Stablecoin Depeg Risk: What Breaks a Dollar Peg

Introduction

Is your "dollar" actually a dollar? For most holders the honest answer is: it is a claim on a dollar, held up by a mechanism — a redemption desk, a pool of collateral, or a hedged trading position. As long as the mechanism works, the claim and the price agree, and the coin earns its name. When the mechanism comes under stress, the two can diverge, and that divergence is a depeg. The word covers everything from a half-cent flicker that closes in an hour to a collapse that never comes back, which is exactly why it confuses people: it names a symptom, not a cause.

This page sorts the causes into three classes, because the class decides the outcome. A design failure means the collateral model itself was unsound — the canonical case is UST, whose algorithmic peg unwound in May 2022 and erased roughly forty billion dollars, permanently. A custody failure means the backing is real but temporarily unreachable — the canonical case is USDC in March 2023, which fell to about 87 cents over a weekend while part of its cash sat in a failed bank, then recovered fully within days.

And a synthetic-carry profile is not a failure at all but a different instrument: a dollar engineered from hedges rather than backed by cash, as with USDe, carrying risks that neither of the first two classes has. The spread between those outcomes — total loss, full recovery, and a live strategy in between — is the whole argument for learning the taxonomy before you need it, rather than during the weekend that reveals which one you were holding.

Most reader harm in this subject comes from lumping the classes together: treating every wobble as a collapse, or worse, assuming every coin that says "stable" fails softly. Neither is true. So the page works through the three stories in full — mechanism, timeline, and in UST's case the now-concluded legal record — and then converts them into three practical tools: a four-criterion checklist for judging any stablecoin before trouble, the early-warning signals worth watching as it approaches, and a short decision framework for the worst moment of all — when a coin you already hold is trading below par and you have minutes, not days, to tell a recoverable wobble from a terminal break. The wider map of designs and reserves lives in our complete stablecoin guide; this page owns the failures.

What a depeg actually is

Every stablecoin defends its dollar with one of four mechanisms, and knowing which one is on duty tells you what a price break means. A fiat-reserve coin like USDC or USDT defends the peg through redemption arbitrage: if the market price slips below a dollar, arbitrageurs buy the discounted token and redeem it at par with the issuer, pocketing the difference and pushing the price back up. An over-collateralised crypto-backed coin like DAI or USDS defends it through liquidation machinery: the collateral behind each token is worth more than a dollar, and smart contracts sell it automatically before it can fall below the claim.

An algorithmic or seigniorage design — UST was the flagship — defends it with a mint-and-burn arbitrage against a volatile sister token, which is to say, with confidence. And a synthetic delta-neutral dollar like USDe holds its value through offsetting positions: collateral long, perpetual futures short, gains and losses cancelling.

Each defence has a signature failure, and naming them in advance is half the work of judging a coin under stress:

• Fiat-reserve (USDC, USDT) — redemption arbitrage. Breaks when the exit at par is blocked: reserves trapped at a failed bank, or redemption gated by policy. Recoverable when the assets are real, as the USDC weekend showed.
• Crypto-collateralised (DAI, USDS) — automated liquidation. Breaks when collateral falls faster than the machinery can sell it, in a crash violent enough that liquidations cannot keep pace with the price.
• Algorithmic (UST) — mint-and-burn against a sister token. Breaks when confidence in the sister token goes, because the shock absorber and the peg are the same variable; this is the class with no floor.
• Synthetic delta-neutral (USDe) — offsetting hedges. Breaks when the hedge does: a prolonged negative-funding regime, or the derivatives venue hosting the position straining under the same stress that is moving the price.

Speed is the other axis, and it maps onto the classes in a way worth internalising before you need it. A design-class break is fast and self-accelerating — UST's spiral ran its course in roughly three days, because each turn of the mechanism fed the next. A custody-class break is bounded by an external clock — USDC's trough lasted a weekend because that was how long the banking rails stayed shut, and the recovery was as quick as the resolution that reopened them. A synthetic dollar breaks on the timescale of its funding regime, which can grind negative for weeks rather than snap in days. Knowing roughly how fast the class in front of you tends to move is the same as knowing how much time you have to think — three days of accelerating collapse and one closed weekend demand very different reflexes.

