Licensed to Fill – How Crypto Took Over the Plumbing of Finance

• Paxos became the first blockchain-native firm registered as an SEC clearing agency in May 2026, entering the DTCC’s regulatory category while the DTCC itself prepares to tokenise Russell 1000 equities on Stellar in 2027, so the convergence now runs in both directions.
• The GENIUS Act turned stablecoin issuers into narrow banks in all but name, and BIS research shows the effect on public debt markets, with each $1 of USDC growth producing roughly $0.75 of T-bill purchases over four months.
• DeFi spent 2026 relearning banking’s oldest lessons, with TVL down 39%, $942 million lost to 121 exploits, and a $15 billion four-day run on Aave that replicated a bank panic without a lender of last resort.
• The industry’s defining conflict is no longer crypto versus the state but the crypto lobby versus the bank lobby, fighting over stablecoin yield in the Senate the way securities firms and banks have fought since Glass-Steagall.
• What genuinely survived of the original design is the settlement layer, since atomic DvP, bearer-style transfer, and programmability drive every serious institutional project, while pseudonymity and decentralised governance fade.

On 28 May 2026, the SEC granted Paxos Securities Settlement Company registration as a clearing agency under Section 17A of the Securities Exchange Act, making a stablecoin issuer’s subsidiary the first new entrant into US securities clearing since the DTCC consolidated the field. The approval took seven years, a 2019 no-action letter, and a settlement pilot run under SEC supervision. It arrived while Bitcoin traded more than 50% below its October 2025 peak, DeFi shed capital for the sixth consecutive month, and the industry’s own lobbyists fought bank lobbyists over yield rules in the Senate.

That timing carries the whole argument. Crypto’s absorption into traditional finance did not pause for the bear market because it was never a function of price. The licenses, charters, registrations, and rulemakings kept moving while the speculative layer halved, and the scoreboard of 2026 wins reads like a list of everything the industry was invented to route around. A sector that spent a decade selling the exit from traditional finance has spent this cycle celebrating admission into it, and almost nobody inside seems to find that strange.

The plumbing is merging from both ends

Paxos makes the cleanest example because clearing and settlement is the least glamorous, most consequential layer of any market. A clearing agency registration does not let a firm issue tokens or run an exchange. It lets the firm stand between both sides of a securities trade and finalise it, which in the US has effectively been the DTCC’s job since Dodd-Frank designated it a systemically important utility in 2012. Paxos can now do that job for eligible US equities on blockchain rails, with quarterly reporting, stress testing, and the rest of the supervisory apparatus attached. The pilot behind the registration settled real equities daily with large global firms from 2020, so the SEC granted it on years of performance data (and still made it temporary, because regulators do not hand out permanence).

In crypto, one bad click is final. Learn to verify before you lose.

Traffic runs the other way too. The DTCC itself announced a partnership with the Stellar Development Foundation to tokenise DTC-custodied assets, including Russell 1000 equities and US Treasuries, on Stellar’s public blockchain in the first half of 2027. The incumbent monopolist of American post-trade infrastructure is putting the assets it custodies onto a public chain while a blockchain-native firm earns the right to compete with it under the same statute. Convergence was supposed to mean crypto growing up into finance, and in practice it also means finance climbing down onto crypto’s rails, with the two motions now hard to tell apart.

JPMorgan has run the same arc in miniature. Its blockchain unit spent years inside a private, permissioned network, then in May 2025 executed a cross-chain delivery-versus-payment test that settled Ondo’s tokenised US Treasuries against dollar deposits at JPMorgan, with Chainlink coordinating between the bank’s permissioned Kinexys network and a public chain. By December 2025 the bank had gone further and launched its own tokenised money market fund on Ethereum, seeded with $100 million of its own capital. The largest US bank now issues a yield-bearing fund on the same public chain that hosts Uniswap, which would have been a compliance officer’s resignation letter five years ago.

