Congress Banned Stablecoin Interest. Your Wallet Pays 4% Anyway.

The most heavily lobbied sentence in American crypto law — the one banning stablecoin issuers from paying interest — has been fully operational for months. It is also, as a practical matter, already beaten.

On June 30, Consensys launched MetaMask Money Account, a self-custody savings-and-spending product that pays up to 4% variable APY on balances of mUSD, MetaMask's dollar-pegged stablecoin. The same balance is spendable at hundreds of millions of Mastercard merchants through the MetaMask Card. Deposit dollars, earn 4%, tap to pay. That is a description of a very good checking account — from a company that is not a bank, on money the law says cannot bear interest.

The trick is that nobody involved is technically paying interest on a stablecoin. And the gap between "technically" and "actually" is where the next two years of stablecoin policy will be fought.

How the machine works

The GENIUS Act prohibits payment stablecoin issuers from paying yield to holders. Congress wrote that ban for a reason banks lobbied hard to keep visible: a stablecoin that pays interest is a deposit substitute, and deposit substitutes drain funding from the banking system that makes loans.

MetaMask's structure routes around the ban in three moves. First, mUSD's reserves — cash and short-term Treasuries held by Bridge, a Stripe-affiliated entity — generate yield that stays with the issuer side, exactly as the law demands. Second, when a user opts in, their balance is routed automatically into Morpho, a decentralized lending protocol, with Aave integration planned. The 4% comes from borrowers on Morpho paying to borrow, not from the issuer paying holders. Third, the whole arrangement lives inside self-custody software — and the GENIUS Act carves out transactions through self-custody wallets from much of its reach.

The rate is honest about what it is: variable and demand-driven. When borrowing demand on Morpho runs hot, the yield approaches 4%; when it cools, the rate sags with it. This is not a teaser rate backstopped by a marketing budget. It's a live feed from a credit market.

But step back from the mechanics and look at the product. A consumer deposits dollars, holds a dollar-denominated balance, earns a return, and spends the balance with a card. Every functional element of an interest-bearing checking account is present. The only thing missing is the word.

Congress left the door open on purpose — or by accident

The GENIUS Act never defined "holder," and it said nothing at all about affiliates or third parties offering rewards tied to a stablecoin. Whether that was drafting fatigue or deliberate ambiguity depends on which lobbyist you ask. Either way, the industry noticed immediately. Forbes called the resulting gap a Coinbase-shaped hole — because Coinbase, which issues nothing, pays USDC rewards to its customers as a distribution partner, not an issuer.

The regulators are now trying to close what Congress left open. The OCC's proposed GENIUS Act rules include a rebuttable presumption that any coordinated arrangement between an issuer and an affiliate or related third party to pay holders yield is itself a prohibited yield arrangement, as Perkins Coie's analysis lays out. The Bank Policy Institute, the large banks' lobby, has been pushing to slam the loophole shut entirely. And the CLARITY Act's drafters have taken their own run at fixing it.

Notice what MetaMask's design does to all of that. The OCC's presumption targets coordination between issuers and affiliates. But Morpho is neither. It's an autonomous lending protocol that neither Consensys nor Bridge controls, paying rates set by open-market borrowing demand. The yield isn't a disguised transfer from reserve income; reserve income stays put. If the presumption is rebuttable, this is what a rebuttal looks like with engineering behind it.

The banks were right about the substance

Here's the uncomfortable part for crypto partisans: the banking lobby's underlying complaint is correct. Money will move toward the higher yield. The Brookings Institution predicted precisely this dynamic — a prohibition on interest creates strong competitive incentives to circumvent the prohibition, the same way corporate treasurers already sweep idle balances into money market funds every night. Prohibitions on paying for deposits have a long, failed history; Regulation Q spent four decades teaching that lesson before it was repealed.

So the question was never whether the yield ban would leak. It was who would build the leak, and whether the plumbing would be sound.

That second part deserves more scrutiny than the launch coverage gave it. A Money Account balance earning 4% is not a deposit at an insured bank, and it is not a claim on Treasuries sitting quietly in reserve. It is exposure to Morpho's lending markets — smart-contract risk, oracle risk, borrower-collateral risk — wrapped in a savings-account interface. The distinction between "backed by T-bills" (the idle mUSD) and "lent into DeFi" (the yielding mUSD) collapses into a single number on a phone screen. When a lending market seizes up someday — one will, eventually — a great many users will discover which side of that distinction their balance was on.

Self-custody softens some failure modes and sharpens others. Nobody can freeze the account, and no bank failure takes it down. But no deposit insurance catches it either, and the product is unavailable in the UK precisely because British regulators have not made peace with retail products shaped like this.

The wallet is the new branch

The larger shift is competitive, not legal. The wallet — the piece of software closest to the user — is claiming the economics that used to belong to banks and exchanges. MetaMask now offers yield, card spending, and trading from one balance. The Open USD consortium of 140-plus payment and banking firms launched its own stablecoin days earlier with reserve yield shared among partners rather than users. Everyone in the stack is fighting over the same pool of reserve income, and the consumer-facing endpoints are winning because they can convert it into things users feel — rates, cashback, zero-fee redemption.

For users, the practical problem is now fragmentation: a yield-bearing balance here, an exchange balance there, a hardware wallet somewhere else, each with a different rate and a different risk profile. Keeping one coherent view of all of it is exactly the job mobile portfolio trackers like The Crypto App exist to do — the dashboard layer matters more, not less, when every app on your phone is trying to become your bank.

The Metal tier of MetaMask's card pays 3% cashback on the first $10,000 of annual spend for a $199 fee. Run the numbers on 4% yield plus 3% cashback and compare them to the 0.4% the average American savings account pays. That spread is not sustainable forever. It doesn't have to be. It only has to last long enough to move the deposits.

What breaks first: the loophole or the banks' patience

Three outcomes are plausible from here. The OCC finalizes its rules broadly enough to catch protocol-routed yield, and products like Money Account geofence the US the way they've geofenced Britain. Or Congress rewrites the ban through CLARITY and discovers, as it did with GENIUS, that defining "yield" precisely enough to stop engineers is harder than passing a bill. Or — the path Regulation Q suggests — the ban erodes into irrelevance and eventually gets repealed, at which point stablecoin issuers pay interest openly and the American checking account gets repriced for the first time in a generation.

Every one of those paths ends with consumers earning more on their dollars than banks currently pay. The only question the yield ban ever really decided was who captures the spread in the meantime.

The banks spent their political capital banning a word. The engineers shipped the thing itself, wrapped in a wallet, eleven months later. Next time the two sides fight over a sentence in a stablecoin bill, watch the commit logs instead.