Arbitrage is the immune system in all four cases, and a depeg is what you see when the immune system stops working. Two things reliably break it. The first is that redemption becomes gated or paused — the arbitrage loop needs an exit at par, and if the exit is closed, even briefly and even for good reasons, the discounted price has nothing to anchor it. That is the custody class. The second is that the backing itself falls in value — collateral crashing faster than liquidations can clear, or a sister token collapsing under the very selling the mechanism creates. That is the design class, and it is the one with no floor.

There is a second ingredient the immune system needs, and it is the one that fails quietly: liquidity. Arbitrage only closes a gap if there is depth on the other side of the trade — enough buyers of the discounted token, enough capacity to redeem or hedge at size. When secondary-market depth thins, the same sell pressure that would move a liquid coin a tenth of a cent moves a thin one several cents, and the wider gap frightens more holders into selling, which thins the book further. That feedback is why a depeg can look like a sudden cliff rather than a gentle slope: the mechanism does not degrade smoothly, it degrades as liquidity evaporates beneath it.

It is also why the depth of a coin's markets, and not only the size of its reserves, belongs on any honest risk assessment — a coin can be fully backed and still gap hard if too few people are willing to stand in front of the sell-off.

Magnitudes matter as much as mechanisms. A deep-liquidity coin drifting a tenth of a per cent from par intraday is not depegging; that is the ordinary texture of a market, and it closes on its own. The signal worth respecting is a sustained break — below roughly 98 cents, say — especially when it arrives together with gated redemptions or visible stress in the backing. And underneath every judgement this page will help you make sits one question, worth memorising because it separates the recoverable from the terminal: is the backing gone, or just temporarily unreachable? The two case studies that follow are the cleanest possible illustrations of each answer.

Case study one: UST and the design-class failure

UST was the largest experiment ever run on the idea that a dollar peg could be held by pure mechanism, with no hard reserves at all. The design was elegant on paper: one UST could always be exchanged for one dollar's worth of newly minted LUNA, the ecosystem's volatile sister token, and vice versa. If UST traded below a dollar, arbitrageurs would buy it, burn it for a dollar of LUNA, and profit — absorbing the sell pressure. The peg's shock absorber was LUNA's market value, which meant the peg's shock absorber was confidence itself.

The subsidised-yield demand trap

What made the structure genuinely fragile was the demand engine bolted onto it. Anchor Protocol — the Terra ecosystem's flagship lending platform, now defunct since the 2022 collapse — paid a subsidised yield of roughly 20 per cent on UST deposits, and at its peak roughly 70 per cent of all UST in existence sat in that single protocol. Read that concentration for what it was: most of the "demand" for the stablecoin was demand for an unsustainable interest rate, parked in one place, ready to leave the moment the rate or the confidence wavered. The yield was not a feature of the system; it was the system.

The three-day death spiral

In May 2022 the confidence wavered, and the mechanism did exactly what it was designed to do — which was the problem. Large UST sales pushed the price below par; arbitrageurs burned UST and minted LUNA; the minting diluted LUNA's price; the falling LUNA made the backing look weaker, prompting more UST sales. Over roughly three days, LUNA's supply hyperinflated from about one billion tokens to some six trillion, its price fell towards zero, and UST followed it down, because there was nothing else beneath it. Approximately forty billion dollars of value was erased. What would a holder have seen on day one? A stablecoin a few cents below par — indistinguishable, at a glance, from the wobble USDC would later survive. The difference was underneath, in what the backing was made of.

The legal record is now closed, which this page states precisely because it is so often garbled. The criminal track: Do Kwon was extradited to the United States on 31 December 2024, pleaded guilty in August 2025, and was sentenced to fifteen years in prison on 11 December 2025 in the Southern District of New York. The civil track, separately: the SEC secured a judgment of approximately 4.55 billion dollars in June 2024, following a jury verdict that April. The two tracks are distinct — the criminal case belonged to the Department of Justice, the civil judgment to the securities regulator — and both concluded on fraud and misrepresentation grounds layered on top of what was, underneath, a design that concentrated reflexivity until it snapped.

