Stablecoins grew up — and agents are the next users

On July 18, 2025, the GENIUS Act — the Guiding and Establishing National Innovation for U.S. Stablecoins Act — was signed into law. It is the first federal framework for payment stablecoins in the United States, and the most consequential digital-asset law passed to date. After years of legal ambiguity, federal law now defines who may issue a dollar stablecoin, how it must be backed, who supervises it, and what happens if an issuer fails.

The questions the Act answers are the ones I have worked on since the earliest design discussions around Libra and Diem: how a peg actually holds, who bears reserve risk, how stablecoins compete with the existing payments stack, and how public and private money should share the rails. The legal framework we built for Libra is now, ironically, the basis for the GENIUS Act. The law is, in effect, a policy answer to the economics of stablecoins.

What the GENIUS Act actually does

The Act creates a federal licensing and supervisory regime for “payment stablecoins” — digital tokens designed for payment that an issuer is obligated to redeem one-for-one for a fixed amount of monetary value. It makes it unlawful for anyone other than a “permitted payment stablecoin issuer” to issue such a token in the United States, and it clarifies that compliant payment stablecoins are neither securities nor commodities. It takes effect the earlier of 18 months after enactment or 120 days after regulators finalize implementing rules.

Crucially, the law treats stablecoins as what they economically are: privately issued claims on the dollar that compete with the payments system, not speculative securities. That framing — stablecoins as money and payments infrastructure rather than investments — is the same one my co-authors and I have argued for in peer-reviewed work since 2021.

Key provisions, in plain terms

Full 1:1 reserve backing. Every outstanding stablecoin must be backed at least one-to-one by high-quality liquid assets — cash, insured deposits, and short-dated U.S. Treasuries. Reserves must be segregated, cannot be rehypothecated, and issuers must publish monthly disclosures of reserve composition examined by third parties.

Permitted issuers only. Issuance is restricted to subsidiaries of insured depository institutions, OCC-supervised nonbank entities, and state-qualified issuers. A dual-track framework lets issuers under $10 billion opt into a certified state regime; larger issuers move to the federal regime. Non-financial public companies are generally barred from issuing without unanimous approval from a new Stablecoin Certification Review Committee.

No yield, par redemption, and holder priority. Permitted issuers cannot pay holders interest or yield, must maintain public redemption policies, and — in insolvency — stablecoin holders are prioritized ahead of other creditors. Issuers are also designated financial institutions under the Bank Secrecy Act, with AML, KYC, sanctions, and seize/freeze/burn obligations.

Why the economic design choices matter

A peg is a promise, and the GENIUS Act is fundamentally a statute about making that promise credible under stress. Mandating segregated, liquid, non-rehypothecated reserves and par redemption is the legislative version of the core finding in our research: confidence in redemption, not code, is what holds a stablecoin to a dollar. Algorithmic designs that tried to manufacture stability without full backing failed precisely because they removed that anchor — and the Act now forecloses them at the federal level.

The reserve and disclosure standards are not bureaucratic detail; they are the difference between digital dollars that expand competition and inclusion, and a system that simply concentrates power in a new set of intermediaries. Getting standards right — reserves, disclosure, interoperability, and clear rules for failure — matters more than picking a winner.

The fight GENIUS didn’t settle: yield

Under GENIUS, issuers themselves cannot pay holders yield. The live fight — now playing out in the CLARITY Act — is whether affiliated exchanges, distributors, or rewards programs can share those economics with users in ways that are functionally equivalent to interest. The banking lobby wants an “airtight prohibition” on anything that resembles interest, without carve-outs.

History suggests they should be careful what they ask for. The margarine laws of the late 1800s and Regulation Q’s 1933 ban on deposit interest were both airtight prohibitions — and both caused the substitution they were meant to prevent. Regulation Q did not protect banks from money market funds; it is the reason money market funds exist, an industry that now holds more than $7.6 trillion. If CLARITY passes with an equivalence standard strong enough to catch anything that has the effect of interest, the stablecoin industry will spend the next decade building products that are formally something else — and the deposit franchise will end up competing with whatever gets built instead.

