Stablecoin yield in DeFi

The research was done by AI agents that open web pages. Some sites refuse them; where that happened we say so. The small numbers point to the source list at the end. Where a figure is our own sum, it is marked “our estimate” and the basis is given.

The claim

Our collection for this scheme holds one tweet¹. That is too few to show how the scheme is promoted, so this page rests on that one example and on how the method is offered generally: by exchanges, protocols, course sellers and the products regulators have acted on. The page judges the scheme, not the person who tweeted.

“If you want to get rich with crypto:

Get a job and save $2,500

Convert it all to USDC

Invest in 5 high-yield DeFi protocols

Earn $500–$1,000 in passive income every month

Reinvest the profits

Do this consistently, and you could become a dollar millionaire.”

@BitcoinIntelX, 4 October 2026, 20,470 views as collected¹

The tweet ends with a link to attached media, which a mirror of the tweet describes as a 16.6-second video². We did not view it. So we do not know whether it names the five protocols, gives a rate or shows any earnings. The tweet text gives two numbers, the money in and the money out. It does not give the rate, a time period, fees, tax or any risk of loss¹.

What the scheme is

A stablecoin is a crypto token meant to stay worth one US dollar. USDC, the one in the tweet, is issued by a company called Circle, which holds dollars and short-term US government debt behind it.

DeFi (“decentralised finance”) means financial services run by programs on a blockchain, with no bank in the middle. These programs are called protocols or smart contracts. The main ones for this scheme are lending protocols such as Aave, Compound, Morpho and Sky. You deposit stablecoins. Borrowers, who put up other crypto as security (called collateral), pay interest. Most of that interest goes to depositors.

Yield is the yearly return on the deposit. It is usually shown as APY, the yearly percentage including interest on interest. A pool or vault is one pot of deposits with one rate.

Holding USDC by itself pays nothing. Circle’s terms say: “you acknowledge that you are not entitled to any interest or other returns earned on such funds”¹⁶. Any yield comes from lending the coin out or taking some other risk with it.

The mechanism is real. Over the past 12 months Aave paid depositors about $620 million, Morpho about $206 million, Ethena about $243 million and Sky about $189 million, on DefiLlama’s figures⁴⁰. DefiLlama is a public data site that tracks these protocols. The Aave and Morpho figures cover all deposited assets, not stablecoins only, and they are DefiLlama’s estimates, not audited accounts⁴⁰. The money comes from borrowers and, for Ethena, from fees paid between traders of futures (contracts that bet on a future price). It does not depend on new depositors arriving.

So the question is what rate it pays, and what the tweet’s income would need.

The arithmetic

Step 1: the rate the claim needs

Monthly income = money deposited x yearly rate ÷ 12. The tweet gives the first and the last and leaves out the rate. Working backwards (our estimate, from the tweet’s own figures¹):

• $500 a month on $2,500 is 20% a month. $1,000 a month is 40% a month.
• Over a year that is 240% to 480% without reinvesting.
• With profits reinvested each month, as the tweet says, it is about 790% to 5,570% a year (1.2 to the power 12, and 1.4 to the power 12).
• At those rates $2,500 would pass $1 million in about 18 to 33 months.

Per day, 20% to 40% a month is about 0.7% to 1.3%.

This reads the tweet as one $2,500 deposit. A kinder reading, saving $2,500 again and again, is tested in Step 4.

Step 2: what stablecoin deposits pay today

We downloaded DefiLlama’s live list of yield pools on 10 October 2026 at 14:34 UTC and worked out the figures below ourselves⁴. “Median” means the middle pool: half pay more, half less.

All figures in the table are from source⁴.

A median counts pools, not money. The average weighted by the money in each pool says more about what a typical deposited dollar earns. It is about 4.0%, leaving out the 92 pools that DefiLlama marks as “outliers” (readings it treats as abnormal)⁴. With those left in, it is about 4.65%, and one pool that showed 898% that hour adds 0.6 points of it alone (Step 3)⁴. An earlier download the same day gave 4.9%, so the unfiltered figure moves within hours. We use 4.0%.

About two thirds of the money sits in pools paying under 5%, and nearly all the rest in pools paying 5% to 10%⁴. One large pool sitting at about 5% moves that split by three points.

The plain case is lending USDC on the largest lenders. On the Ethereum blockchain Aave paid 4.30%, Compound 4.38% and Fluid 6.16%⁴. The three largest stablecoin pools paid 3.8% (Sky’s savings token), 5.25% (Maple) and 3.23% (USYC, Circle’s fund of US government debt in token form)⁴. The highest rate among the 76 pools of $100 million or more was 13.6%, on apxUSD, a newer dollar token whose pool held $162 million⁴.

Limits of this data: it is one snapshot. Nearly a fifth of the “stablecoin” money is in tokenised Treasury funds, which are funds of US government debt sold as tokens and are not DeFi lending⁴. About 13% of the money is in pools showing a rate of zero or no rate, which pulls the average down; leaving those out gives about 4.6%⁴. Part of some rates is paid in reward tokens, extra tokens handed out by a project, whose price can fall. We did not check the rates on the protocols’ own apps.

Step 3: how rare the claimed rate is

Of the 2,997 pools, 7 showed 240% a year or more, which is 0.2%⁴. Together they held $32.7 million, or 0.075% of the money in stablecoin pools⁴. The count was 6 in a download a few hours earlier.

Most of that $32.7 million is one pool holding $30.5 million⁴. Its daily readings were 0% on 7 and 8 October, 5,085% on 9 October and 898% on the 10th⁴. Our reading is that this is a spike in the reported number and not a rate anyone earned for a month. We did not open that protocol’s own page to confirm. The other six pools held between about $12,000 and $2 million each⁴. DefiLlama marks all seven as outliers, and its own forecast for every one of the ten highest-paying pools is “Down”⁴.

Lower down, 114 pools (3.8%) showed 20% a year or more, holding 0.19% of the money⁴. The tweet needs 20% every month; these pools show 20% for a whole year.

So “five high-yield protocols” paying 240% or more do not exist at any real size today.

Step 4: what $2,500 earns, and what the income needs

These are our sums from the rates in Steps 2, 5 and 6⁴⁵⁹.

The tweet’s income is 60 to 120 times what the average rate pays on $2,500 (our estimate: $500 and $1,000 against about $8).

The “dollar millionaire” line, read two ways (our estimates).