The lesson: in an algorithmic design, the peg and its shock absorber are correlated — stress hits both at once, and there is no reserve to fall back on. When the backing of a stablecoin is another asset of the same ecosystem, the honest description is not "stable"; it is "leveraged on confidence". Design-class risk does not announce itself with a small depeg; it announces itself in the architecture, visible from day one to anyone who asks what actually stands behind the token.

Case study two: USDC and the custody weekend

Ten months later the market ran the counter-experiment. USDC was everything UST was not: fully reserve-backed, transparent about its holdings, issued by a regulated company. And it still broke its peg — which is exactly why the episode is so instructive, because it shows what a depeg means when the backing is real.

The exposure: real cash, trapped for a weekend

The mechanics were banal, which is the point. In March 2023, roughly 3.3 billion dollars of USDC's reserves — about eight per cent — were held as cash at Silicon Valley Bank. On Friday 10 March, SVB failed, the second-largest bank failure in US history at the time. The cash was not lost in any final sense, but it was suddenly unreachable, trapped inside a bank resolution over a weekend when the traditional banking rails were closed and redemptions could not operate normally. The market did what markets do with uncertainty: USDC's price fell, troughing near 87 cents, as holders priced in the worst case they could not yet rule out.

The resolution came fast. Between 12 and 15 March 2023, the Federal Reserve, the FDIC and the Treasury announced that all SVB depositors would be made whole; Circle confirmed full access to the trapped reserves; and USDC returned to its dollar peg within days, with the redemption backlog clearing as the rails reopened. Nothing about the coin's solvency had ever changed — the dollars existed throughout. What failed, for one weekend, was access. The precise framing matters enough to italicise in your memory: this was reserve-custody risk at a banking partner, not issuer insolvency, and "USDC failed" is simply the wrong sentence about what happened.

The rebuild, and what it can and cannot promise

The aftermath reshaped the coin's structure in ways a careful reader can verify today. As of July 2026, USDC's reserves sit predominantly in short-dated US Treasuries held through a dedicated, BlackRock-managed fund with custody at BNY, with the cash portion spread across regulated banks, and the whole structure attested monthly. Circle itself became a public company in June 2025, listing on the New York Stock Exchange as CRCL, and holds e-money authorisation in the EU through its French entity — a licence effective since July 2024. You do not have to take any of that as a guarantee; the 2023 lesson is precisely that even a fully backed coin inherits its banking partners' risk. But the difference between a coin that concentrates its cash in one fragile bank and one that publishes where every dollar sits is a difference you can check before the weekend it matters.

The lesson: a fully backed stablecoin can still depeg, because backing and access are different things. The recoverable class of depeg looks like this one — real assets, temporary barrier, fast normalisation once the barrier lifts. The evaluation habit it teaches: do not just ask whether reserves exist; ask where they are, how concentrated they are, and what has to stay open for you to reach them.

The synthetic-dollar risk profile

The third profile in the taxonomy is not a failure story — it is a category that too many readers file in the wrong drawer. A synthetic dollar such as Ethena's USDe maintains a dollar-shaped price without holding dollars: it holds crypto collateral and hedges it with short perpetual-futures positions of equal size, so that when the collateral falls the short position gains, and the combined value stays near one dollar. It is a genuinely clever construction, and nothing about this section says otherwise. What this section says is that it is a different instrument — a delta-neutral trading strategy with a peg-shaped price target — and that treating it as a peer of USDT or USDC is the single most common category error in the stablecoin conversation as of 2026.

