Is It Safe to Keep Your Money in Stablecoins? An Honest Guide to the Real Risks

Why we are the ones telling you this

Most companies in this category solve the risk problem by not mentioning it. When we reviewed the public help pages of several companies in this category in August 2026, we could not find the words stablecoin, depeg or FDIC on the pages a customer actually reads: the product is described as "digital dollars" and the security section stops at licensing and encryption. Pages change, and we are describing what we found on the date we looked — not making a claim about any company today.

That is a defensible commercial strategy and a bad deal for the person holding the balance. If you do not know which asset you own, you cannot look up its risks.

We take the opposite position, and not out of virtue: our own published disclaimer already says out loud what most of this industry buries. It states that we are a technology and software provider and not a bank, that the value of a digital asset "may fluctuate and may even go to zero", that we do not guarantee parity, liquidity, convertibility or backing, and that stablecoin balances are covered by no deposit insurance scheme. That document is public, it is linked from every page of this site, and everything in this article is consistent with it.

What follows is what can actually go wrong. Not to talk you out of anything, and not to sell you anything — so that when you decide, you decide with the full picture. This is educational content, not financial, legal, tax or investment advice.

Three kinds of "stablecoin", and only one is what you probably think

This distinction does more work than everything else in this article combined, because the disasters cluster almost perfectly by category.

Category	Examples	What is behind it	Worst historical outcome
Fiat- and Treasury-backed	USDC, USDT	Cash at banks, US Treasury bills and overnight repo against Treasuries; USDT's reserve also includes gold, bitcoin and secured loans	Temporary depegs; the largest tokens returned to par, but not every token in this category fully recovered
Algorithmic	UST (Terra)	Nothing off-chain — a smart contract that swapped 1 UST for $1 of LUNA, a token from the same system	Total, permanent loss
Synthetic / yield-bearing	USDe, xUSD, deUSD, msUSD	Hedged derivative positions or leveraged strategies run by third-party managers	Collapses, frozen withdrawals, issuers winding down

UST is the case worth understanding properly, because it is the one everybody half-remembers. Researchers at MIT and LSE described it as convertible debt of infinite maturity with a $1 face value backed by LUNA, and explicitly contrasted it with Tether and Circle, which are "backed by off-chain liquid assets" — UST "was not supported by off-chain collateral" (Liu, Schoar and Makarov, published via the Harvard Law School Forum on Corporate Governance, May 2023).

Its demand engine was Anchor Protocol, paying depositors 19.5% a year — a subsidised return, not a generated one. By April 2022 Anchor was paying out roughly $6 million a day, most of it drawn from a yield reserve that was visibly draining rather than earned by the protocol, and the community was already debating cutting the rate. That is the reusable warning sign: if a "stable" dollar pays far above market rates, somebody is funding that gap, and that somebody can run out of money.

The run started on 7 May 2022 with two large withdrawals from Anchor, and the collapse took days, not weeks. As the mechanism minted LUNA to defend the peg, LUNA's supply expanded by orders of magnitude and its price fell to near zero. The US Department of Justice puts investor losses at more than $40 billion. Do Kwon pleaded guilty to fraud on 12 August 2025 and was sentenced in December 2025 (US Attorney's Office, SDNY).

One more detail from the DOJ indictment that reframes the whole episode: UST had already lost its peg in May 2021, falling below $0.92, and was quietly rescued by a high-frequency trading firm buying UST at scale — while Kwon publicly credited the algorithm. The proof that the mechanism worked was manufactured, and that manufactured confidence is what fed the 2022 catastrophe.

The 2025–2026 wave landed in the third category. Several yield-bearing tokens marketed as dollars lost most of their value; in some cases withdrawals were frozen, in others the issuer wound the token down. We do not publish figures for the size of those falls: the estimates we found varied too widely to rely on.

None of those were bank-style failures of a dollar reserve. They were failures of a strategy. A token pegged to a dollar and a token that earns you money are different products wearing the same word.