• One deposit. At 4.0% a year with interest reinvested and no new deposits, $2,500 takes about 150 years to reach $1 million. At 10% a year for ten years it grows to about $6,500.
• Saving $2,500 every month. The tweet says “Do this consistently”, which can mean saving the sum again and again. At 4.0%, monthly deposits of $2,500 reach $150,000, enough for $500 a month, after about four and a half years. They reach $1 million after about 21 years, of which about $637,500 is the saver’s own deposits. On this reading the outcome comes mostly from the saving, and $2,500 still does not earn $500 a month.

Step 5: have rates ever been near the claim?

No. They have been higher than today, for a while.

These are our averages of DefiLlama’s daily readings⁵. They are not what any one wallet earned, and they leave out fees. Today’s rates of 4.2% to 4.4% on Aave and Compound are above the 2026 average so far⁴⁵.

On single days Aave’s USDC rate touched 56.7% in 2023 and 41.4% in 2024, read as a yearly rate⁵. The best sustained stretch for a large product was sUSDe in early 2024. It averaged about 38% a year across March 2024, roughly 2.7% for that month, and about 23% a year in April (our estimate from⁵). Its average for 2024 from launch was about 17.5% a year, or about 1.4% a month⁵. The tweet needs 20% to 40% a month. The best month on record paid about a tenth of that.

Ethena’s own documents explain why its rate moves. Its income is a fee that traders in futures markets pay each other, called funding. Ethena writes that it “is able to earn revenue from funding, but could also be required to pay funding”, and that in its historical data the fee was negative on 17.5% of days for one asset and 15.9% for another, with a long-run average of 9.15% and 7.80% a year⁵³.

Step 6: the comparisons

• US government debt. A Treasury bill is a short loan to the US government. A 4-week bill yielded 4.03% and a 13-week bill 4.16% on 9 October 2026⁶. The money-weighted DeFi rate of about 4.0% is the same or slightly less, and that average itself includes tokenised Treasury funds⁴⁶. So the extra pay on $2,500 is roughly nil. Coin Metrics, a crypto data firm, reported in September 2026 that USDC on Aave had paid less than a one-year Treasury bill for 78% of 2026 so far¹³.
• A bank. The US national average savings rate is 0.37%, and the average 12-month certificate of deposit (a fixed-term savings account) pays 1.73%⁷. Stablecoin lending pays roughly ten times the average savings account. But that average is mostly branch accounts. Online savings accounts with US deposit insurance up to $250,000 paid about 4.25% to 4.34% with no minimum balance on a list dated 4 October 2026⁶⁰. That is the same as or more than USDC on Aave. The true part of the pitch is that stablecoin lending beats an ordinary branch savings account. It does not beat an online one.
• Routes through an exchange app. These need no wallet skills. Rates are from a consumer finance site’s July 2026 list unless marked⁹.
  • Coinbase. Its own page offers 3.75% on USDC, only to members of its paid Coinbase One plan, “with plans starting at $4.99/month”⁸. The July list gave 3.50% and a yearly price of $49.99⁹. On $2,500, 3.75% is $7.81 a month. After the monthly fee that leaves about $2.82 a month. On the yearly plan it leaves about $44 a year (our estimate).
  • Kraken. “Anyone can earn 1.75% on their USDC with Kraken”, with no membership⁹. On $2,500 that is about $3.65 a month (our estimate).
  • Robinhood Earn. It advertises an “estimated 7% APY” on a dollar token called USDG, with a $1 minimum and no paid membership, and it is not offered in New York or Texas⁹⁵⁹. Robinhood’s help page says: “Your USDG is deposited into a Morpho lending vault curated by Steakhouse Financial”⁵⁹. The rate “can change at any time, including dropping to zero”, and the product “involves substantial risk, including loss of principal”⁵⁹. DefiLlama showed that vault at 7.14%, of which 3.75 points were lending interest and 3.39 points reward tokens⁴. On $2,500, 7% is about $15 a month (our estimate).

Step 7: fees

Network fees are small today. Etherscan, which tracks the Ethereum network, priced a token swap at about $0.06 and a borrow at about $0.05 on 10 October 2026¹⁰. A second site, l2fees.info, showed $5.48 for a swap on Ethereum and $0.18 to $0.27 on two cheaper networks built on top of it¹¹. The two conflict for Ethereum. Etherscan’s reading is dated to the minute, so we prefer it¹⁰¹¹. If the higher figure applied, ten transactions (five deposits and five withdrawals) would cost about $50, about half of what $2,500 earns in a year at 4% (our estimate)¹¹. This is a snapshot. We did not price the cost of buying USDC with a bank transfer or card.

Splitting $2,500 across five protocols does not raise the rate. It limits what one failure can take to a fifth of the money, at the price of more transactions and five things to watch.

What people who tried it report

We found no dataset of what individual depositors end up with: no median earnings per wallet, and no share who come out ahead. What exists is market rates, a few studies of real wallets, and a record of incidents.

Personal reports. In four differently worded searches we found no published, proof-backed report from an ordinary person earning a living-size monthly income from stablecoin yield on a small deposit. That is absence of evidence, not proof. Reddit, where such threads mostly live, refused our tools both directly and through a reader service, so no forum reports were read.

A study of real wallets in a collapse. In May 2022 the stablecoin TerraUSD (UST) collapsed. Its issuer had advertised “as much as 20 percent interest through the Anchor Protocol”²⁵. This was not plain lending of a backed coin like USDC: the coin was held at $1 by a trading mechanism, not by dollars in reserve. Researchers who traced the wallets found: “Wealthier and more sophisticated investors were the first to run and experienced much smaller losses. Poorer and less sophisticated investors ran later and had larger losses.”²⁶ The paper counts $50 billion in value lost in three days²⁶. The US securities regulator counts $40 billion in market value²⁵.

A study of yield chasers. Researchers studied 262 “yield farms” (pools that pay reward tokens to depositors) on one exchange, using blockchain records from March 2021 to July 2022²⁷. The average advertised yield was 80.53%, and the average farmer put about $7,732 into 2.64 farms²⁷. They found: “investors chase farms with high yields and that farms with the highest headline rates record the most negative risk-adjusted returns.”²⁷ (“Risk-adjusted” means the return after allowing for how much the value swung.) They also found that farming was profitable on average before transaction costs, and that flat fees “are more costly for small stake investments and frequent rebalancing”²⁷. Limits: it is a working paper on one platform. Those farms held pairs of volatile tokens, not stablecoin loans, so it bears on chasing headline rates in general.