BlackRock supplies the asset management leg. Its tokenised Treasury fund BUIDL, launched in March 2024, sits at roughly $2.5 billion in assets and is accepted as collateral across major trading venues. In May 2026 the firm filed with the SEC for a new blockchain-native stablecoin reserve fund and for onchain shares of a $7 billion conventional money market fund, with BNY Mellon maintaining the official shareholder registry on Ethereum as ERC-20 tokens. An ownership record that constitutes the legal register of a mainstream BlackRock fund would live on a public blockchain, maintained by a custodian bank older than the republic. The tokenized real-world asset market these products anchor has grown to roughly $26 to 30 billion, small against global fund assets but tripling year on year while nearly everything speculative in crypto shrank.

The distribution layer converged earlier and faster. Spot Bitcoin ETFs, approved in January 2024, moved crypto exposure into brokerage accounts and model portfolios, and by the end of 2025 BlackRock alone managed close to $80 billion in digital asset exchange-traded products while backing more than $65 billion in stablecoin reserves through its money market vehicles. The same wrapper works in both directions, which the first half of 2026 demonstrated less comfortably, as ETF holders pulled several billion dollars during the drawdown and transmitted TradFi-style redemption pressure straight into crypto’s order books. Institutional access was sold as a stabiliser. It behaves like what it is, another channel through which capital arrives quickly and leaves quickly, priced by people who treat Bitcoin as one line in a risk budget.

Stablecoins got the bank treatment because they behave like banks

The GENIUS Act, signed in July 2025 as the first US federal crypto statute, settles the definitional argument by refusing to have it. Permitted issuers must be bank subsidiaries or federally or state-qualified nonbanks, reserves must be held one-to-one in dollars and low-risk assets, redemption rights must be enforceable, and compliant stablecoins are explicitly neither securities nor commodities. Strip the vocabulary and the law describes a narrow bank, an institution that takes dollar liabilities and holds only safe short-term assets against them. Economists have proposed narrow banks for ninety years and regulators have refused to charter them, so it took a crypto instrument with a nine-figure lobbying budget to get one through Congress. The implementing regulations are due by 18 July 2026, covering issuer licensing, capital, custody, and anti-money-laundering standards, which means the operating detail of American stablecoin law lands within days of this piece publishing.

Total stablecoin supply sits around $315 billion as of June 2026, with Tether’s USDT at roughly $186 billion and Circle’s USDC near $75 billion, the two together controlling 83% of a market that lists 382 coins but behaves like a two-name oligopoly with an ornamental tail, and scale on that order justifies the bank treatment. Those reserves stopped being an abstraction for the Treasury market some time ago. A BIS working paper published this year estimates that a $1 increase in USDC’s market capitalisation produces roughly $0.75 of T-bill purchases cumulated over four months, placing stablecoin issuers alongside government money market funds and foreign official holders as a measurable bid for American short-term debt. An instrument invented as an exit from sovereign money has become one of the more dependable buyers of sovereign debt.

Ethereum holds roughly 60% of global stablecoin liquidity, around $170 billion as of April 2026, while Tron carries about $87 billion of which more than 97% is USDT, the corridor chain for remittances and dollar savings across Asia, Africa, and Latin America, so the geography underneath the supply is as concentrated as the issuers above it. Two issuers on two chains now constitute the effective dollar clearing system for a meaningful slice of the emerging world, an arrangement no committee designed and no regulator fully reaches. The GENIUS framework governs the onshore issuer well and the offshore giant barely at all, which is the same perimeter problem eurodollar markets posed in the 1960s, replayed on faster rails.

Worth noticing is what the United States chose not to build. The same July 2025 legislative push that produced GENIUS saw the House pass a bill to block a retail central bank digital currency, so American policy now runs on regulated private dollar tokens where China runs on the state-issued e-CNY. Both countries concluded that digital money needed a sovereign anchor, and they anchored it at opposite ends, one in supervised private balance sheets holding T-bills and one in the central bank itself. The stablecoin, dismissed for years as crypto’s plumbing fluid, turned out to be the instrument through which the dollar system extends itself onchain without the Fed writing a line of code.