Will stablecoins drain bank deposits?

The original fear — the one that greeted Libra in 2019 — was that instantly available digital dollars would break the banks’ hold on deposits. The evidence says otherwise: despite the meteoric rise in stablecoin market cap, empirical studies have found little sign of deposit erosion. Deposit stickiness is a powerful force; most customers value the convenience of the banking bundle too much to move their savings for a few basis points.

What stablecoins do instead is discipline the incumbents — a credible exit option that pressures banks to price deposits competitively — while overhauling the plumbing itself: atomic settlement, 24/7 availability, and cross-border liquidity that no longer sits trapped in the correspondent-banking void. The GENIUS Act is the bridge that makes this safe: by legislating full backing and enforceable redemption, it addresses the run and liquidity risks regulators rightly worried about, and converts an offshore shadow-banking anxiety into a transparent upgrade for the global dollar. The smartest banks have noticed — several are already building tokenized-deposit products on public and private blockchains rather than lobbying against the rails.

Issuance is a commodity. Distribution wins.

Stablecoins were never meant to be a profit center. Issuance grants no additional economics beyond the distribution — the balances and payments volume — you already bring to the table. That logic leaves two possible universes: one in which an endless number of stablecoins struggle for relevance, and one in which companies converge on a standard. That was the core idea behind Libra, and it now lives on with Open Standard: more than 140 companies — Visa, Mastercard, and American Express in the same consortium, alongside Stripe, BlackRock, Coinbase, Google, BNY, and a long roster of banks and fintechs — backing OUSD, a stablecoin that charges nothing to mint or redeem, returns nearly all reserve income to distributors, and is governed collectively.

The GENIUS Act solved many of the critical dimensions a standard like this needs; many more — the kind of rules the Principles for Financial Market Infrastructures exist for — still have to be worked out. The hard part is not the announcement; it is coordinating governance among parties that must collaborate first and then compete fiercely.

Corporate chains and the ghost of Libra

At the same moment the ecosystem converges on open standards, a new generation of challengers — Stripe’s Tempo, Circle’s Arc — is betting the opposite: that the future belongs to slick, branded, proprietary chains. The lesson from Libra is not about timing. Any network with a single architect is living on borrowed time: once it has a captive market, the temptation to tilt the playing field becomes irresistible. The only thing that truly separates crypto from the systems it aims to replace is that it is permissionless. If corporate chains win, the market structure will be eerily familiar — an old monarchy of card networks swapped for a new one of fintech giants.

The next users won’t be human

GENIUS regulates who issues digital dollars. The next set of design questions comes from who spends them: AI agents transacting on our behalf raise problems of spending authority, disputes, and liability that the current card-era playbook was never written for — questions closer to 17th-century maritime commerce than to consumer chargebacks. Regulated, fully backed digital dollars are the natural settlement asset for agentic payments, but the missing layers are portable identity, programmable spending constraints, and verifiable trust. That is where the stablecoin agenda meets the economics of AGI: as execution gets cheap, verification becomes the binding constraint — in payments as everywhere else.

Related research & writing

• The Economics of Stablecoins — the companion hub: how pegs hold, who bears reserve risk, and what digital dollars mean for payments
• Why An Open Standard Will Win The Stablecoin Race — Forbes, 2026
• The Banks Would Like To Dye Your Stablecoins Pink — Forbes, 2026
• How Banks Learned To Stop Worrying And Love Stablecoins — Forbes, 2025
• Stripe’s Tempo And The Ghost Of Facebook’s Libra’s Past — Forbes, 2025
• Bot Chargebacks, Voyages, and AI Liability — The Korea Herald, 2026
• Some Simple Economics of Stablecoins — Annual Review of Financial Economics, 2021 — with Alonso de Gortari and Nihar Shah
• Stablecoins and the Future of Money — Harvard Business Review, 2021 — with Jai Massari
• Setting Standards for Stablecoin Reserves — working paper, 2021 — with Nihar Shah
• Are Stablecoins Winner-Take-All? — Competition Policy International, 2024 — with Jai Massari

See all research → · See all writing →