Higher-yield dollar products that failed. Two terms first. Bad debt means loans that the collateral no longer covers. An exploit is a theft that uses a flaw in a system.

• Stream Finance, November 2025. Stream offered a yield-bearing dollar token, xUSD. One commentary says it promised 18% on $382 million²⁸. On 4 November 2025 Stream said an outside fund manager had disclosed “the loss of approximately $93 million in Stream fund assets” and that it had “temporarily suspended withdrawals”²⁹³⁰. About $160 million of user deposits were frozen²⁸²⁹. The token’s price was reported at $0.26 and at a low of $0.51, readings taken at different times on 3 and 4 November²⁸²⁹³⁰. Stream’s account is that the loss was made by the outside manager and that it had hired a law firm to investigate³⁰. Almost eight months later depositors had not been paid, the token traded near 8 cents, and Stream had begun collecting creditor claims with a notice that this “does not establish or confirm any claim, entitlement, or right to payment”³¹. Vault managers on Euler, Morpho, Silo and Gearbox had accepted Stream’s tokens as collateral, so the loss spread to those lending markets³¹. A research group estimated $285 million owed across them; that is an estimate, not an audited figure³⁰³¹. What depositors in those managers’ vaults lost is not stated in the sources we read.
• Resolv, March 2026. An attacker with a stolen key created 80 million units of the dollar token USR with almost nothing behind them³³. About $25 million was taken³². The token fell to about $0.025 on one exchange within 17 minutes and was trading around $0.19 to $0.25 a day later³²³³⁵⁷. An insurer’s report says: “Eighteen audits had reviewed Resolv’s contracts. All of them confirmed the code worked as designed.”³³ The lender Fluid took more than $10 million of bad debt and committed to covering its users’ losses in full³²³³. Fifteen vaults on Morpho were hit; Morpho’s co-founder said those vaults were “all in high-risk, long-tail collateral strategies”, meaning they lent against lesser-known tokens³². Resolv’s reply, as reported, was that its collateral pool “remains fully intact”⁵⁷. It paused redemptions (swapping the token back for dollars), “meaning users could not freely exit their positions for an extended period after the event”³⁴. The Defiant, a crypto news site, wrote of such tokens: “when those stablecoins depeg, the losses fall on depositors, not on the curator.”³² (“Depeg” means the token stops being worth $1. A “curator” is the manager who chooses where a vault’s money goes.) What individual depositors lost was not counted.

The largest lender froze, then recovered. In April 2026 an exploit of about $290 million at another project left Aave with bad debt³⁵. Galaxy, a crypto research firm, put the estimate at about $123.7 million, “concentrated in the wETH reserves”, which is Aave’s market for the crypto asset ETH and not its stablecoin markets³⁵. The stablecoin markets were squeezed all the same: “USDT and USDC pools reached 100% utilization as users sold other assets for stablecoins they then withdrew”, so every deposited dollar was lent out and none was free to withdraw³⁵. “Approximately $5.1b of stablecoin deposits are now subject to withdrawal constraints”, Galaxy wrote³⁵. A company that connects to Aave reported on 4 May 2026: “Stablecoin liquidity is restored.”³⁶ On the sources we read, stablecoin lenders faced a wait of about two weeks and no loss of their deposit³⁵³⁶. Aave’s community proposed covering the shortfall from partners and its treasury, with the note that the figures “are not guaranteed”³⁷. We did not confirm that the vote passed.

USDC itself has slipped below $1 once. In March 2023 Circle could not withdraw $3.3 billion of USDC’s reserves, about 8%, from a failed US bank. A Federal Reserve note records: “At its trough, USDC traded at 86 cents to the dollar.”²⁴ The price recovered within days after the US government protected that bank’s depositors²⁴. A holder who waited lost nothing. One who had to sell that weekend did.

Thefts in general. DefiLlama’s list of hacks holds 1,299 incidents totalling about $21.1 billion¹². Our sums from it: 148 incidents and about $2.7 billion in 2025, and 292 incidents and about $2.2 billion in 2026 up to 7 October¹². The list covers all of crypto, including exchanges, not stablecoin pools alone. It records about $1.45 billion of the 2025 total as later returned¹².

And the plain version has held up. The sources we read report no loss of deposits for stablecoin lenders on Aave’s main market in the April 2026 squeeze³⁵³⁶. We did not check Aave’s record back to its launch. Aave’s USDC market has paid a positive single-digit average rate in every year since 2023⁵. One large vault manager, Steakhouse Financial, says it “finished the half year with zero bad debt” on its Morpho deposits in the first half of 2026³⁸. That is the company’s own statement, and it earns fees on deposits³⁸. Its larger USDC vaults on Morpho showed about 3.5% to 4.6% today⁴.

What the rules allow now

Depositing is allowed. We found no rule that bars an ordinary person from holding a regulated stablecoin and lending it through a protocol. The limits fall on others, and they shape what is on offer.

• United States. A 2025 law, the GENIUS Act, says no stablecoin issuer “shall pay the holder of any payment stablecoin any form of interest or yield”¹⁷. The same law says these coins are not backed by the US government and are not covered by deposit insurance¹⁷. The ban is on issuers, not on a person lending coins. Whether exchanges may keep paying rewards on balances is unsettled. A wider crypto bill (H.R. 3633, the Clarity Act) stalled in the Senate on 15 September 2026 when a vote to end debate failed; the tracker notes “The Senate may try again”¹⁸. A consumer finance site wrote in July 2026 that the latest draft “prohibits platforms from paying out passive yields just for holding stablecoins” and that most of the exchange programmes in Step 6 “may have to change or shut down” if it became law in that form⁹. We did not read the bill text or check whether the GENIUS Act’s detailed rules are final.
• European Union. The crypto law known as MiCA says: “Crypto-asset service providers shall not grant interest when providing crypto-asset services related to e-money tokens.”¹⁹ Dollar and euro stablecoins fall in that category. Lending through a decentralised protocol is outside that article; we did not check how the EU treats such protocols.
• Who may use the apps. Aave’s terms bar users in sanctioned countries and bar using a VPN “for the purpose of circumventing geographic blocking”¹⁵. Whether use is legal where you live is left to you¹⁵.

The depositor carries the risk.