Losing the peg: what actually happened, and what it cost

Here is the depeg record for fiat-backed stablecoins, which is the category most people actually hold. Read the last column first. Every low below is an exchange-level print on the venue named, not a single global price. Each row carries its source and the date we consulted it; where we could not attach a verifiable source to a print, we left the row out.

Token	Date	Low (and where)	Cause	Outcome
USDT	15 Oct 2018	~$0.85 on Kraken, ~$0.87 on Bittrex	Concerns over Tether/Bitfinex banking	Recovered
USDT	12 May 2022	~$0.95 on Binance.US, ~$0.92 on Kraken	Contagion from the UST collapse	Back to ~$0.998 the same afternoon
USDC	11 Mar 2023	~$0.87	$3.3B of cash reserves stuck at Silicon Valley Bank	Par restored 13 Mar; $3.8B redeemed 13–15 Mar
USDT	15 Jun 2023	~$0.996	Liquidity imbalance in Curve's 3pool (>70% share)	Not a reserve event at all
TUSD	15–16 Jan 2024	~$0.97, low of $0.926 on Poloniex	Loss of structural demand plus attestation problems	Partial recovery
FDUSD	2 Apr 2025	~$0.87	Public doubts about the issuer's strength, which the issuer denied	~$0.99 the next day after redemptions were honoured

The USDC episode is the reference case, and its details are more instructive than its headline. On 11 March 2023 Circle disclosed that SVB was one of its six custodial banks and held $3.3 billion — around 8% of total reserves, per the Federal Reserve's own December 2025 research note. The token fell over a weekend, hourly trading volume approached $2 billion on 11 March concentrated in decentralised exchanges, and the contagion reached tokens with zero SVB exposure: DAI fell to roughly $0.88–0.90 because more than half its collateral was USDC. In the same episode, USDT and BUSD traded slightly above $1.00 on flight-to-safety flows.

Three honest readings of that weekend:

1. It was resolved by a discretionary government decision, not by the stablecoin's own mechanism. On Sunday 12 March, Treasury, the Fed and the FDIC guaranteed that all SVB depositors would be made whole — beyond the ordinary $250,000 insurance limit. That is not a protection you can count on repeating.
2. The redemption machine worked as soon as banks reopened. Circle redeemed $3.8 billion and issued $800 million between 13 and 15 March. Whoever waited was paid $1.00; whoever panic-sold on Saturday realised the loss themselves.
3. The risk was never in the Treasury bills. It was in the slice of cash sitting in a commercial bank. Bank risk is the real risk of a fiat-backed stablecoin.

The Federal Reserve economists' conclusion is the honest anchor for this whole article: stablecoins with high-quality backing hold their peg "during normal times" but remain "fragile during periods of significant stress". Fiat backing reduces the risk. It does not eliminate it.

⚠️ There is no such thing as "the" price of a stablecoin. Every exchange and every pool had its own low. On 12 May 2022 USDT printed around $0.92 on Kraken while trading near $0.95 on Binance.US the same day. On 10 October 2025 USDe printed well below $1.00 on Binance for a matter of minutes, during one of the largest market-wide liquidation events on record, while trading near $0.99–1.00 on-chain and on other venues; Ethena publicly attributed that print to the exchange's internal oracle reading its own thin order book. Be suspicious of dramatic lows quoted without a venue attached.

The practical rule: a depeg is not determined by the price on a screen, it is determined by whether you can still redeem 1:1 at the source. USDC came back because there were dollars to claim. UST did not because there was nothing to claim.

Reserves and attestations: what the accountant actually signs

This is where marketing does the most damage, because "audited" and "attested" get used interchangeably and they are not the same thing.

USDC. The most recent examination report, signed 27 August 2026, covers two dates — 8 and 31 July 2026 — and reports 71,826,453,410 USDC in circulation against 71,904,101,332 USD of reserve assets. Circle publishes detail down to the CUSIP of each individual Treasury bill, a level of granularity we did not find in the public reports of the other large issuers we reviewed in August 2026. A detail worth knowing: as of 31 July 2026 that reserve is not mostly directly-held T-bills. By our own arithmetic on the report's figures, roughly 73% is overnight repo against Treasuries, about 16% cash, and only about 11% T-bills held outright. Overnight repo is over-collateralised short-term lending, not outright ownership — a real nuance, not a scandal.