Follow the yield to see the difference. The staked form, sUSDe, earns from three streams: the funding rates that long traders pay short positions on perpetual futures, the staking rewards on the collateral, and the yield on backing assets. Notice what is absent: reserve interest from cash and Treasuries, the engine behind a fiat-backed coin's yield. And notice what is present that a fiat-backed coin does not have: a direct dependency on derivatives-market conditions. Funding rates are not a law of nature — they are the market's mood, and they flip negative in bearish stretches, at which point the strategy pays to hold its own hedge rather than earning from it. A prolonged negative-funding regime is to a synthetic dollar what a failing bank is to a fiat-backed one: the specific stress its design cannot fully engineer away.

Two further risks complete the profile. The hedging positions have to live somewhere, and that somewhere is centralised derivatives venues, reached through custodial arrangements — so the construction carries venue and custody exposure at its core, not at its edge; the general discipline for evaluating centralised-venue exposure is covered in our operational security guide. And in fast markets, the neutrality itself is tested — hedges must be maintained and rebalanced under exactly the conditions when venues strain. Add the regulatory dimension: USDe has faced regulatory action in Germany, where the financial regulator moved against the issuer's German entity under MiCA, affecting its European availability — a live situation that readers in the EU should verify as of the time they read this rather than assume in either direction.

The lesson: a synthetic dollar's peg and its yield come from the same live strategy, so you cannot hold the stability and decline the strategy risk — they are the same position. Held knowingly, at a size that respects what it is, USDe is a legitimate instrument. Held as "another stablecoin", it is a misfiled derivative. The question this page keeps asking — is the backing gone or just unreachable? — has a third answer here, and it is the one to memorise: the backing is a trade, and it is only as good as the conditions the trade needs.

The issuer-risk checklist

Three stories, three classes — now the payoff. These four criteria turn the case studies into a repeatable evaluation you can run on any stablecoin in twenty minutes, before the weekend when it matters. Each criterion comes with what good and worrying look like, grounded in the failures above.

1. Reserve transparency: attestation or audit?

Start with who verifies the reserves, how often, and at what level of assurance. An attestation is a point-in-time confirmation by an accounting firm that reserves matched liabilities on a given date; a full independent audit is a deeper, standards-based examination — and the two words are not interchangeable. The live example, as of July 2026: Tether publishes quarterly attestations from BDO Italia, with the first-quarter 2026 report showing roughly 191.7 billion dollars of assets against about 183.5 billion of liabilities — around 104.5 per cent coverage, roughly 80 per cent of it in US Treasuries and cash-like holdings alongside gold and Bitcoin positions — and in March 2026 it engaged a Big Four firm, reported to be KPMG, for its first full independent audit, which is under way, not completed.

Circle publishes monthly attestations of USDC's Treasury-heavy reserve fund. Good looks like: frequent verification, a named firm, a published breakdown, and ideally a real audit. Worrying looks like: gaps, vagueness about composition, or no third party at all. The side-by-side reserve record for the two big coins lives in our USDT versus USDC comparison.

2. Redemption terms: who can actually exit at par?

A peg is only as strong as its redemption path, so read the terms as if you will need them under stress. Who may redeem directly with the issuer — anyone, or only institutional partners above a minimum size? What rights does the issuer reserve to gate, delay or fee redemptions? Which banking rails does redemption depend on, and when are they closed? The case studies bracket the range: UST had no redemption into hard assets at all — its "exit" was a mint of more volatile tokens — while USDC's redemption was gated for a weekend by a bank closure rather than by policy, and normalised as soon as the rails reopened. Good looks like: broad redemption rights, clear terms, multiple banking partners. Worrying looks like: redemption as a privilege of insiders, or terms that let the issuer close the exit precisely when you would want it.

3. Jurisdiction: does the issuer answer to anyone?

Regulatory standing does not guarantee a peg, but opacity plus no regime is a tail-risk multiplier. The consequence-level map as of July 2026: in the EU, MiCA's stablecoin rules have applied since 30 June 2024, and an issuer's authorisation status decides whether compliant venues may sell its coin at all. In the US, the GENIUS Act was signed on 18 July 2025 but is not yet in force — the earliest backstop date is 18 January 2027. In the UK, the FCA's final issuer rules were published on 30 June 2026 and come into force on 25 October 2027.

