demand engine wearing a crypto costume, and that's the only pitch Washington should need to hear. Forget the ideology. Forget the bitcoin maximalist bravado. The dollar token business is, at its core, a distribution channel for U.S. government debt, and it's scaling faster than any diplomatic effort to sell Treasuries abroad.

Sam Lyman puts it in one image that's hard to shake. Dollar stablecoins are Starlink for the dollar system. If you've got an internet connection, you've got the dollar. That's not a crypto talking point, that's a monetary policy fact with a balance sheet attached to it.

The Mechanics Nobody Bothers to Explain

Here's what actually happens when someone in Lagos, Buenos Aires, or Istanbul buys a dollar stablecoin. The issuer takes that cash and parks it in short-dated Treasuries. So the buyer, whether they know it or not, is holding a claim on a T-bill. Lyman's framing is blunt and correct. These buyers are, in effect, purchasing U.S. debt.

Now scale it. Estimates put the U.S. stablecoin market above $3 trillion by the end of the decade. Read that as a funding forecast rather than a crypto forecast. That's $3 trillion of fresh demand for the paper the Treasury auctions every week, and it arrives without a single foreign central bank having to return a phone call.

In traditional markets, this would be called a captive buyer. Foreign official holders of Treasuries sit somewhere in the high single-digit trillions, and their appetite moves with geopolitics, sanctions risk, and reserve diversification politics. Stablecoin issuers don't have those problems. Their demand is mechanical. More tokens in circulation means more bills on the books. The comparable in TradFi is an index fund that has to buy the underlying no matter what the price does.

That's the whole thesis, and it's a strong one. The dollar's reserve status has always depended on access. Stablecoins turn access into a software product.

The Gateway Drug Line Is the Weak Link

Here's where I push back. Lyman calls stablecoins, in many ways, a gateway drug to Bitcoin. It's a nice line. The data has never really supported it.

Stablecoin holders tend to stay stablecoin holders. People who want dollar stability in an unstable currency regime aren't shopping for volatility, they're escaping it. Look at transfer volumes. Dollar-denominated stablecoin settlement dwarfs Bitcoin's on-chain dollar volume by a wide margin, and that gap has widened, not narrowed, through every bull market since 2020. The on-ramp exists. Not everybody walks up it.

But here's the thing. The gateway drug claim doesn't need to be true for the Treasury argument to work. It's a bonus, not a load-bearing wall. And the people pushing back on the whole thesis tend to attack the wrong target.

So what's the real bear case? Three things. First, reserve composition rules. If regulation pushes issuers toward overnight repo, bank deposits, or money market funds instead of bills, the Treasury demand story deflates quietly, and nobody notices until the auctions get soft. Second, competitive supply. If T-bill issuance shrinks or yields compress toward zero, the economics of running a stablecoin business get thin, and the float stops growing. Third, policy drag. Lyman's own read is that big crypto legislation stalls until after 2028. That's two more years of the same circular arguments in Washington.

And there's a fourth risk that gets almost no attention. The same political class that's skeptical of crypto is also skeptical of data centers. A moratorium on AI data center buildout doesn't just hit hyperscalers, it hits the hybrid miners who've been converting power contracts into compute revenue. Lyman floats a data center dividend of up to $9,000 a year per household in some markets. Whether that number holds up or not, the political temptation to grab the revenue and ban the infrastructure is real. Crypto keeps learning the same lesson. Access is only durable when the policy is.

Where the Trade Actually Sits

Strip away the jargon and it's a credit product. Stablecoin issuers are running a short-duration carry trade funded by retail users who don't want the duration. The Sharpe ratio tells a sobering story, because the issuer captures the yield and the holder gets a token that's supposed to be worth a dollar. That spread is the entire business model, and it's the same spread that makes the Treasury demand so reliable.

So which side do I land on? I'm long the Treasury demand thesis, and I'm short the gateway drug thesis, and I don't think those positions conflict. The dollar access story is structural. It's tied to inflation in emerging markets, remittance corridors, and cross-border B2B settlement, none of which care about Bitcoin's halving cycle. The adoption ladder story is cultural, and culture is fickle.

Winners here are clear enough. Treasury gets a new buyer class that grows with smartphone penetration. Stablecoin issuers get a spread business with sovereign-grade collateral. Dollar users in weak-currency economies get optionality they've never had. Losers are the correspondent banking networks collecting fees on wires that now look absurd, and any government that thought capital controls were enforceable in a world where a token can hold a T-bill.

Can those three interests actually stay aligned for a decade? That's the question nobody in the room wants to answer directly. History says alignment on the upside, friction on the downside.

Agents Don't Hold Dollars

One more thread worth pulling. Lyman makes the point that when software agents start saving, they want Bitcoin. It sounds like science fiction until you think about the plumbing. An autonomous agent doing thousands of microtransactions a day needs a settlement rail, and that rail will be a dollar token. But an agent holding a reserve balance for a year isn't going to sit in something yielding nothing against 3% inflation. Crypto is pricing in what equities haven't, which is a world where machine treasuries behave a lot like corporate treasuries, with a risk asset on the balance sheet and a stablecoin for working capital.

That's years out. Fine. But it's the direction, and it makes the stablecoin buildout look less like a payments feature and more like a monetary base layer.

Watch one thing over the next few quarters. Look at the reserve breakdown in the next round of stablecoin attestations. If the bill share climbs, the $3 trillion number is real Treasury demand. If it slides toward repo and deposits, then what we've got is money market fund assets with extra steps and a marketing budget. That distinction is worth more than any keynote.