USDT. Tether reported $187,751,426,411 in assets against $183,641,897,215 in liabilities as of 30 June 2026, leaving a reserve buffer of $4,109,529,196 — roughly half the $8.23 billion of the prior quarter. The announcement does not explain the drop, and we are not going to guess at a cause. The structural point stands on its own: around 20% of those assets — gold, bitcoin and secured loans — are not cash equivalents, so part of the backing moves with the market. That is a different risk profile from a reserve held entirely in cash and Treasuries, and it is worth knowing which one you are holding.

Now the part almost nobody explains. Here is what each document actually covers:

	Monthly USDC attestation	Quarterly USDT attestation	A full financial audit
What is examined	Management's assertion that reserve fair value ≥ USDC in circulation	The reserves report and total assets/liabilities at one date	The complete financial statements for a full year
Who sets the criteria	Circle's management, defined inside the report	Company management	An accounting framework (US GAAP / IFRS)
Assurance level	Reasonable assurance (AICPA examination)	Reasonable assurance (ISAE 3000 Revised)	Reasonable assurance
Time covered	Two timestamps a month, 23:59 UTC	One date per quarter	The whole financial year
Internal control	Not opined on	Not opined on	Understood and considered
Going concern	Not addressed	Explicitly excluded	Assessed

The scope limitation in Tether's report is stated by the auditor itself, in plain language: "Activity prior to and after this time and date was not considered when testing the balances." Translated for a reader: if a reserve were short on 29 June and covered on the 30th with money returned on 1 July, the attestation would still be technically correct. That is not an accusation — it is what the document says it does not cover.

BDO goes further and excludes the two things a worried holder most wants to know. It states that it provides no assurance on management's going-concern assessment, and that asset valuation "is based on normal trading conditions and does not reflect unexpected and extraordinary market conditions, or the case of key custodians or counterparties experiencing substantial illiquidity". In other words: can the reserve survive a run? is precisely the question an attestation declines to answer. The Notes section — the narrative part the press quotes most — is expressly outside the scope of the assurance engagement.

To be fair in both directions: an ISAE 3000 attestation is not a rubber stamp. BDO performs direct confirmations with banks and custodians, physically inventories and assays precious metals on a sample basis, reconciles against on-chain ledgers, and verifies collateral on a sample of secured loans. The correct criticism is not "they check nothing" — it is "they check existence and valuation at one instant, against criteria management defines, without opining on internal control as a system, on the rest of the period, or on the company's viability."

And history shows what only a regulator finds. In October 2021 the CFTC ordered Tether to pay $41 million and Bitfinex $1.5 million over misleading statements about USDT's backing, finding that Tether held sufficient fiat reserves to back USDT for only 27.6% of the days in a 26-month sample period from 2016 through 2018. Eight months earlier, Tether and Bitfinex had paid $18.5 million to settle a New York Attorney General investigation — and Tether's quarterly reporting obligation originates from that settlement, not from a voluntary transparency initiative. On Circle's side, the enforcement history relates to Poloniex, the exchange it owned in 2018–2019 (about $10.4 million with the SEC in 2021 and $7.59 million with OFAC in May 2023 over 65,942 transactions involving sanctioned jurisdictions); there is no known enforcement action against Circle over USDC's reserves.

The two issuers also sit under different legal roofs, which is a comparison we go into properly in USDC vs USDT in Latin America. USDC is issued by subsidiaries of a NYSE-listed company that files with the SEC; USDT is issued by Tether International, S.A. de C.V., a Salvadoran company that relocated from the British Virgin Islands in January 2025.

The risk nobody publishes: your address can be frozen

This is the heart of the article, and the part least often explained anywhere.

Both major issuers can freeze balances, and the capability is written into the smart contract. USDT exposes addBlackList(), removeBlackList() and destroyBlackFunds(), executable only by Tether's owning multisig. A blacklisted address cannot transfer anything. destroyBlackFunds() permanently burns the balance of a frozen address. USDC has an equivalent blacklist mechanism.

