Stablecoins: The Dollar on the Blockchain

:::section{mins=3} What Is a Stablecoin :::

:::lead Bitcoin and ETH never promised to be worth tomorrow what they are worth today. A stablecoin promises exactly that. As of September 2026 about $305 billion of these tokens are in circulation, and not one of them is a dollar in a bank account. :::

:::p{html}

Imagine you sold ETH and want to sit out a drawdown without moving money back to a bank. Or you live in a country where the local currency loses a third of its value in a year, and you want dollar savings without an American bank. Or you need to send $50,000 to a partner on another continent on a Saturday night. All three cases call for an asset that does not move with the crypto market. DEXs, lending protocols and payment services need the same unit of account, because debts, collateral and invoices are easier to manage in something that will not drop 20% overnight.

:::p{html}

This is the job of the peg mechanism. It rests on two pillars, the backing and the right to exchange the token at the peg target with the issuer or the protocol. The backing is where they differ. Some tokens are backed by cash and short-term US government securities, others by crypto collateral worth more than the debt issued against it, others by a portfolio of assets with hedging positions. So why does the price stay near $1? Not because the name contains "stable" but because the second pillar works: the moment the market price drifts, arbitrage profits from the gap and pulls it back to target.

:::p{html}

Where the market stands as of September 2026: about $305 billion of stablecoins are in circulation (trackers differ by a few billion depending on methodology), with roughly $185 billion in USDT and $75 billion in USDC.

:::p{html}

Tokens pegged to the euro and other currencies exist, but dollar stablecoins dominate so thoroughly that in this card "stablecoin" almost always means a token targeting $1. The numbers themselves move fast, which is why the date next to a figure matters more than the figure.

:::callout Picture a barbershop down the street selling a punch card: ten haircuts for $90 instead of $120. Each voucher in that booklet is worth about $9, even though it is just cardboard with a stamp. What holds the price is not the cardboard but the fact that any day you walk in, they cut your hair for that voucher. While the shop is open, vouchers circulate almost like money: people gift them to coworkers and resell them in the neighborhood group chat slightly below face value. But the moment a rumor spreads that the owner is leaving and the space is being turned into a coffee shop, the vouchers drop in price instantly. Someone will dump the whole booklet at half price just to get something back. A stablecoin works the same way. The "$1" printed on the token holds while the redemption mechanism holds. The value comes not from the label but from the confidence that someone will honor it. :::

:::keypoints{label="Key points"}

• "Stable" in the name describes a target, not a guarantee: the issuer has no way to forbid the market from trading the token at $0.95.
• The peg is held not by contract code but by the economics around it (issuance, redemption and arbitrage). The ERC-20 itself has no idea what it is worth.
• Stablecoins have become the unit of account for the whole crypto market: DeFi prices, debts and collateral are denominated in them, so a failure of a major stablecoin hits the entire system at once, not one protocol.
• Any stablecoin market figure goes stale within months. The habit of checking current supply on DefiLlama beats any memorized number. :::

:::details{badge=Advanced title="Why the Price Comes Back to $1"}

:::further{label="Further reading"}

• DefiLlama Stablecoins | defillama.com
• Circle Transparency | circle.com
• Tether Transparency | tether.to :::

:::quiz{id=q-5-2-s1 label="Check yourself" correct=1 explanation="Correct. The contract does not control the market price. The peg is maintained by the economics around the token: as long as redemption at face value works, any deviation creates profit for whoever pushes the price back."} Why does a stablecoin usually return to $1 after a small deviation?