• Aave’s terms: “No particular value, liquidity, yield or return is guaranteed.” And: “Any of these can result in total loss of assets.”¹⁵ The operator’s liability is capped at “THE GREATER OF: (A) USD $1,000” or the fees paid in the past year¹⁵.
• Aave’s risk page: “Smart contracts can contain software bugs or other vulnerabilities within the protocol code and the underlying reserve tokens.”¹⁴ The same page lists audits and a bug bounty (a reward for reporting flaws) as protections¹⁴.
• Robinhood’s help page on its lending product: “No deposit insurance. Your assets in a decentralized lending protocol are not covered by FDIC insurance, SIPC protection, or any government-backed guarantee.”⁵⁹
• Circle’s terms: “Only Users Type A can redeem USDC directly with Circle.” Those are institutional accounts. Everyone else sells on the market, and “Circle does not guarantee that the value of one (1) USDC will always equal 1 USD ($1) on other platforms.”¹⁶
• The UK regulator’s consumer page: “you should be prepared to lose all your money”, and crypto “still remains high risk with no protections if something goes wrong”²¹.

Yield products have a different legal status from the coin. US securities staff said in April 2025 that fully backed dollar stablecoins are not securities, partly because they do “not pay or guarantee to pay interest”²⁰. Tokens and products that do pay a yield are outside that statement.

Tax. The US tax agency says: “If you have digital asset transactions, you must report them whether or not they result in a taxable gain or loss.”²² Swapping one token for another counts as a sale²². The UK tax agency says it “does not consider the return received by the lender to be interest for tax purposes”, and that lending tokens can itself count as a sale²³. The US pages we read do not address lending returns specifically. Every deposit, swap, reward and withdrawal may need recording. Other countries were not checked.

Look-alike sites. The FBI recorded $7.2 billion lost to cryptocurrency investment fraud in 2025, its largest loss category⁵¹. In 2022 it warned of what it called a “scam” using apps posing as “liquidity mining” services that promise “one to three percent daily”: “victims unknowingly authorize scammers to pull an unlimited amount of funds out of their cryptocurrency wallets”⁵¹. That warning is about apps reached through unsolicited messages. It is not about any tweet or account on this page. Its use here is as a general check: an income of 20% to 40% a month, which the claim on this page needs (Step 1), works out at about 0.7% to 1.3% a day, and offers of a fixed daily percentage are the mark the FBI describes.

Cases about “deposit and earn” offers. These cases concern only the companies and people named in them. None involves any account quoted on this page, and we suggest no connection. None of them is plain lending on a large protocol. They show what has happened to offers promising far more than the base rate, and to companies that held customers’ deposits themselves.

• Terraform Labs. A jury found the company and its founder Do Kwon liable for fraud, and they agreed to pay more than $4.5 billion. The regulator’s words: “Terraform and Do Kwon’s fraudulent activities caused devastating losses for investors”²⁵.
• Celsius. It advertised rewards “as high as 18 percent annual percentage yield”. The Federal Trade Commission alleged in its complaint that Celsius “gave most participants far less than promised” and took more than $4 billion of customer deposits for its own use. Celsius, then bankrupt, settled; the settlement carries a $4.7 billion judgment, suspended so that remaining assets could go back to customers. Three former executives had not settled when the release was published, and the claims against them were allegations⁴⁹.
• BlockFi. It agreed in 2022 to pay $100 million, $50 million to the securities regulator and $50 million to 32 US states, after the regulator found its interest accounts were unregistered securities. It settled without admitting or denying the findings⁴⁸.
• DeFi Money Market. A 2021 US securities case about a DeFi lender concerned tokens paying 6.25%. The regulator’s order found unregistered sales and misleading statements to investors. It was settled without admitting or denying the findings, and holders could redeem their principal and interest⁴⁷.
• Goliath Ventures, 2026. The firm promised returns from crypto “liquidity pools” (pots of tokens that traders swap against). US tax investigators say “the funds were primarily used to pay purported returns to earlier investors”, and its chief executive pleaded guilty⁵⁰. A second regulator says about 1,600 customers put in at least $397 million⁵⁰. The civil cases are pending and their claims are alleged, not decided.

Who makes money from it

In our collection. The one tweet shows nothing for sale¹. The account’s profile reads “Independent Bitcoin research and market intelligence platform delivering timely insights. DM for business” and links to a Telegram channel². The channel showed 6.69K subscribers³. We saw no yield product and no price list there³. Its pinned post is a risk notice saying the content “does not constitute investment advice” and that “there is a possibility of losing your entire investment”³.

Beyond the tweet, these are the parties with documented income from stablecoin deposits.

The issuers. They keep the interest on the dollars behind the coins. DefiLlama estimates Tether’s income at about $5.94 billion over the past year and Circle’s at about $2.38 billion, the two highest of anything it tracks⁴⁰. Converting savings to USDC, the tweet’s first step, hands that interest to the issuer.

The exchanges. An article citing Circle’s annual filing says Circle’s distribution costs connected to Coinbase were $1.4 billion in 2025⁴¹. We did not open the filing. Coinbase pays its users 3.75% and charges for the membership that unlocks it⁸.

Protocols and vault managers. They take a cut before the depositor is paid. Aave generated about $710 million in fees over the past year and kept about $90 million; Sky generated about $391 million and kept about $202 million⁴⁰. Yearn takes a performance fee from the yield. Its documents give 10% for one class of vault and say older vaults charged 20%, with fees now set vault by vault⁴². On Morpho, “Curators set performance and management fees”⁴². These cuts are small next to the gap between 4% and 240%.

Referrers. Platforms pay people who bring in depositors. Nexo’s page says: “You can earn up to $2,500 in NEXO per referral, based on your friend’s average portfolio balance over 30 days.”⁴³ The reward is 0.5% of the newcomer’s average balance, shared between the two, so $2,500 each needs a $1 million balance, and the newcomer must add $5,000 or more⁴³. OKX offers affiliates “up to 50% commission” on the trading fees of people who sign up through their link⁴³. The referrer is paid on the newcomer’s deposit or trading, not on the newcomer’s result.

Paying for referrals is ordinary and lawful. The following case does not involve the companies above. It shows how commission-paid promotion worked in one programme the regulator acted against. In a 2021 case about a crypto “lending program”, the securities regulator alleged in its complaint that the promoters “received commissions based on their success in soliciting investor funds”⁵². We did not check how the case ended.

Courses and communities.