	USDT	USDC
Can freeze at contract level	Yes	Yes
Published scale	The issuer does not publish a periodic total of blocked balances	129,764,354 USDC as "Access Denied Tokens" (Circle reserve report, 31 Jul 2026)
Issuer's own figure	2,300+ cases with 340+ agencies in 65 countries, $4.4B+ frozen (Tether, 23 Apr 2026)	Circle deducts restricted tokens from circulating supply in its own report
Declared policy	Acts "immediately and decisively" on credible links to sanctioned entities or criminal networks, including its own monitoring	Freezes "only at the direction of law enforcement or the courts" (CEO Jeremy Allaire, April 2026)

The two published figures above are not measured the same way and are not directly comparable — one is an address count from a third-party on-chain analysis, the other is a dollar amount Circle deducts from its own circulating supply. What they do show is a large difference in scale, and that difference measures compliance policy and usage profile, not architecture. The correct conclusion is not "USDC is safer against freezes" — it is that both can do it, and the difference is judgement, not capability.

Three things make this more serious than a bank freeze:

It can destroy and reissue, not just block. Tether has publicly described burning frozen USDT and reissuing an equivalent amount to addresses designated by courts or victims. We describe the capability, not its scale: we could not verify a figure for how much was handled that way. An issuer can not only stop you moving a balance, it can extinguish it and mint an equivalent to somebody else. There is no equivalent power in bitcoin or ether.

Innocent third parties get caught. A block lands on an address, not on a guilty person. When that address is a company's operating wallet, the effect reaches every one of its customers; when it is a contract, everyone with a balance behind it loses access at once. And when the court order behind the freeze is sealed, the person affected may not even be able to obtain an explanation of why their money stopped moving.

And it is not a magic button either. On-chain analyses of freeze proposals have found that a meaningful share of the targeted funds moves out before the freeze executes: bots watch the multisig and react in minutes. For a legitimate user, the mirror image of that finding is worse — an erroneous freeze is not reversed with one click either.

The counterweight, in fairness: issuers also face the opposite claim — that they should have frozen and failed to do so in time. After large on-chain thefts, that argument has already been taken to court by people who lost money. The same power that reads as censorship in one story is demanded as a duty in the next, and neither outcome is designed around protecting you.

In the US this is now mandatory. The GENIUS Act prohibits issuing a payment stablecoin without the technological capability to comply with legal orders to seize, freeze, burn or block transfers, and FinCEN and OFAC proposed rules in April 2026 extending that obligation into secondary markets. The counterintuitive conclusion: a US-regulated stablecoin that could not be frozen would be illegal. Anyone promising you "a stablecoin nobody can freeze" is describing a product that cannot be regulated in the largest market on earth.

Our position on this is section 13 of our disclaimer, and it is deliberately blunt: D1C is not liable for freezes, blocks, holds, AML reviews or court orders adopted by authorities, banks or regulated providers that affect your account or your transactions. We would rather you read that here than discover it on the day it happens.

Who holds the keys, and why a mistake has no undo

Two questions with uncomfortable answers, and both are already written into our published disclaimer.

We are not the custodian. Section 3: digital assets are custodied by Bridge; D1C does not custody your assets, has no access to private keys, does not move funds on its own initiative, and does not guarantee redemptions or the sufficiency of reserves. Section 1: DIGITALROCKETS LLC, a Florida limited liability company, is a technology and software provider — not a bank, not a deposit institution, not a money services business or money transmitter, not a custodian, not an investment adviser. Section 2: the registered MSB status with FinCEN belongs to Bridge Building Inc., not to us; account services may be provided by different Bridge entities depending on your region — for Latin American users, by Bridge Building Limited. Section 14: using the service creates no banking, fiduciary, custodial or advisory relationship.

That chain matters when something goes wrong, because it tells you who can actually act. It also explains section 12: we do not control and are not responsible for the acts of third parties who provide the financial services — stablecoin issuers explicitly included. If Circle freezes an address, we cannot unfreeze it. Saying otherwise would be a comfortable lie. The mechanics of how that infrastructure produces a dollar balance are in digital dollars explained and virtual dollar accounts in Latin America.