• Exchanges automatically cancel trades priced below one dollar
• Minting and redemption at face value make arbitrage profitable
• The $1 target price is hard-coded into the token contract
• Network validators manually push the rate back to a dollar :::

:::section{mins=3} Four Ways to Hold a Dollar Price :::

:::lead The phrase "this stablecoin is backed" says little. What matters is the backing: a dollar of Treasuries, a dollar fifty of ETH, or a futures hedge. In May 2022 the collapse of UST and LUNA showed the price of the fourth option, about $40 billion in a week. :::

:::p{html}

Splitting stablecoins into "safe" and "dangerous" gets you nowhere. The better question is where the peg's stability comes from, and the market has four answers:

:::p{html}

• Reserve-backed (fiat-backed) — USDC and USDT are issued by centralized companies that hold reserves against every token: cash, short-term US government securities and other disclosed assets. It is the most capital-efficient model, since a dollar of debt does not require a dollar fifty of collateral. The price of that convenience is trust in the issuer, its banks and its jurisdiction, plus the issuer's ability to block addresses when its rules require it.
• Crypto-collateralized (overcollateralized) — why would anyone mint a dollar against their own crypto? A typical situation: you hold ETH, you believe it will grow and do not want to sell, but you need dollars now. Protocols like MakerDAO (issuer of DAI, the pioneer of the model) and Aave (issuer of GHO) solve this. A protocol here is simply a smart contract that works the same for everyone: you deposit ETH as collateral, mint stablecoins against it, spend them, and when you repay the debt you take the collateral back. No bank, no forms, no manager's approval.

:::p{html}

The collateral together with the debt minted against it is called a CDP (Collateralized Debt Position). The construction has one hard rule, and it runs like this. The collateral must always be worth noticeably more than the debt. If ETH drops and the cushion shrinks, the position is liquidated before the debt goes underwater. The cost of stability is that the collateral locked up always exceeds the debt issued.

:::p{html}

• Synthetic dollars — USDe by Ethena holds a basket of assets and simultaneously opens short futures positions of roughly the same size. A rise in spot is offset by the short's loss, a fall by the short's profit, and on net the portfolio aims to keep its dollar value. There is no bank reserve of "a dollar per token" here. What stands in its place is a market hedge and the exchanges hosting it.
• Algorithmic without external backing — the peg rests on mint-and-burn rules between two linked tokens. The best-known example is TerraUSD: while the market trusted the UST-for-LUNA swap, the scheme held. When trust vanished, minting LUNA to redeem UST crushed its price, and the recovery mechanism turned into an accelerant of the fall.

:::callout Picture four firms, each selling insurance on a million-dollar apartment. The first keeps a million in cash on a separate supervised account. The second holds no cash but took the owner's country house, worth one and a half million, as collateral: if it fails to pay, the house is sold. The third built a portfolio of assets and bought reinsurance that covers the loss if those assets sink. The fourth promises to settle a fire claim in its own shares, which are worth something as long as clients believe in the firm. In a calm year all four policies look identical and cost the same. The difference shows on the day an entire block burns and everyone must pay at once. Stablecoins differ in exactly this way: not by name, but by who absorbs the loss under stress and with what. :::

:::keypoints{label="Key points"}

• "Backed" does not mean "risk-free": the reserve model carries bank and issuer risk, crypto collateral carries liquidation risk, synthetics carry exchange and hedge risk. The models differ both in where the risk sits and in how much of it there is, so no single number can rank them.
• The reserve model is at once the most capital-efficient and the most centralized, and both traits grow from the same root: one company holds and controls the reserve dollar and answers for it to a regulator.
• Liquidation in the crypto-collateralized model is not a failure but a standard mechanism: one borrower's loss pays for the stability of the whole stablecoin.
• UST did not collapse because of a bug. The mechanism worked exactly as designed; the design simply contained a loop that, once trust broke, accelerated the fall instead of damping it. :::

:::details{badge=Advanced title="CDP, Safety Margin and Liquidation"}

:::further{label="Further reading"}

• Aave — GHO documentation | aave.com
• Ethena — USDe Overview | docs.ethena.fi
• Liquity — LUSD | liquity.org :::

:::quiz{id=q-5-2-s2 label="Check yourself" correct=2 explanation="Correct. Liquidation closes the risky position while the collateral still covers the debt. The loss stays with the individual borrower instead of spreading across all holders of the stablecoin."} What protects a crypto-collateralized stablecoin when collateral prices fall?