• Yield Farming Coop sells a “49 Weeks to Freedom” challenge at $152 a month for 12 months, which is $1,824 (our sum), with the first week free, and a basic tier at $30 a month⁴⁴. Its pages say liquidity pools offer “returns ranging from 1% to over 30% per month”, that “Users starting with $250 are projected to earn over $120 weekly”, and that a token given with the basic membership carries a “Guaranteed 7.5% return each month”⁴⁴. These statements are the seller’s own and are not checked. The pages carry a 2024 footer, so prices may have changed. We found no refund policy on the pages we read. Its listing on the community platform Skool showed $360 a year and 133 members⁴⁴. A partner group’s event page offers “$10 off” through “our affiliate link”⁴⁵. It teaches liquidity pools in general, not only stablecoins.
• A search for “defi” on Skool showed paid communities from $7 to $145 a month. One, at $47 a month, showed 1.5k members and the line “Build your first income streams in 7 days!”⁴⁶ We did not open their sales pages or refund terms.

We found no published results from any of these sellers showing what share of members reach an income.

The upside

Stablecoin lending pays. These are the best documented results. None reaches the tweet’s rate.

Depositors as a group are paid hundreds of millions of dollars a year. The four protocols named under “What the scheme is” paid depositors about $1.26 billion between them over the past year (our sum from⁴⁰). A news article says Ethena has paid out over $751 million since launch, at an average of 10.9% since January 2024⁵⁴. DefiLlama’s figure of $728 million is close to that⁴⁰. These are group totals from public data and the companies’ own figures, not individual results.

The best documented year was 2024. Someone lending USDC on Aave for that year averaged about 7.2%⁵. A holder of sUSDe averaged about 17.5% a year over the ten and a half months from its February launch, with March the best month at about 2.7%⁵. On $2,500 that is about $180 for the year on Aave and about $380 over the ten and a half months on sUSDe (our estimate). Both rates have since fallen to between 3% and 5%⁵.

Large listed companies sell the plain version. In September 2025 Coinbase began placing customers’ USDC into Morpho lending vaults; a news site reported yields “currently as high as 10.8%” at launch⁵⁵. In July 2026 Coinbase advertised about 7% to paying members, and the vault manager said about $368 million of its deposits came from the Coinbase app³⁸. DefiLlama today shows that manager’s two large USDC vaults on the same network at about 3.5% and 4.6%⁴. We could not confirm which vaults hold Coinbase customers’ money or what Coinbase advertises today. Robinhood’s version is in Step 6.

Named people with checked results: none found. One search for a profiled stablecoin lender with a public wallet and stated capital returned only strategy guides. The evidence that it works is at the level of whole protocols.

The top tenth and top hundredth of pools. There is no data on what the best depositors earn, so we use pools. Among pools of $1 million or more, the top tenth pay about 10.2% or more and the top hundredth about 30% or more⁴. Pools paying 20% or more hold 0.16% of the money in that group⁴. In practice the money in the highest-paying pools is small, and the readings are unstable.

The skilled versions.

• Fixed-rate tokens on newer dollar products. On the protocol Pendle, tokens that lock in a rate until a set date showed about 10% to 16% today, in pools of $12 million to $50 million⁴. Someone able to judge what stands behind a new dollar token can roughly triple the base rate. On $2,500 that is about $30 a month (our estimate).
• Looping. This means borrowing against a deposit to deposit again, so the gap between two rates is earned several times over. A Bank of Canada staff paper finds that “many users engage in recursive leverage” on Aave and that forced sales “occur in concentrated waves”³⁹. Deposits of one such token on Aave and Morpho “peaked at $2.81B on Sep 19, 2025” and had fallen to $403 million by April 2026 as the gap closed³⁴. We found no published results for people who did this.
• Tokens handed to early depositors. Some new products have also paid early users in the project’s own token, called an airdrop. In April 2024 Ethena gave 750 million of its ENA tokens, 5% of the supply, to early users after a six-week points campaign⁵⁸. We have no figure for what a small depositor received. This is a one-off that cannot be planned as monthly income, and most products never do it. It is covered on our airdrop farming page.

A realistic good result (our estimate). In the first year, about 3.5% to 4.5% on the largest lenders with no skill beyond handling a wallet safely, going by 2026 so far⁴⁵. Perhaps 6% to 10% for someone who picks among newer products and accepts a real chance of losing part of the deposit⁴. On $2,500 that is about $90 to $250 in the year. Time to the first interest is days. Time to $500 a month is set by the size of the deposit, about $60,000 to $150,000 at 10% down to 4%, and not by experience.

What it takes to compete

Nobody needs to beat anyone to earn 4%. What separates the people who draw a monthly income from those who earn a few dollars is mostly the first item.

1. Capital. Income is the deposit times the rate. At today’s average rate, $1,000 a month takes about $300,000 and $10,000 a month about $3 million (Step 4). The scheme pays people who already hold six or seven figures in dollars.
2. Knowing where each yield comes from. Rates above the base come from somewhere: reward tokens, futures funding that can turn negative⁵³, borrowing to deposit again³⁹, or an outside manager trading the money²⁹. The products that failed in 2025 and 2026 were in this group³¹³². The people who do well with them can read what backs a token and who holds the keys. Eighteen audits did not prevent one of those failures³³.
3. Wallet safety. There is no insurance and no one to claim from¹⁵¹⁷²¹. Approving the wrong site can empty a wallet⁵¹.
4. Attention. The plain version needs little. The skilled versions need daily watching of rates and of the point at which a loan is forcibly closed³⁴³⁹. A deposit may also be stuck for days in a squeeze³⁵.
5. A reason to be here and not in a Treasury bill or an online savings account. Right now there is little or no extra pay over either⁴⁶⁶⁰. The case is stronger for someone who cannot easily hold dollars or buy that debt where they live. That is our judgement; we found no data on who depositors are.

What it costs. Network fees are cents on the cheapest routes today¹⁰¹¹. A course is optional and runs from $7 a month to $1,824 a year on the pages we read⁴⁴⁴⁶. At 4.0%, $2,500 earns about $100 a year (our estimate). A $30 monthly membership costs more than three times that.

How to tell early which side you are on. These checks are our judgement, built from the figures above.