And nothing on a blockchain can be undone. Section 6: digital asset transactions "are in no case reversible". Once broadcast, an operation cannot be cancelled, reversed or modified — and a transfer sent to the wrong address or over an incompatible network is not recoverable. Not by us, not by Bridge, not by support, not by a court in practical terms.

This is the risk that actually costs ordinary people money, far more often than a depeg. A transfer usually completes in under two minutes, and those two minutes are the whole window of consequence. The address is not the only thing to check: the network is. A perfectly valid address on the wrong chain is a permanent loss. Our own send flow marks Base as the recommended network for a reason — one recommended network is one fewer thing to get wrong.

🔎 The practical habit: before any meaningful transfer to a new destination, send a small test amount, wait for it to land, and only then send the rest. The send fee is 0.40%, so on a $20 test that costs eight cents. It is the cheapest precaution in this entire article — and, as the next section explains, a precaution is all it is: nothing here is insured.

No insurance covers this. None.

Section 5 of our disclaimer, verbatim in substance: digital asset balances and services are not covered by the Federal Deposit Insurance Corporation (FDIC), the Securities Investor Protection Corporation (SIPC), Colombia's Fondo de Garantías de Instituciones Financieras (Fogafín), or any other deposit, investor or government-backed insurance scheme.

That is not a defensive clause we invented. It is the state of the law in every jurisdiction we could verify, as of 28 August 2026:

Jurisdiction	Scheme	Covers a stablecoin balance?	What exists instead
United States	FDIC / SIPC	No — prohibited by statute	1:1 reserves, redemption right, bankruptcy priority over reserves
European Union	Deposit guarantee (DGSD)	No — e-money tokens sit outside its perimeter by design	1:1 reserves with a minimum in bank deposits, segregation, free redemption at par
United Kingdom	FSCS	No — "FSCS protection will not cover the value of a stablecoin"	Backing, segregation, statutory trust, custody and liquidity rules
Mexico	IPAB	No — no stablecoin framework in force	Nothing specific; virtual assets are not legal tender
Colombia	Fogafín	No	Nothing specific; the regulating bill was archived in 2026
Peru	Fondo de Seguro de Depósitos	No	AML supervision of providers, which is not prudential protection

Some detail worth having, because the nuances are where people get misled:

The US closed the back door explicitly. The GENIUS Act, signed 18 July 2025, prohibits representing a payment stablecoin as backed by the full faith and credit of the United States, government-guaranteed, or subject to federal deposit insurance. The same section fines marketing a product as a payment stablecoin when it was not issued under the Act at up to $500,000 per violation. Then in April 2026 the FDIC proposed that reserve deposits backing a stablecoin are deposits of the issuer, insured as corporate deposits — not passed through to you. FDIC Chairman Travis Hill said the quiet part in March 2026: it would be hard to rationalise the Act's firm marketing prohibition if stablecoins were meant to be an access mechanism for insured accounts.

What US law does give you is real but is not insurance: 1:1 reserves in defined assets, a redemption right, a prohibition on the issuer paying yield, and a claim on reserve assets outside the bankruptcy estate. Even that priority is academically contested — legal scholarship has argued that in practice holders could rank well below first, once other statutory and administrative claims are accounted for. The honest framing is that the law gives holders standing over the reserves, and that standing is a lawsuit with timelines and haircut risk, not a payout.

And a marketing trap worth dismantling: "we keep the dollars at FDIC-insured banks" does not mean you are insured. A dollar balance credited on a platform is generally a claim against that platform. And for stablecoins specifically, even the pass-through route is being closed by rule.

One last timing caveat, because this section ages fast. As of 28 August 2026, no US federal agency has issued final GENIUS Act rules — several agencies published proposals and regulators missed the statutory one-year deadline of 18 July 2026, with the OCC targeting November 2026. Absent final rules, the Act's default operative date is 18 January 2027. In the EU, MiCA is already fully applicable and under an open review consultation. In Mexico there is a Senate initiative from May 2026 with no committee report and no framework in force; Banxico, the SHCP and the CNBV maintain that virtual assets are not legal tender and carry no backing from any financial authority or government institution. In Colombia the bill regulating virtual asset providers was archived in August 2026 — while the tax authority adopted the OECD's crypto reporting framework, with data collection from the 2026 tax year. Read that last pair together: in Colombia, reporting obligations on providers are advancing faster than any protection for holders. What that means for your own tax position is a question for a professional — nothing here is tax advice.