• A ban on selling the pledged assets until the debt is repaid
• An automatic increase in the stablecoin's own market price
• Liquidation of the position before the debt goes underwater
• Unlimited minting of the governance token for the protocol :::

:::section{mins=4} The Big Three: USDT, USDC, DAI/USDS :::

:::lead Three token families cover almost 90% of the stablecoin market, and all three promise the same dollar. The difference sits one level down: who issues the token, where the backing lives, and who has the right to ask. :::

:::p{html}

On any exchange you meet the same three dollars, and the difference between them is not the price but who stands behind the promise.

:::p{html}

• USDT — the largest stablecoin by a wide margin, about $185 billion in circulation as of September 2026. The issuer, Tether, is registered in El Salvador and keeps its reserves mostly in short-term US Treasuries, remaining one of the largest holders in the world, with bitcoin and gold as a buffer. USDT's strength is liquidity: it remains the most widely used stablecoin on global crypto markets, especially where access to a bank dollar is limited. The weakness is the mirror image: for years Tether published only quarterly attestations instead of a full audit, and venues in the EU authorized under MiCA, the European rulebook for crypto assets covered in §6, no longer offer USDT. In 2026 Tether announced the completion of the first full audit in the company's history, and what that actually changed is the subject of the advanced block.
• USDC — about $75 billion in circulation. The issuer, Circle, has traded on NYSE since June 2025; reserves sit in a dedicated fund managed by BlackRock and in bank accounts, with monthly reports signed by Deloitte. This is the built-for-regulators model, and it has a price. In March 2023, $3.3 billion of Circle's reserves got stuck in the collapsing Silicon Valley Bank, and USDC sank to $0.87 over a weekend until US authorities guaranteed the deposits. Transparency did not prevent the depeg; it merely let the market price the problem precisely. We will walk through that story in §4.
• DAI and USDS — the only branch of the big three that grew out of a protocol rather than a company. In 2024 MakerDAO, familiar from §2, rebranded to Sky and launched USDS: DAI did not go anywhere, the tokens convert 1:1 (and major exchanges auto-converted DAI balances into USDS in spring 2026), the family's combined supply hovers around $11 billion, roughly $6.5 billion of it in USDS. The system is governed by holders of the governance token, not a board of directors. The caveat is that a large share of the backing has long consisted of USDC and tokenized Treasuries, so independence from the banking system here is a matter of degree rather than a binary property.

:::p{html}

So which of the three should you hold? The question is put wrong, because choosing among them is not a search for "the safest" but a choice of which risk you understand: issuer jurisdiction and concentration with USDT, US banks and regulators with USDC, mechanism complexity and governance with USDS.

:::callout On a pharmacy shelf sit three boxes of the same medicine in the same dosage. The first is an old international brand: sold longer than anyone, stocked in every pharmacy on earth, but the factory is in another jurisdiction, and inspectors were let in for the first time only this year. The second is a certified local manufacturer: it publishes quality-control protocols every month, the plant is nearby, the oversight strict and fussy. The third comes from the compounding counter, made to an open recipe: anyone can inspect it, but some ingredients are bought from those same two factories. The active substance in all three is identical, exactly one dollar. The difference is not in the pill but in who made it, who inspects the plant and how, and what happens to supply if the maker loses its license. :::

:::keypoints{label="Key points"}

• Market share is not a safety ranking: USDT wins where liquidity and access decide, USDC where reporting decides, USDS where resistance to address freezing matters.
• The 2023 USDC depeg exposed an awkward paradox: the most transparent issuer suffered first because the market knew exactly which bank held its money. With Tether, the market simply saw no such specific point of impact that day.
• "Decentralized" for DAI/USDS comes with an asterisk: as long as USDC and Treasuries sit in the backing, censorship resistance runs through the same banks and the same issuers.
• Tether's first audit sharply reduced the uncertainty around the decade-old question "do the reserves exist at all," but the KPMG opinion itself is unpublished, and an audit of last year and redemption on a specific day of stress remain different guarantees. :::

:::details{badge=Advanced title="Attestation, Audit, and What They Are (Not) Worth"}

:::further{label="Further reading"}

• Tether — KPMG audit announcement (Aug 2026) | tether.io
• Circle — monthly reserve reports | circle.com
• Sky — USDS documentation | developers.sky.money :::

:::quiz{id=q-5-2-s3 label="Check yourself" correct=3 explanation="Correct. Disclosing reserves does not protect the reserves themselves. The market knew exactly that $3.3 billion sat in Silicon Valley Bank, and when the bank collapsed, that transparency instantly turned into a price of $0.87."} Why did USDC fall to $0.87 in March 2023 despite fully disclosed reserves?