The full machinery — who must do what, from when, and what it means for your access — is the subject of our stablecoin regulation guide. For this checklist, the question is simpler: is there a regulator anywhere with the power to make this issuer tell the truth about its reserves? If the answer is no, the transparency criterion above is running on the issuer's honour alone.

4. Concentration: what single failure breaks the whole thing?

Concentration is the through-line of every story on this page. UST concentrated its demand in one yield sink — Anchor, holding roughly 70 per cent of supply at roughly 20 per cent subsidised yield. USDC concentrated a slice of its cash in one bank, and that slice set the price for the whole coin over a weekend. A synthetic dollar concentrates its solvency in the venues that host its hedges. So for any coin you evaluate, ask where the single points of failure are: one bank, one custodian, one protocol, one exchange, one sister token. Good looks like: named diversification you can verify. Worrying looks like: a structure whose honest diagram has one box that everything else plugs into. If you also earn yield on the coin, the same concentration test applies to the yield's source — a discipline our earning-yield guide builds into a full checklist of its own.

The checklist applied: UST and USDC scored

Run the four criteria across the two case studies and the difference that decided everything was visible on every line, months before either coin moved. On reserve transparency, UST had nothing to verify — there were no hard reserves, only a mechanism — while USDC published its holdings, so a reader who insisted on seeing real assets would have declined UST on the first criterion alone. On redemption terms, UST's only exit was a mint of more LUNA, which is no exit at all under stress, whereas USDC's redemption path was sound and merely gated by a closed bank for a weekend. On jurisdiction, UST answered to no regulator with power over reserves it did not have, while USDC's issuer was a supervised company with those reserves attested by a named firm.

And on concentration, both failed in different registers — UST parked roughly 70 per cent of its supply in one subsidised yield sink, USDC held about 8 per cent of its cash at one bank — yet only one of those concentrations was fatal, because only one sat behind a peg with no floor beneath it. The checklist does not forecast the day of a break; it tells you in advance, and in cold blood, which coin is built to survive one.

The twenty-minute version is mostly knowing where to look. For transparency, find the issuer's reserves or attestations page and check the date and the named firm behind it, not just the headline coverage figure. For redemption, read the terms of service for who may redeem and under what conditions — the gating rights are usually written down, if rarely advertised. For jurisdiction, check whether the issuer names a regulator and a licence, and whether that claim survives a search of the regulator's own public register rather than the issuer's marketing copy. For concentration, read the reserve breakdown for single-bank or single-custodian exposure. None of this requires expertise; it requires the habit of reading the primary document instead of the summary — which is exactly the habit that separated the holders who stepped away from UST early from the ones who read only the yield.

Early-warning signals

The checklist is for before; these are the tells for during. Which of these did UST show, and how early? Most of them, for the better part of a year — which is the point of writing them down.

• Market tells: a sustained price below roughly 98 cents rather than a passing flicker; secondary-market depth thinning so that modest sells move the price; large, visible redemptions by sophisticated holders while retail inflows continue.
• Redemption tells: queues lengthening, minimums rising, processing slowing, or redemptions paused outright — the immune system being switched off, whatever the stated reason.
• Yield tells: an above-market rate with no visible source — Anchor's roughly 20 per cent was the loudest warning in the whole UST story — or a synthetic yield that quietly depends on derivatives funding staying positive.
• Structural tells: heavy reserve concentration at a single bank or custodian; attestations overdue or getting vaguer; governance keys with the power to pause redemption; a sister-token dependency anywhere in the backing.
• Regulatory tells: an issuer withdrawing from a licensing process, a regulator moving against an issuer or venue, or compliant platforms delisting the coin — each a signal that someone with more information than you has repriced the risk.

One calibration note, because over-reaction has costs too. The signals are cumulative, not binary: a transparent, fully backed coin showing one market tell during a general panic is a "reduce and watch" situation, as the USDC weekend proved for those who sold the bottom at 87 cents and bought their dollars back a week later at par. A coin showing yield, structural and redemption tells together is a different conversation entirely. The skill this page has been building is precisely that discrimination — reading which class of trouble you are looking at, because the right response to a custody wobble and the right response to a design break are opposites.