• Do the sum before anything else. Take the monthly income you want, multiply by 12 and divide by your deposit. If the answer is above about 14% a year, no pool holding $100 million or more pays it today⁴. If it is above 20%, you are looking at pools that hold 0.16% to 0.19% of the money⁴.
• Ask what pays the rate. If a page promises a monthly or daily percentage and does not say who pays it, compare it with the cases under “What the rules allow now”. If part of the rate is reward tokens, ask what the rate is without them.
• Check the rate’s history, not today’s number. The one large pool showing 898% today showed 0% two days earlier⁴.
• Count what you keep. Subtract membership fees, network fees and tax. On small sums a $4.99 monthly fee takes most of the yield (Step 6).
• Ask whether you could get out. Stablecoin withdrawals on the largest lender were limited for about two weeks in April 2026³⁵³⁶.
• Compare with the dull alternative. If a Treasury bill or an insured online savings account open to you pays about the same as the pool, the extra risk is buying nothing⁶⁶⁰.

What we did not verify

• The tweet’s video. We did not view the attached media, so the five protocols, any rates and any proof shown in it are unknown.
• Only one tweet. Nothing here describes how often or by whom this pitch is posted.
• What individual depositors earn. We found no data on earnings per wallet, the share of depositors who come out ahead, or how many hold enough to earn $1,000 a month. A study of depositor sizes on Aave redirected to a login page and returned nothing through a reader service; its figure is not used.
• Personal reports. Reddit refused our tools by both routes. No forum accounts were read. One first-person write-up of a one-week test was read and is not used: it was about a different crypto asset and its figures could not be checked⁵⁶.
• DefiLlama figures. All rates are one snapshot on 10 October 2026, computed by us from the raw data, and they move within hours. The stablecoin flag includes tokenised Treasury funds and some paired pools. Yearly averages are simple means of daily readings. Rates were not cross-checked on the protocols’ own apps. Payout and revenue figures are DefiLlama’s estimates, and for Aave and Morpho cover all assets.
• The pools above 240%. We did not open any of those protocols to find out why they show those rates or whether anyone could have earned them.
• Exchange routes. Robinhood’s 7% is from a consumer site’s July 2026 list; Robinhood’s marketing page was not read, and the site’s writers said they could not sign up themselves⁹. The Kraken rate and Coinbase’s yearly price are from the same list and were not checked on the companies’ pages. Online savings rates are from one list of accounts chosen by a site that earns referral fees from banks⁶⁰.
• Stream, Resolv and the Aave incident. What Stream’s depositors finally recover is unknown; nothing after 29 June 2026 was found. The 18% and $382 million figures are from a commentary, not from Stream. What depositors in other managers’ vaults lost through Stream is not in the sources we read. Resolv’s own account of the incident was not found; one source is an insurer that sells cover. We did not confirm that Aave’s funding proposal passed. An article we read mentions a legal complaint about Stream; we did not open a court document and leave it out.
• Hack totals. They cover all of crypto. We have no loss rate for stablecoin pools alone.
• Law. The Senate vote is from a bill tracker; congress.gov refused our tools by both routes. What the Clarity Act says about rewards is from a July 2026 article, not the bill text. The GENIUS Act’s start date and final rules, and the EU’s treatment of decentralised protocols, were not checked. Terms of lenders other than Aave were not read. Whether X has rules on income claims like this was not checked.
• Tax. No source read states how US tax treats lending returns from a protocol. Countries other than the US and UK were not checked.
• Fees. The two fee sources conflict for Ethereum. Etherscan’s page refused a second check, directly and through a reader service, so the $0.06 and $0.05 figures rest on one reading. The cost of buying USDC with a bank transfer or card, and of moving between networks, was not priced.
• Coinbase and Circle. Coinbase’s page refused us directly and was read through a reader service. Circle’s annual filing was not opened. The Block’s report of the 10.8% launch rate was not re-read.
• Token distributions. A report that depositors in one 2025 pre-launch vault each received tokens worth several thousand dollars could not be opened (the page returned an error), so it is left out.
• Sellers. Refund terms were not found for the course named above, and the Skool sales pages were not opened. Member counts may include people who do not pay. Listings on other course sites could not be read. The Binance and Bybit affiliate pages did not load.
• Cases. The sentencing in the Goliath Ventures case was set for 8 October 2026; we found no report of it. Final judgments in the 2021 promoters case, the outcome for the former Celsius executives, and whether BlockFi paid the full sum before its later bankruptcy were not checked. A US investor bulletin on crypto interest accounts could not be opened.
• Exact quotes. Quotes from the Coin Metrics, Cointelegraph, Kiln, Aave governance, NBER, Decrypt, The Block and Crypto Briefing pages and the US tax agency pages were read through a tool that summarises text and were not checked word for word. Quotes from Aave’s terms, Circle’s terms, the two laws, the Federal Reserve note, the UK regulator, the FBI, the US securities and trade regulators’ releases, Galaxy, the Defiant, Steakhouse, Dune, Ethena’s documents, Robinhood, NerdWallet, Nexo, the course pages and the yield-farm paper were matched in the page text.
• Whether the people named have seen this page. The people and companies named here were not contacted.

Sources

“Read at source” means we opened the page itself on 10 October 2026. “Secondary” means we read an article or summary about the source. “Estimate” means our own sum. “Through a reader service” means a service (r.jina.ai) fetched the page for us because the site refused our tools directly.