• The auditor found fictitious assets in Circle's reporting
• The token's smart contract was hacked and then paused
• Circle permanently switched off redemption for all clients
• Part of the reserves was stuck in the collapsed SVB bank :::

:::section{mins=4} Depeg: The Day a Dollar Stops Being a Dollar :::

:::lead Over one March 2023 weekend USDC fell to $0.87, then returned to a dollar as if nothing had happened. A year earlier UST fell from a dollar to 30 cents and never came back. Understanding the difference between these two stories is worth more than any stablecoin ranking. :::

:::p{html}

A depeg is a sustained deviation of a stablecoin's market price from its declared target, usually $1. The key word is "sustained": even on calm days the price wobbles within fractions of a cent, and that is not a depeg but the normal life of a market. A depeg is when the deviation reaches a noticeable size, roughly a cent or more, and is not closed by arbitrage within minutes. That price is the market's answer to an honest question: "can this token really be exchanged for a dollar?" While the answer is obvious, the arbitrage from §1 erases any deviation. A depeg begins when arbitrageurs start doubting, and by the source of that doubt depegs split into two fundamentally different types.

:::p{html}

The first type is an access crisis with the backing alive. This is the USDC story. On Friday, March 10, 2023, Silicon Valley Bank collapsed, and Circle confirmed that $3.3 billion of its reserves were stuck inside. A small share, about 8% of reserves, but the weekend turned it into an abyss: banks closed, redemption offline, and nobody knew whether deposits above the insured limit would be returned.

:::p{html}

Sellers outnumbered the brave, and by Saturday night USDC traded at $0.87. DAI sank alongside: the USDC share in its backing that §3 warned about worked as a contagion channel.

:::p{html}

On Sunday evening US authorities announced that SVB depositors would get all their money back, and redemption of USDC at $1 became guaranteed again. Then simple arithmetic took over: the token traded at $0.90 on exchanges while Circle would redeem it at $1.00, so every purchased token carried 10 cents of nearly risk-free profit. So many rushed for that profit that the wave of buying itself lifted the price, and by Monday USDC was a dollar again. The arbitrageurs' gain was paid by those who sold in panic at the bottom.

:::p{html}

The second type is a crisis of the backing itself. Its defining example is TerraUSD (ticker UST), the algorithmic stablecoin from §2. It lived on its own blockchain, Terra, and LUNA was that network's native coin, like ETH for Ethereum. The entire peg mechanism rested on the swap between them: at any moment the protocol exchanged 1 UST for a dollar's worth of freshly minted LUNA. The second source of demand was Anchor, a protocol on the same Terra chain paying about 19.5% a year on UST deposits.

:::p{html}

In May 2022 large UST sales nudged the price down, holders ran for the exit, and the recovery mechanism turned against the system: to redeem UST the protocol printed LUNA, the growing supply pressed LUNA's price down, and falling LUNA scared the remaining UST holders even more. In a week about $40 billion evaporated from the UST and LUNA ecosystem. There was no full external backing for the price to return to: the main redemption mechanism remained LUNA minting, and the partial bitcoin reserve, covered in the advanced block below, was not enough for everyone.

:::p{html}

Hence the practical takeaway. When you hear about a depeg, ask one question: does the backing exist with the problem in access, or is the backing itself melting? The first is a story about nerves and hours of waiting. The second is about whether you reach the exit in time.