A practical way to hold that discrimination in mind is to weight the tells by how hard they are to fake. A market tell on its own can be noise, or manipulation, or a general panic dragging good coins down with bad. A redemption tell — a paused or gated exit — is harder to explain away, because it touches the arbitrage loop directly. A structural tell in the backing, or an issuer quietly withdrawing from a licensing process, is harder still, because it means someone with better information than yours has already repriced the risk. One soft tell is a reason to look closer; a hard tell, or two tells from different rows of the list, is a reason to act on the classification you have already made rather than wait for confirmation the chart will only give you too late.

Reading a live depeg: recovery or terminal decline

The checklist prepares you and the signals alert you, but the hardest moment is the one in between — when a coin you hold is trading at ninety-four cents and you have minutes, not days, to decide. The two case studies are the templates here, because they are the two endings, and telling them apart in real time reduces to a small number of questions you can actually answer while the price is moving.

Is the backing an asset or a mechanism?

This is the single most decisive question, and it is answerable in advance from the coin's design alone. If the dollar is backed by cash and short-dated government debt held at named custodians, a price break is a problem about access, and access problems resolve — the USDC weekend is the template, where the trough near 87 cents reversed within days of the SVB depositors being made whole. If the dollar is held up by a mint-and-burn loop against a sister token, a price break is a problem about the mechanism itself, and mechanism problems compound — UST is the template, where the same arbitrage that was meant to defend the peg is what destroyed it. A holder who had settled this one question beforehand knew, on day one of each event, which story they were in, even though both charts looked identical at ninety-four cents.

Is the arbitrage loop still open?

A recoverable depeg keeps a path back to par: redemptions still function, or will the moment a closed rail reopens, so arbitrageurs have a reason to buy the discount and push it up. A terminal depeg has lost that path — either redemption is into a collapsing asset, as UST's mint of ever-cheaper LUNA was, or it is gated with no reopening in prospect. When redemptions pause, the useful question is not whether they are paused but why: a closed banking rail over a weekend is temporary and dated; a solvency doubt has no reopening time attached to it.

What is the honest worst case?

Size the downside before you act, because both directions of mistake carry a bill. For a fully backed coin behind a temporary access barrier, the worst realistic case is a haircut and a wait — which is exactly why selling USDC at 87 cents and buying the same dollars back at par a week later was the expensive answer. For a design-class break, the worst case is zero — which is why holding UST through a recovery that could not arrive was the far more expensive one. Over-reaction and under-reaction each have a price; matching the response to the class you have already identified is how you avoid paying the wrong one. The discipline is not prediction, which is impossible mid-panic, but classification, which the rest of this page has made possible before the panic starts.

Conclusion

Three classes, three outcomes. A design failure is terminal: when the backing is confidence in a sister asset, there is no floor, and UST's forty billion dollars did not come back. A custody failure is usually recoverable: when real assets are temporarily unreachable, the price snaps back once access does, as USDC's did within days of the SVB resolution. And a synthetic dollar is its own third thing — not a failure but a live strategy, whose peg is exactly as strong as the trade behind it and whose holders have accepted a carry risk whether they know it or not. Same symptom on a price chart; utterly different diseases.

If you keep one habit from this page, keep the question: is the backing gone, or just unreachable — and can I see it? Everything else is elaboration. The checklist gives that question structure: verified reserves, a real redemption path, a regulator with teeth, and no single box in the diagram that everything else plugs into. The early-warning list gives it timing. And when the moment arrives anyway — a coin you hold below par, the clock running — the three questions turn panic into classification: is the backing an asset or a mechanism, is the arbitrage path still open, and what is the honest worst case? Answer those and you are responding to the disease rather than the symptom, which is the whole difference between the holder who sold USDC at eighty-seven cents and the one who did not.