The risk in your account, not just in the asset

Everything above is about the token. This section is about you, and statistically it is the more likely way to lose money.

Sign-in uses a one-time code sent to your email. There is no password. That design removes the risk of a reused or leaked password, and it concentrates everything into one place: whoever controls your inbox controls your account. Our own documentation says it plainly and tells you what to do about it — protect that mailbox with two-factor authentication. If you take one operational action after reading this article, make it that one.

The rest of the security stack: funds are held under institutional custody at Bridge.xyz, with AES-256 encryption and TLS 1.3; identity verification runs through Persona (operated by Bridge), which is covered in what KYC verification is; and the mobile app supports Face ID and fingerprint. We will never ask you for your access codes, on any channel. Anyone who does is not us.

Two structural risks worth stating alongside the technical ones. Availability is not guaranteed: geographic and regulatory limits apply, the service is unavailable to sanctioned persons and jurisdictions (OFAC, UN, EU), and payout rails and currencies are enabled progressively country by country — if an option is not yet active in your account, the app shows it as "in progress". And the regulatory treatment of digital assets is evolving and uncertain; changes in law can affect availability, use or value. Both are sections 8 and 9 of the disclaimer, and neither is something we can promise our way out of.

What a careful reader actually does

This is not financial, legal, tax or investment advice, and it is not a recommendation to hold, buy or allocate anything. It is a list of operational habits that reduce the specific, named risks described above — each row is tied to a failure mode, not to a decision about your money.

Habit	The risk it addresses
Know which token you hold, by name	You cannot research the risk of an asset you only know as "digital dollars"
Treat an unexplained yield as a risk signal, not a feature	Anchor's advertised 19.5%, xUSD, deUSD — when a return is subsidised rather than generated, the subsidy ends before the marketing does. This describes a risk; it is not a recommendation about what to hold
Read the issuer's report, not the press coverage	Circle publishes CUSIP-level detail; Tether publishes quarterly plus one annual audit. Both are public
Understand that issuer risk is concentrated	A freeze, a lawsuit or a bank problem hits one issuer at a time, so a balance in a single token concentrates that exposure. This describes the risk; it is not a recommendation about what to hold. Note also that at D-ONE CASH the balance is credited in USDC, so this exposure is not one you pick per operation
Check the network before every send, and test with a small amount first	Wrong network is the one mistake with no undo
Turn on 2FA for the email address on your account	The one-time code is the only key to the account
Know your exit before you need it	Bank withdrawals are usually processed in 1–2 business days, with a 20 USD minimum amount and the exchange rate applied when the order is processed. The currencies you can actually withdraw to a bank in are USD, EUR, MXN and COP — in any other the app will not let the withdrawal complete, even where a fee is published — and rail availability is enabled progressively by country: if an option is not yet active on your account, the app shows it as "in progress"
Know what a balance is, and is not	A stablecoin balance is a payment tool. It is not a deposit, it is not insured, and it does not become either by sitting still. How much to hold is your decision, and nothing here is a recommendation about it

If you want the full cost picture rather than the risk picture, the leg-by-leg arithmetic is in the true cost of sending $200.

D-ONE CASH is a product of DIGITALROCKETS LLC. We are not a bank, a deposit institution, a custodian or an investment adviser, and we are not legal, tax or investment advisors. Everything in this article about our product is verified against our published documentation and our disclaimer; third-party facts are attributed to their source and date wherever we could verify them, and where we could not we have said so or left the figure out. This is educational content, not financial, legal, tax or investment advice.

Frequently asked questions

Are stablecoins FDIC-insured like a bank account?