:::callout You paid a deposit for a banquet hall for an anniversary, and a week before the date the hall stops answering the phone. Rumors start going around town, and one of the other customers is already selling their booking at half price just to recover something. Two endings are possible. In the first, the hall's phone system broke; the clients' money sits in the account, on Monday the manager calls everyone back, those who dumped their booking at half price kick themselves, and those who bought up other people's bookings get a banquet at a 50% discount. In the second, the owner gambled away the prepayments long ago, and the silence was not a glitch but a symptom. In that case the one who sold first and salvaged something had it right. From the outside both stories start identically, with silence on the line. The whole difference is whether the money is in the account, and you need to find out before the crowd does. :::

:::keypoints{label="Key points"}

• Panic looks the same in both cases, but the math of the two depeg types differs: with backing alive the token has an anchor the price can return to; with backing melting there is nowhere to return.
• The USDC depeg punished those who sold at the bottom and rewarded those who bought. The UST depeg punished those who waited. The same "hold to the end" strategy produces opposite results in the two types of crisis.
• Contagion through backing moves faster than rumor: DAI sank after USDC above all because of the USDC share in its own collateral, not because its holders panicked harder than anyone else.
• The death spiral is not unique to Terra; it is a property of any scheme where one asset's stability is bought by issuing another that depends on the same faith. :::

:::details{badge=Advanced title="Anatomy of a Death Spiral"}

:::further{label="Further reading"}

• Circle — SVB events blog (March 2023) | circle.com
• Jump Crypto — The Depegging of UST | jumpcrypto.com
• Chainalysis — 2022-2023 depeg analysis | chainalysis.com :::

:::quiz{id=q-5-2-s4 label="Check yourself" correct=2 explanation="Correct. USDC's reserves existed in full; the problem was temporary access to part of them, and the price returned to its anchor. UST had no full external anchor: the main redemption printed LUNA, which got cheaper from the redemption process itself, and the partial LFG reserve burned through in days."} Why did USDC return to $1 in three days while UST never came back?

• Circle bought back every USDC token on the market with its own money
• US regulators banned exchanges from trading UST after the collapse
• USDC was fully backed externally, while UST redemption depended on LUNA
• UST holders simply sold faster than USDC holders did in March :::

:::section{mins=4} Yield: Every Percent Has a Payer :::

:::lead An ordinary payment stablecoin like USDT or USDC pays nothing, it is a token in a wallet, not a deposit. Short-term US Treasuries yield about 4% a year, and that is where issuer reserves sit. Who pockets that interest, and where bigger promises come from, is the economics of stablecoins. :::

:::p{html}

Start with an awkward fact: the USDT and USDC in your wallet pay you nothing. The reserves behind them sit in Treasuries, the interest drips in every day, and the issuer keeps it. This is why Tether, with its compact team, consistently ranks among the world's most profitable companies per employee. Regulators have cemented the arrangement: a ban on paying interest to holders of payment stablecoins is written both into the already-operating European MiCA and into the enacted American GENIUS Act, so that the tokens do not become an unregulated twin of bank deposits.

:::p{html}

Is there a way to reach that interest? The market answered with wrappers. Put USDS into a savings contract and you receive sUSDS, whose exchange rate against USDS slowly rises: the Sky protocol passes the holder income from its reserves and lending operations. sUSDe by Ethena and tokenized money-market funds work along the same lines. Formally this is no longer a payment stablecoin but a different product, and in that sense the ban worked: it did not remove yield, it drew a border between a "token for payments" and a "token as deposit."

:::p{html}

From there the ladder climbs, and on every rung the payer changes. Lend stablecoins through a protocol like Aave: the interest comes from borrowers, mostly leveraged traders, so the rate swings with the market's appetite. Hold sUSDe: the protocol's income comes from three sources — funding and basis on the hedging positions from §2, staking on part of the backing assets, and yield on the liquid stablecoins in the reserve. If the combined result turns negative, the loss is taken by the reserve fund, not by sUSDe holders: their rate simply drops toward zero, and the open question is how long the fund lasts in a prolonged slump.