1. Does It Pay collection of 1 tweets on this scheme, collected 2026-10-06. The tweet: https://x.com/BitcoinIntelX/status/2106758434989240687 , 2026-10-04. Read from our collected copy.
2. The same tweet and the account’s profile, through a mirror. https://api.fxtwitter.com/BitcoinIntelX/status/2106758434989240687 . 2026-10-04. Read at source on 2026-10-10 (x.com itself refused our tools). The attached media was not viewed.
3. Telegram channel linked from the profile. https://t.me/s/BitcoinIntelX . Pinned post dated 2025-11-13; latest post 2026-10-08. Read at source on 2026-10-10.
4. DefiLlama, live list of yield pools. https://yields.llama.fi/pools . Downloaded 2026-10-10 at 14:34 UTC. Read at source on 2026-10-10; medians, averages, shares and counts are our own calculations (estimate) on the downloaded data. The daily readings for the one pool in Step 3 are from an earlier download the same day.
5. DefiLlama, daily rate history for USDC on Aave v3 (https://yields.llama.fi/chart/aa70268e-4b52-42bf-a116-608b370f9501), USDC on Compound v3 (https://yields.llama.fi/chart/7da72d09-56ca-4ec5-a45f-59114353e487), sUSDe (https://yields.llama.fi/chart/66985a81-9c51-46ca-9977-42b4fe7bc6df) and sUSDS (https://yields.llama.fi/chart/d8c4eff5-c8a9-46fc-a888-057c4c668e72). Data to 2026-10-10. Read at source on 2026-10-10; yearly and monthly averages are our estimate.
6. US Treasury, daily Treasury bill rates. https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_bill_rates&field_tdr_date_value_month=202610 . 2026-10-09. Read at source on 2026-10-10.
7. FDIC, national rates and rate caps. https://www.fdic.gov/national-rates-and-rate-caps . 2026-09-21. Read at source on 2026-10-10.
8. Coinbase, USDC page. https://www.coinbase.com/usdc . Undated page. Read at source on 2026-10-10 through a reader service.
9. NerdWallet, stablecoin yield programmes. https://www.nerdwallet.com/investing/learn/stablecoin-yield-programs . Updated 2026-07-29. Read at source on 2026-10-10 through a reader service; its figures about the exchanges and the bill are secondary.
10. Etherscan gas tracker. https://etherscan.io/gastracker . 2026-10-10, 14:12 UTC. Read at source on 2026-10-10.
11. L2 Fees. https://l2fees.info/ . Live page, undated. Read at source on 2026-10-10 through a reader service; the $50 total is our estimate.
12. DefiLlama, hacks database. https://api.llama.fi/hacks . Latest entry 2026-10-07. Read at source on 2026-10-10; yearly totals are our sums (estimate).
13. Coin Metrics, State of the Network, issue 379. https://coinmetrics.substack.com/p/state-of-the-network-issue-379 . 2026-09-01. Read at source on 2026-10-10, through a summarising tool.
14. Aave documentation, risks. https://aave.com/docs/concepts/risks . Undated page. Read at source on 2026-10-10.
15. Aave terms of service. https://aave.com/terms-of-service . Undated page. Read at source on 2026-10-10.
16. Circle, USDC terms. https://www.circle.com/legal/usdc-terms . Last updated 2025-12-12. Read at source on 2026-10-10.
17. GENIUS Act, Public Law 119-27. https://www.govinfo.gov/content/pkg/PLAW-119publ27/html/PLAW-119publ27.htm . 2025-07-18. Read at source on 2026-10-10.
18. GovTrack, H.R. 3633. https://www.govtrack.us/congress/bills/119/hr3633 . Vote dated 2026-09-15. Secondary (a tracker, not the Senate record), read on 2026-10-10 through a reader service.
19. Regulation (EU) 2023/1114 (MiCA), Article 50. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32023R1114 . 2023-05-31. Read at source on 2026-10-10.
20. US Securities and Exchange Commission, staff statement on stablecoins. https://www.sec.gov/newsroom/speeches-statements/statement-stablecoins-040425 . 2025-04-04. Read at source on 2026-10-10 through a reader service.
21. UK Financial Conduct Authority, crypto basics. https://www.fca.org.uk/investsmart/crypto-basics . Updated 2026-01-29. Read at source on 2026-10-10 through a reader service.
22. US Internal Revenue Service, digital assets. https://www.irs.gov/filing/digital-assets . Reviewed 2026-09-02. Also the virtual currency FAQ, https://www.irs.gov/individuals/international-taxpayers/frequently-asked-questions-on-virtual-currency-transactions , updated 2026-06-30. Read at source on 2026-10-10, through a summarising tool.
23. HM Revenue and Customs, Cryptoassets Manual, CRYPTO61110 and CRYPTO61620. https://www.gov.uk/hmrc-internal-manuals/cryptoassets-manual/crypto61110 and https://www.gov.uk/hmrc-internal-manuals/cryptoassets-manual/crypto61620 . Updated 2025-11-28. Read at source on 2026-10-10.
24. Federal Reserve, FEDS Notes on the Silicon Valley Bank failure and stablecoins. https://www.federalreserve.gov/econres/notes/feds-notes/in-the-shadow-of-bank-run-lessons-from-the-silicon-valley-bank-failure-and-its-impact-on-stablecoins-20251217.html . 2025-12-17. Read at source on 2026-10-10.
25. US Securities and Exchange Commission, Terraform Labs: complaint, https://www.sec.gov/newsroom/press-releases/2023-32 , 2023-02-16; judgment, https://www.sec.gov/newsroom/press-releases/2024-73 , 2024-06-13. Read at source on 2026-10-10 through a reader service; quotes matched in the page text.
26. Liu, Makarov and Schoar, “Anatomy of a Run: The Terra Luna Crash”, NBER Working Paper 31160. https://www.nber.org/papers/w31160 . 2023-04. Read at source on 2026-10-10 (abstract).
27. Augustin, Chen-Zhang and Shin, “Reaching for Yield in Decentralized Financial Markets”. https://www.udel.edu/content/dam/udelImages/star/fintech-innovation-hub/Reaching_for_Yield_in_Decentralized_Financial_Markets.pdf . Draft of 2024-03-21. Read at source on 2026-10-10.
28. CoinShares commentary on Stream Finance. https://coinshares.com/insights/the-node/edito-07-11-25/ . 2025-11-07. Secondary.
29. Cointelegraph, Stream Finance pauses platform. https://cointelegraph.com/news/stream-finance-pauses-platform-finds-93m-loss . 2025-11-04. Read at source on 2026-10-10, through a summarising tool.
30. Decrypt, Stream Finance. https://decrypt.co/347285/stream-finance-stablecoin-plunges-77-protocol-fund-manager-loses-93-million . 2025-11-04. Secondary (a news article quoting the company’s statement), through a reader service.
31. The Defiant, Stream Finance creditor claims. https://thedefiant.io/news/defi/stream-finance-starts-collecting-creditor-claims-in-step-toward-global-resolution . 2026-06-29. Read at source on 2026-10-10 through a reader service; the exposure figures in it are secondary.