No, and in the United States that is now positive law rather than interpretation. The GENIUS Act prohibits representing a payment stablecoin as government-guaranteed or subject to federal deposit insurance, and in April 2026 the FDIC proposed that reserve deposits backing a stablecoin are insured to the issuer as corporate deposits, not passed through to holders. The same answer holds elsewhere: EU e-money tokens sit outside the deposit guarantee perimeter by design, the UK's FSCS explicitly does not cover the value of a stablecoin, and Mexico, Colombia and Peru have no framework at all. What exists instead is structural — 1:1 reserves, segregation, a redemption right, and in the US a claim on reserve assets in a bankruptcy. That reduces risk; it does not transfer it to the state.

What happens to my money if Circle or Tether goes bankrupt?

Two very different scenarios get confused here. If the issuer fails, you are a claimant against the reserve assets — under the GENIUS Act those assets sit outside the bankruptcy estate and holders get a priority claim, though how strong that priority really is remains academically contested. Either way it is a legal process with timelines and haircut risk, not a payout. If the platform where you hold the balance fails, the exposure is different and historically worse: an unsecured claim against a company, which is what holders of Celsius, BlockFi, Voyager and FTX Earn balances discovered in 2022. Our own disclaimer is explicit that D1C does not custody assets — custody sits with Bridge — and that we do not guarantee redemptions or the sufficiency of any issuer's reserves.

Can I lose money if USDC or USDT loses its peg to the dollar?

You can, but the mechanism matters more than the headline. In March 2023 USDC traded around $0.87 after $3.3 billion of its cash reserves were stuck at Silicon Valley Bank; par was restored within roughly 72 hours and Circle redeemed $3.8 billion between 13 and 15 March. Whoever waited was paid $1.00; whoever sold at the bottom converted a scare into a realised loss. USDT has had several episodes — around $0.85 in October 2018, around $0.92–0.95 in May 2022 — and recovered each time. UST never recovered, because it had no off-chain collateral to claim. The test is not the price on a screen, it is whether redemption at par is still open at the source. And note there is no single price: dramatic lows usually come from illiquid venues.

Can my stablecoins be frozen or blacklisted?

Yes. Both USDT and USDC can freeze balances at the smart-contract level, and USDT can additionally burn a frozen balance and reissue it to an address designated by a court. Circle publishes the figure in its own reserve report: 129,764,354 USDC restricted by legal order as of 31 July 2026. How much each issuer freezes reflects its compliance policy and user base, not a difference in capability. It also reaches innocent parties, because a block lands on an address: if that address is a company's operating wallet or a smart contract, everyone behind it loses access at once. In the US, this capability is legally required — an unfreezable payment stablecoin would be illegal to issue. Our disclaimer states that D1C is not liable for freezes, blocks, holds, AML reviews or court orders.

Is it safer to keep stablecoins in an app or in my own wallet?

Neither is safe in the abstract, because they fail in different ways. With self-custody you carry the key risk: lose the seed phrase and no support team on earth can recover the funds, and an on-chain mistake is irreversible. With an app where you do not hold the keys you carry counterparty risk: the solvency of whoever actually holds the assets, the access controls and the compliance decisions. In that structure the custodian is Bridge, not D1C. In our case that split is on the record — Bridge holds custody, D1C has no access to private keys and does not move funds on its own initiative, and the balance is not a bank deposit in either arrangement. What is identical in both models is the issuer risk: a freeze or a depeg reaches your balance regardless of who holds the keys.

Are high advertised yields on "stablecoin" products safe?

Treat a high advertised return on a "stable" dollar as the risk signal, not the feature. UST paid 19.5% through Anchor, with payouts of roughly $6 million a day by April 2022 drawn from a draining yield reserve, and the collapse cost investors more than $40 billion by the DOJ's figure. The 2025–2026 wave — xUSD, deUSD, msUSD — was the same shape: yield generated by third-party strategies, not by the dollar itself. Under the GENIUS Act, US-permitted issuers are prohibited from paying interest or yield on a payment stablecoin at all, which means anyone paying you a return is not the issuer and the risk lives entirely with whoever is paying. We do not offer yield on balances and this article is not investment advice — the point is only that the question to ask is always "who is funding this, and what happens when they cannot?"