:::p{html}

See 30% a year at a young protocol: almost certainly you are being paid in its own token, printed to attract liquidity, and that yield will die when the printing stops. You already know the story of Anchor, its 19.5% and the $40 billion it cost the Terra ecosystem, from §4.

:::p{html}

Hence the rule this section exists for. Every yield has a source: bond interest, a borrower's rate, the futures market, or someone's token printing. If you cannot name the source and explain why it pays above the risk-free rate, then the missing part is paid by you; the bill just has not arrived yet.

:::callout Imagine you inherited the rights to a few songs. A distributor sends streaming royalties: a couple percent of the catalog's value a year, boring, but every month and without surprises. An agency offers twice as much by placing the songs in ads and TV shows: the income is higher but depends on the season and clients' tastes, and in a bad year there may be nothing. Then a studio appears offering "30% a year on your catalog": it says it invests the songs in promoting young artists, and for the first months the money really arrives. Except the source of that money is not listeners but the deposits of new rights holders, lured by the same 30%. The music is identical; the difference is who pays and for what. If nobody can show you a listener paying for the song, the payment comes from the next client. :::

:::keypoints{label="Key points"}

• The interest ban in the GENIUS Act and MiCA is not a ban on yield but a product border: payment token on one side, yield wrapper on the other. Knowing which side your token is on matters more than the rate itself.
• USDT and USDC do not pass reserve income to the holder, so a separate market of products has grown around stablecoins that return part of the capital's yield to the user — at the cost of extra risk. Competition for that interest is what drives the savings-token market.
• A fixed rate does not remove the volatility of the underlying income; someone absorbs it. With fixed yield the key questions are who guarantees the rate and out of what buffer they cover the gap in a bad month.
• The higher the rate above Treasury yields, the shorter the list of possible explanations, and "we found a way to pay more without risk" is not on that list. :::

:::details{badge=Advanced title="The Risk Ladder and the Leverage Loop"}

:::further{label="Further reading"}

• Sky — Savings Rate documentation | developers.sky.money
• Ethena — sUSDe and funding mechanics | docs.ethena.fi
• DefiLlama — Yields (current stablecoin rates) | defillama.com :::

:::quiz{id=q-5-2-s5 label="Check yourself" correct=3 explanation="Correct. The protocol's income comes from three sources: funding and basis on the short futures, staking on part of the backing assets, and yield on the liquid stablecoins. When their sum turns negative, the loss is absorbed by the reserve fund, so the sUSDe rate can drop to zero, while a prolonged negative period drains the reserve itself."} Where does the yield of Ethena's sUSDe mainly come from?

• From Ethereum network fees charged on token transfers
• From subsidies Ethena pays out to attract new users
• From a fixed rate guaranteed by the token's issuer
• From hedge funding, staking, and yield on the liquid reserve :::

:::section{mins=4} Payments and Regulation: Stablecoins Move into the Foundation :::

:::lead Visa settles with banks in stablecoins on nine chains, Stripe bought Bridge for $1.1 billion, and in mid-2026 over a hundred companies announced the Open Standard consortium. These experiments ran from 2021; in 2026 they became infrastructure, alongside new stablecoin regimes in America and Europe, and not by accident. :::

:::p{html}

A stablecoin starts life as a trader's instrument and ends up inside the payment button you press without thinking. Getting there took a law.

:::p{html}

Before MiCA and the GENIUS Act, stablecoins had no dedicated legal framework in the major Western jurisdictions: they lived inside a patchwork of banking, trust and money-transmission rules. That did not stop the experiments — Visa piloted USDC settlement back in 2021, PayPal launched PYUSD in 2023 through a regulated New York trust — but every attempt to scale ran into legal uncertainty.

:::p{html}

Dedicated rules arrived, though at different speeds. In the EU the MiCA regulation has applied in full since late 2024: issuing currency-pegged tokens requires authorization, and reserves and redemption are tightly prescribed.

:::p{html}

In the US the GENIUS Act was signed on July 18, 2025 — the first federal framework for payment stablecoins with the same logic: a license, reserves 1:1 in cash and short-term Treasuries, monthly disclosure of composition, an interest ban.