32. The Defiant, Resolv USR exploit. https://thedefiant.io/news/hacks/defi-has-seen-resolv-s-usd25m-usr-exploit-many-times-before . 2026-03-23. Read at source on 2026-10-10 through a reader service.
33. Nexus Mutual, Resolv incident report. https://nexusmutual.io/blog/resolv-protocol-incident-report-by-nexus-mutual . 2026-03-23. Secondary (an insurer’s report), through a reader service.
34. Dune, “The Pendle Effect”. https://dune.com/blog/the-pendle-effect . Undated; data to 2026-04-23. Read at source on 2026-10-10 through a reader service.
35. Galaxy Research, the Kelp exploit and DeFi. https://www.galaxy.com/insights/research/kelpdao-layerzero-exploit-defi . 2026-04-22. Read at source on 2026-10-10 through a reader service; quotes matched in the page text.
36. Kiln status page. https://status.kiln.fi/incidents/01KPN7ACX5K88ETEJGYYF51SKB . Updates from 2026-04-20 to 2026-05-12. Read at source on 2026-10-10, through a summarising tool.
37. Aave governance forum, rsETH incident funding update. https://governance.aave.com/t/arfc-rseth-incident-funding-update/24740 . 2026-04-24. Read at source on 2026-10-10, through a summarising tool.
38. Steakhouse Financial, DeFi markets update. https://kitchen.steakhouse.financial/p/defi-markets-update-2026-07-14 . 2026-07-14. Read at source on 2026-10-10 through a reader service. The company’s own statement.
39. Bank of Canada, Staff Analytical Paper 2026-13. https://www.bankofcanada.ca/2026/04/staff-analytical-paper-2026-13/ . 2026-04. Read at source on 2026-10-10 (abstract only).
40. DefiLlama, fees and revenue data. https://api.llama.fi/overview/fees?excludeTotalDataChart=true&excludeTotalDataChartBreakdown=true&dataType=dailyFees and https://api.llama.fi/summary/fees/aave-v3?dataType=dailySupplySideRevenue (and the same for morpho-blue, ethena-usde, sky-lending). 2026-10-10. Read at source on 2026-10-10. DefiLlama’s estimates.
41. CryptoSlate, Circle and its distributors. https://cryptoslate.com/usdc-keeps-winning-but-circles-distributors-are-taking-more-of-the-prize/ . 2026-07-15. Secondary, through a reader service.
42. Yearn documentation, https://docs.yearn.fi/getting-started/products/yvaults/overview , and Morpho documentation, https://docs.morpho.org/learn/concepts/curator/ . Undated pages. Read at source on 2026-10-10.
43. Nexo referral page, https://nexo.com/referral , and OKX affiliates page, https://www.okx.com/affiliates . Undated pages. Read at source on 2026-10-10.
44. Yield Farming Coop: https://yieldfarmingcoop.com/freedom , https://yieldfarmingcoop.com/basic and https://www.skool.com/yield-farming-coop/about . Footer dated 2024. Read at source on 2026-10-10.
45. Event page with an affiliate link. https://luma.com/rxtzxsj6 . Event dated 2025-02-07. Read at source on 2026-10-10.
46. Skool, search for “defi”. https://www.skool.com/discovery?q=defi . 2026-10-10. Read at source on 2026-10-10.
47. US Securities and Exchange Commission, DeFi Money Market. https://www.sec.gov/newsroom/press-releases/2021-145 . 2021-08-06. Read at source on 2026-10-10 through a reader service; quotes matched in the page text.
48. US Securities and Exchange Commission, BlockFi. https://www.sec.gov/newsroom/press-releases/2022-26 . 2022-02-14. Read at source on 2026-10-10 through a reader service; quotes matched in the page text.
49. US Federal Trade Commission, Celsius. https://www.ftc.gov/news-events/news/press-releases/2023/07/ftc-reaches-settlement-crypto-platform-celsius-network-charges-former-executives-duping-consumers . 2023-07-13. Read at source on 2026-10-10 through a reader service; quotes matched in the page text.
50. IRS Criminal Investigation, Goliath Ventures plea, https://www.irs.gov/compliance/criminal-investigation/goliath-ventures-ceo-pleads-guilty-to-cryptocurrency-fraud-scheme-conspiracy , 2026-06-30; Commodity Futures Trading Commission release 9280-26, https://www.cftc.gov/PressRoom/PressReleases/9280-26 , 2026-08-11. Read at source on 2026-10-10.
51. FBI Internet Crime Complaint Center: 2025 annual report, https://www.ic3.gov/AnnualReport/Reports/2025_IC3Report.pdf , published 2026-04 (through a reader service); public service announcement on liquidity mining, https://www.ic3.gov/PSA/2022/PSA220721 , 2022-07-21. Read at source on 2026-10-10.
52. US Securities and Exchange Commission, promoters of a crypto lending programme. https://www.sec.gov/newsroom/press-releases/2021-90 . 2021-05-28. Read at source on 2026-10-10 through a reader service; quote matched in the page text.
53. Ethena documentation, funding risk. https://docs.ethena.fi/solution-overview/risks/funding-risk . Undated page. Read at source on 2026-10-10.
54. Crypto Briefing, Ethena rewards. https://cryptobriefing.com/ethena-750m-rewards-since-launch/ . 2026-07-28. Secondary.
55. The Block, Coinbase USDC lending. https://www.theblock.co/post/371281/coinbase-usdc-onchain-lending . 2025-09-18. Secondary.
56. “I tried yield farming for a week”, The Daily Dollar. https://thedailydollar27.substack.com/p/i-tried-yield-farming-for-a-week . 2025-02-24. Read on 2026-10-10 through a summarising tool; not used for any figure on this page.
57. Coin360, Resolv USR. https://coin360.com/news/resolv-usr-stablecoin-exploit-depeg-80m-mint . 2026-03-23. Secondary, read on 2026-10-10 through a reader service.
58. The Defiant, Ethena’s ENA token distribution. https://thedefiant.io/news/defi/ethena-labs-ena-launches-at-usd1-billion-post-airdrop-as-sats-campaign-kicks-off . 2024-04-02. Secondary (a news article), read on 2026-10-10 through a reader service.
59. Robinhood support, Robinhood Earn. https://robinhood.com/us/en/support/articles/robinhood-earn/ . Undated page. Read at source on 2026-10-10.
60. The Motley Fool, savings account rates. https://www.fool.com/money/banks/articles/top-savings-account-rates-today-oct-5-2026/ . 2026-10-05, rates as of 2026-10-04. Secondary (a list of banks’ rates), read on 2026-10-10 through a reader service.

Corrections

If you are quoted or named on this page and think something is wrong, want your reply shown beside it, or want to be removed, write to [email protected]. We aim to reply within 14 days, and always within one month. Factual errors are corrected with a dated note here. A sentence that is seriously disputed comes down while we check it.

Change log: (empty)