Transparency, diversification and redemption rights are the durable defences, and they are all checkable in advance, which is the quiet good news of the whole subject: the differences that decided UST's holders' fate against USDC's were visible before either coin moved a cent off its peg. The wider framework — designs, reserves, regulation and choosing a coin in the first place — lives in our complete stablecoin guide; keep the checklist from this one within reach, because the coin that finally tests it will not send a calendar invitation first.

Sources

• US Department of Justice — Terraform Labs / Do Kwon criminal proceedings (SDNY): primary record for the extradition (31 December 2024), the August 2025 guilty plea and the fifteen-year sentence of 11 December 2025.
• US SEC — Terraform Labs civil action: the separate civil track — April 2024 jury verdict and the ~$4.55 billion judgment of June 2024.
• US Federal Reserve — FEDS Note on the SVB failure and the March 2023 stablecoin run: regulator analysis of the USDC depeg, the ~$3.3B SVB exposure, the ~$0.87 trough and the recovery after the 12–15 March backstop.
• Circle — USDC transparency, reserve structure and attestations: issuer source for the monthly attestations, the Treasury-heavy reserve fund, and the post-2023 custody diversification.
• Tether — reserve attestations and audit-engagement announcement: issuer source for the quarterly BDO Italia attestations, the Q1 2026 figures (~$191.7B vs ~$183.5B), and the March 2026 first-full-audit engagement.
• Ethena — USDe and sUSDe documentation: issuer documentation for the delta-neutral mechanism and the funding-plus-staking yield composition behind the synthetic-dollar profile.
• ESMA — statement on non-compliant stablecoins (17 January 2025): the offering-versus-custody clarification behind the jurisdiction criterion's EU row.
• US Congress — GENIUS Act (S.1582, Public Law 119-27): primary record for the signed-not-yet-effective status and the January 2027 backstop date.
• UK FCA — PS26/10, final rules for qualifying stablecoin issuers (30 June 2026): source for the UK dates in the jurisdiction criterion — rules published 2026, in force 25 October 2027.
• Kraken — official security statements: source for the April 2026 statement on the attempted insider-assisted extortion incident (no breach, no client funds at risk) disclosed alongside the venue mention.

Frequently asked questions

What does it mean when a stablecoin depegs?

A depeg is when a stablecoin's market price moves away from the value it promises to hold, usually one dollar. The distinction that matters is between transient and structural. A brief dip of a fraction of a per cent on a deep, liquid coin is noise — arbitrage closes it within hours. A sustained break, especially with redemptions paused or gated, is a signal, because it means the mechanism that normally pulls the price back to a dollar is not working. The single most useful question to ask during any depeg is: is the backing gone, or just temporarily unreachable? When the backing is real but briefly trapped, prices tend to recover, as USDC's did in March 2023. When the backing never existed in a hard form, as with UST in May 2022, there is nothing for the price to recover to.

Has a stablecoin ever gone to zero?

Yes. UST, the algorithmic stablecoin of the Terra ecosystem, collapsed to near zero in May 2022, erasing roughly forty billion dollars of value in about a week. It was a design-class failure: UST held its peg through a mint-and-burn mechanism with a sister token, LUNA, rather than through hard reserves, and when confidence broke the mechanism hyperinflated LUNA's supply from around one billion to some six trillion tokens in roughly three days. The legal aftermath is now concluded: Do Kwon was extradited to the United States at the end of 2024, pleaded guilty in August 2025, and was sentenced to fifteen years in December 2025, with a separate civil judgment of about 4.55 billion dollars entered in 2024. No fully reserve-backed fiat stablecoin has gone to zero — but reserve-backed coins can still wobble, as USDC showed in 2023.

Did USDC fail during the SVB collapse?

No — and the difference between what happened and failure is the most instructive lesson in stablecoin risk. In March 2023, roughly 3.3 billion dollars of USDC's reserves — about eight per cent — were held as cash at Silicon Valley Bank when the bank failed. Over the weekend, with redemptions effectively gated by the bank closure, USDC's market price fell to around 87 cents as holders feared the cash was lost. It was not: when US authorities guaranteed SVB's deposits between 12 and 15 March 2023, Circle confirmed full access to the funds and USDC returned to its dollar peg within days. The backing was real the whole time; it was temporarily unreachable. That is custody risk at a banking partner, not issuer insolvency, and calling it a failure gets the category wrong.