:::p{html}

One caveat matters as of September 2026: the American framework is still being rolled out. The implementing rules are being finished, and by law the regime takes effect no later than January 18, 2027 — or earlier, 120 days after the final rules are published.

:::p{html}

What does regulation change in practice? It quickly revealed a property worth remembering, namely that it governs access without touching the token itself. The story of USDT in Europe is the textbook case. The contract runs, the reserves are in place, but MiCA-authorized venues delisted the token: trading it inside the regulated perimeter became impossible, though holding and transferring it remains allowed. For a European user the world's largest stablecoin moved into the hard-to-reach category. Not a line of code changed. What changed was the perimeter inside which that code can be legally used.

:::p{html}

Inside the perimeter, meanwhile, things started moving at a pace that would have sounded like fiction three years ago. Visa expanded stablecoin settlement with participating banks to nine blockchains and, together with Bridge, issues stablecoin-linked cards in 18 countries. Stripe, after acquiring Bridge, is embedding stablecoins into its APIs as an ordinary payout method, alongside cards and bank transfers. In June 2026 more than a hundred companies, from Visa and Stripe to BNY Mellon and BlackRock, announced the Open Standard consortium with its own stablecoin. Note the common denominator: none of these products show the word "stablecoin" to the user. The user sees a card, a payout, a transfer to a merchant.

:::p{html}

That is the main shift. The stablecoin stops being an asset bought for its own sake and becomes a settlement layer under familiar buttons. For you as a user the practical conclusion follows: the question "which stablecoin should I use" gradually turns into "which stablecoin is wired into my service, and what happens to my money if its issuer runs into trouble." After §3 and §4 you already know how to answer the second half.

:::callout When you send money abroad through a bank, you press one button and see "transfer sent." Under that button lives an invisible world: a message enters the SWIFT network, then a chain of correspondent banks passes the money along day after day, each link taking its fee. You never see that world, and you probably will not recall the name SWIFT until a transfer gets stuck. Stablecoins are bidding for exactly that place in new services: to be the pipes under the button. The button stays the same, "pay" or "withdraw," while the path beneath it switches from a chain of correspondents to a token on a blockchain arriving in a minute. Good infrastructure is noticeable only on the day it breaks. Which is why the more successful stablecoins become, the less often you will meet the word itself. :::

:::keypoints{label="Key points"}

• The GENIUS Act carves payment stablecoins into a dedicated federal regime; MiCA is built wider, creating a general regime for crypto-assets with the special EMT and ART categories. What the two frameworks share is one thing: rules that are banking in spirit, not exchange rules.
• Regulation works through the access perimeter: exchanges, banks and payment services. The token itself can live anywhere, as USDT leaving authorized EU venues with a live contract and reserves demonstrated.
• The interest of Visa, Stripe and banks is explained not by fashion but by the economics from §5: a stablecoin settlement layer beats correspondent chains on speed, and reserves parked in Treasuries earn money by themselves.
• The deeper stablecoins sink into the foundation of services, the less consciously users choose them, and the more it matters who answers to the user when something breaks: the token's issuer, the service, or no one. :::

:::details{badge=Advanced title="GENIUS Act and MiCA: Two Frameworks, One Logic"}

:::further{label="Further reading"}

• GENIUS Act text (US Congress) | congress.gov
• ESMA — MiCA guidance for issuers | esma.europa.eu
• Visa — Onchain Analytics | visaonchainanalytics.com :::

:::quiz{id=q-5-2-s6 label="Check yourself" correct=0 explanation="Correct. Regulation governs the access perimeter, not the token itself. USDT's contract and reserves did not change, but trading access through regulated venues narrowed sharply, and that proved enough to push the token to the margins of the EU market."} What changed for European USDT holders when MiCA came into force?

• Authorized EU venues stopped offering the token
• The USDT smart contract was blocked on EU territory
• Tether's reserves were placed under a European regulator
• All USDT in European wallets was forcibly redeemed :::