Regulation is already redistributing the market. Visa’s onchain analytics show USDC took roughly 67% of adjusted stablecoin transaction volume in June 2026, a record $1.79 trillion month in which the compliant, US-domiciled issuer ran away with the settlement flow while USDT kept the larger supply. Tether still dominates balances, particularly across Tron’s remittance corridors in Asia, but the transactional economy that banks, card networks, and fintechs plug into is consolidating around the issuer that looks most like a regulated financial institution. Visa’s own stablecoin settlement pilot reached a $7 billion annualised run rate by April 2026, growing 50% quarter on quarter. Payments executives adopted the instrument because a bearer token that settles in seconds beats correspondent banking on cost and hours, and because the GENIUS Act made the counterparty risk legible enough for a compliance committee to sign off.

New Scam Prevention Course. Never make a rushed crypto decision again.

DeFi is relearning the banking syllabus the hard way

While the regulated layer compounded, the part of crypto that was supposed to replace banks spent 2026 demonstrating why banking regulation exists. DeFi total value locked has fallen every single month this year, from about $115 billion in January to roughly $71 billion by mid-June, a 39% drawdown that tracks the broader market but was amplified by the sector’s own failures. The year has produced 121 exploits and about $942 million in losses, with the second quarter alone counting 85 incidents, the busiest quarter on record by count. Two April attacks, the $295 million Drift Protocol breach and the $293 million KelpDAO exploit, accounted for more than half of the annual total between them.

The KelpDAO aftermath deserves attention from anyone who believes code replaced trust. Aave, one of the most established lending protocols in existence, watched users withdraw about $15 billion in deposits within four days of the exploit, and its TVL collapsed from $26.4 billion to $14.3 billion. Aave itself was never breached, but its depositors repriced trust across the whole category at speed, pulling liquidity from a solvent protocol because a neighbouring one had failed. Financial historians have a name for that dynamic, and the regulated side of crypto already met it in March 2023 when Silicon Valley Bank’s collapse briefly depegged USDC over $3.3 billion of trapped reserves. DeFi rebuilt the bank run with better settlement and no deposit insurance, no discount window, and no lender of last resort, so the runs resolve in days instead of quarters and the losses stay where they land.

None of this makes DeFi worthless, and the honest reading of the TVL decline is mixed. Some of the exit reflects recursive leverage unwinding, circular liquidity that inflated the metric on the way up and deflates it on the way down without touching real users, and stablecoin supply held near record levels through the same period, which suggests parked capital more than departed capital. The one segment attracting inflows against the trend was the one carrying regulated instruments, with tokenised Treasuries growing to an estimated $11 billion by May 2026 inside the broader $26 billion real-world asset total, a category that for the first time overtook the value locked on decentralised exchanges. The protocols that survive this reset will be the ones with fee revenue, audited risk, and insurance arrangements, which is to say the ones that internalised the boring disciplines of the industry they set out to obsolete.

The turf war tells it all

You can measure how traditional crypto has become by looking at who its enemies are now. The fight over the CLARITY Act, the market structure bill that cleared the House in July 2025 and advanced through Senate Banking on a 15-9 vote in May 2026, is no longer a fight between crypto and the state. It is a fight between the bank lobby and the crypto lobby over deposit economics, conducted through comment letters, coalition memos, and last-minute amendment pushes, which is exactly how the securities industry and the banking industry have fought each other since Glass-Steagall.

Yield is the specific battleground, because GENIUS prohibits stablecoin issuers from paying interest on balances while exchanges sitting next to issuers can pass rewards through, and Coinbase’s arrangement with Circle does precisely that. Banks call it a loophole that threatens their deposit base and want the Senate to close it. The crypto side answers with research arguing that most stablecoin growth originates offshore and imports foreign capital into US banking infrastructure at a rate that outweighs any domestic deposit migration. Brian Armstrong pulled Coinbase’s support for the Senate Banking draft in January 2026 and the committee postponed its markup the next day, a move indistinguishable from any incumbent flexing on a bill that touches its margins. What remains between the bill and the president’s desk is a conflict-of-interest provision covering officials’ crypto ties, a merge with the Senate Agriculture Committee’s version, and the sixty floor votes that will need a meaningful number of Democrats, with the November midterms deciding how long the window stays open.