Is USDe from Ethena a stablecoin like USDT or USDC?

No, and filing it in the same drawer is a genuinely costly mistake. USDe is a synthetic dollar: it holds crypto collateral hedged one-for-one with short perpetual-futures positions, so its dollar value is engineered by a live trading strategy rather than backed by cash and Treasuries. Its staked form, sUSDe, earns yield from perpetual funding rates and staking rewards — which means the yield and the peg depend on the same strategy. Its risks are correspondingly different: funding rates can turn negative, the hedging positions sit with centralised venues and custodians, and fast markets stress the basis. It has also faced regulatory action in Germany affecting its European availability, which readers there should verify as of the time they read this. USDe can be a deliberate, understood allocation; it is not a cash-backed dollar.

Is USDT audited?

As of July 2026 the precise answer is: attestation-based, with the first full audit under way. Tether publishes quarterly attestations from BDO Italia — point-in-time confirmations rather than a full financial-statement audit — and its first-quarter 2026 report, as of 31 March 2026, showed roughly 191.7 billion dollars of reserve assets against about 183.5 billion dollars of liabilities, around 104.5 per cent coverage, with US Treasuries the dominant holding. On 24 March 2026 Tether announced it had engaged a Big Four firm, reported to be KPMG, for its first full independent audit — a process that had not been completed at the time of writing. Both of the popular one-liners are therefore wrong: neither has USDT simply never been audited with nothing changing, nor is it audited today.

How can I tell if a stablecoin is risky before it depegs?

Run it through four criteria, each drawn from a real failure. Reserve transparency: who verifies the reserves, how often, and is it an attestation or a full audit? Redemption terms: who can actually redeem at par, at what minimum size, and what rights does the issuer have to gate redemptions? Jurisdiction: does the issuer answer to a real regulatory regime where you live? Concentration: is there a single yield sink, a single bank, or a single venue whose failure breaks the whole structure — Anchor for UST, Silicon Valley Bank for USDC's cash, the hedging venues for a synthetic dollar. Then watch the live tells: a sustained price below roughly 98 cents, redemption queues lengthening or pausing, overdue attestations, and above-market yield with no visible source. UST displayed several of these for a year before it broke.

Are stablecoins regulated now?

Increasingly, but unevenly, and the stages matter as of July 2026. In the EU, MiCA's stablecoin rules have applied since 30 June 2024 and already determine which coins compliant exchanges may sell. In the US, the GENIUS Act was signed on 18 July 2025 but is not yet in force — it takes effect by 18 January 2027 at the latest, or 120 days after final regulator rules if sooner, and only proposed rules existed as of mid-2026. In the UK, the FCA published final issuer rules on 30 June 2026, with the regime coming into force on 25 October 2027. Two cautions travel with all three: regulation does not guarantee a peg — it improves transparency and issuer standing, which are inputs to your judgement, not substitutes for it — and regulatory status is time-sensitive, so always check the current position.

Related resources

• Stablecoins: the complete guide — the cluster hub: the three designs, how reserves work, the full risk framework and how to choose a coin deliberately.
• Stablecoin regulation guide — the full mechanics behind the jurisdiction criterion: GENIUS, MiCA and the UK rules, with dates and worked examples.
• Earning yield on stablecoins — where stablecoin yield comes from, engine by engine, and the red flags that Anchor's twenty per cent should have raised.
• USDT vs USDC comparison — the dated fact-rows behind this page's stories: reserves, attestations, depeg history and jurisdiction access side by side.

Financial Disclaimer

This content is not financial advice. All information provided is for educational purposes only. Cryptocurrency investments carry significant investment risk, and past performance does not guarantee future results. Always do your own research and consult a qualified financial advisor before making investment decisions.