Europe ran the same convergence with less theatre. MiCA has been in application since 2024, and its effect was to sort issuers and service providers into the familiar categories of authorisation, reserve requirements, and conduct supervision, treating crypto firms as financial institutions with unusual technology rather than a new species. The firms that complained loudest during the drafting now market their MiCA licenses as a competitive moat, the way banks market their charters, and passporting across twenty-seven member states turned the compliance cost into a distribution asset. Regulatory arbitrage still exists at the margins, and the offshore share of Tether’s business shows how much activity a strict perimeter simply pushes elsewhere, but the direction across the US, EU, UK, Singapore, and the UAE points the same way, toward function-based rules that make a crypto lender answer the same questions as any lender.

Fireblocks, which secures transfers for more than 2,400 financial institutions, spent the past year buying Dynamic for its embedded wallet stack and then the accounting platform TRES for around $130 million, assembling what it markets as a complete custody-to-consumer operating system for banks and fintechs. That is a fintech consolidation playbook, executed with fintech language, aimed at bank procurement departments, and it tells the same story at the corporate layer that CLARITY tells at the legislative one. The infrastructure companies of crypto stopped behaving like protocols years ago and now behave like the enterprise software vendors of financial services, because that is who signs the contracts.

What survives of the original design

Calling crypto the new traditional finance risks flattering both sides, so it helps to be precise about what converged and what did not. The functions converged, since custody, clearing, issuance, lending, and asset management now exist in blockchain-native form under recognisable regulation. The risks converged too, as the runs, the two-issuer concentration, and the T-bill sensitivity all demonstrate, and a supervisor stress-testing a stablecoin in 2026 is asking the same questions a money market fund examiner asked in 2008. What did not converge is the settlement layer itself, and that layer is doing the work.

Atomic delivery versus payment is a genuine improvement over a T+1 batch cycle, and the capital freed by instant settlement is real money to anyone running a large book. Bearer-style instruments that move at any hour without a correspondent chain are a genuine improvement for cross-border payments, which is why the volume growth is in stablecoins and tokenised cash and not in anything with a volatile price. Programmability, the ability to make the ownership record itself execute logic, is the piece traditional infrastructure cannot replicate by working weekends. Those three properties explain every serious institutional project on the list above, from Kinexys to the DTCC’s Stellar plan, and none of them requires a token economy, a governance forum, or an ideology.

The future of finance is here. Trade crypto and more

Convergence has not answered the questions supervision exists to answer, and it would be dishonest to score 2026 as a clean win for stability. No dollar stablecoin has yet processed a genuine crisis-scale redemption wave at par, since the USDC depeg resolved because a federal backstop rescued the bank holding the reserves, an intervention nobody should count on twice. GENIUS takes full effect only after its rulemakings finish, leaving a transition window in which the largest issuer by supply operates offshore and the enforcement perimeter is aspirational. And the concentration that makes the system efficient in calm markets, two issuers, two chains, a handful of custody vendors, is precisely the topology that turns an operational failure into a systemic one. Traditional finance learned those lessons through episodes with names and dates. Crypto has adopted the license structure much faster than it has accumulated the loss history that taught banks their caution.

The parts of the original design that are fading are the ones that defined the culture. Pseudonymity is incompatible with the travel rule and dying wherever regulated money touches the chain. Decentralised governance survives mostly as theatre above increasingly professional operating companies. The dream of a parallel system keeps losing votes in the only election that counts, since the capital flows of 2026 run through ETFs, tokenised funds, and licensed issuers rather than through anything permissionless. A total market cap of about $2.2 trillion still carries a large speculative core, and Bitcoin trading at half its October 2025 peak above $126,000 shows speculation has not left the asset class. The infrastructure build no longer depends on it.

The convergence thesis, stated plainly, is that crypto won distribution by surrendering structure. It got the ETFs, the charters, the clearing registration, and the federal statute, and the price was becoming legible to the system it was designed to escape. Whether that counts as victory depends on what you thought the project was. If the project was better financial plumbing, it is winning through a bear market, which is the strongest form of winning available. If the project was an alternative to finance itself, that version now lives mainly in the conference talks, and the money stopped attending some time ago.