Lloyds Settled $750,000 in USDC. Visa's Chain Has No Name.
Lloyds Banking Group and Visa finished a seven-day live pilot and said so on Sept. 30. The number they are willing to print is $750,000. The asset is USDC. The clock they want you to remember is under an hour, including a weekend. The chain Visa used on its side of the transfer does not have a name in any of the writeups I have.
I have not seen a ledger export. What follows is the Visa UK press release, PYMNTS the same day, Electronic Payments International on Oct. 1, and The Block the same morning. The Block points at Lloyds’ own release. I am treating the four secondary accounts as the record, because they agree on the operational facts and disagree mainly by omission.
This is not the card story. We already wrote SoFi settling its Mastercard volume in SoFiUSD. Merchants in that setup never hold the coin. This is also not Visa’s pitch to underwrite stablecoin cards from settlement data, and it is not U.S. Bank moving money on Stellar under the name USBDC. Lloyds is moving its own obligation to Visa. The customer payment already happened somewhere else, or it is not in these pages at all.
What the USDC settlement actually moved
The Visa release dates the announcement to Sept. 30, 2026, in London. PYMNTS uses the same Wednesday. EPI’s byline is Oct. 1, Aninda Chakraborty. The Block’s timestamp is Oct. 1, 7:13 a.m. EDT, edited by Brian Danga. Four clocks, one event. I am not going to treat the Oct. 1 pieces as a second pilot.
The volume is a series of U.S. dollar settlement obligations totalling $750,000. Visa, EPI, PYMNTS, and The Block all print that total. None of them prints the count of payments inside it. A single $750,000 obligation and seventy-five smaller ones would look identical in these paragraphs. I will not pick a count.
Lloyds bought the USDC through Archax. Visa calls Archax a UK-regulated digital asset exchange. EPI and The Block use the same description. I do not have Archax’s order ticket, the price Lloyds paid above par, or the hour the coins were purchased. “Purchased through” is the verb they share. It is not a disclosure of the spread.
The booking path is the part I would put on a diagram before the slogan. The settlement volume was booked through Lloyds’ Corporate Markets branch in Jersey, then transferred to Visa in the United States. All four accounts say Jersey, and all four say the receiving side is Visa in the U.S. That is a branch and a counterparty, not a corridor map. Jersey to the United States is the only geography these pages actually commit to.
Peter Left, head of digital assets at Lloyds Banking Group, is the quote that travels. The Visa release has the long version. Stablecoins could be particularly valuable for cross-border payments, where moving money between markets, currencies, and infrastructures adds time and complexity. Settling $750,000 of live payment obligations with Visa let them move beyond theory. Greater visibility and certainty over the movement of funds can transform liquidity management. Interoperability between blockchain networks helps unlock later applications of digital money at scale. EPI reprints that sequence almost intact. The Block keeps the “beyond theory” sentence and drops the rest. PYMNTS keeps the liquidity sentence and attributes the shorter version to the release. I am using the Visa text as the full quote and the others as confirmation that the sentence was in the release, not invented by a blog.
Under an hour is the claim. A day is the baseline they chose.
The speed sentence is identical in spirit across the four pieces, and close enough in wording that I am not going to hunt a synonym. Funds reached Visa in under an hour, including over the weekend. Visa puts “under an hour” in the body and in the summary bullets. EPI puts the phrase in quotation marks and attributes it to the statement. The Block says the bank said it Wednesday. PYMNTS says “per the release.”
The baseline they compare it with is also shared. Under traditional cross-border processes, settlement can take a day or more if it is initiated outside banking hours. That is a conditional, not a median. It does not say the average Lloyds-to-Visa settlement last quarter took 26 hours. It says the failure mode they are selling against is the one that starts after the window closes and waits for the next banking day. A pilot that includes a weekend is aimed at that failure mode. It does not, by itself, tell you what happens to a Tuesday morning payment that was already going to clear the same day.
EPI adds a sentence the others leave implicit. The pilot evaluated how round-the-clock settlement could affect treasury and liquidity management. The companies tested whether stablecoin settlement could provide clearer information on the status of funds and more predictable arrival times when transactions occur outside standard banking windows. “Clearer” and “more predictable” are the test questions, not the measured results. I do not have a before-and-after timestamp table. I have a claim that arrival was under an hour, and a description of what the pilot was trying to learn.
PYMNTS also quotes a Wednesday post on X from Canton Network. The post, as PYMNTS renders it, repeats the day-or-more baseline, the seven-day pilot, the $750,000, Lloyds’ own node on Canton, and funds reaching Visa in under an hour including the weekend. I have not opened that post. I am not going to treat a quote of a quote as a fifth primary source. It matters only as evidence that Canton, the network, was willing to attach its name to the same numbers the bank printed.
Canton has a name. Visa’s public chain does not.
This is the sentence I do not want smoothed over. Lloyds used its own node on Canton and used the network’s configurable privacy capabilities. Visa supported settlement on a separate public blockchain. Visa’s release says “a separate public blockchain” and stops. EPI says the same, and adds that Lloyds said the approach demonstrated interoperability across both networks. PYMNTS says “a separate public blockchain, according to the release.” The Block is the one that adds the adjective the press release avoided: unnamed.
I am not going to guess. Visa’s stablecoin work has touched more than one public chain, and The Block’s Oct. 1 piece says that in April Visa added Canton and four other blockchains to its settlement pilot, taking the total to nine. That April figure is The Block citing its own earlier story, not a line in the Sept. 30 release. Even if you accept it, nine candidates is not an identification. A pilot that puts Lloyds on Canton and Visa on “a public chain” can be interoperable without telling the reader which public chain received the USDC. Interoperability, in these pages, means the two sides did not have to be on the same network. It does not mean the second network was disclosed.
Canton is the named half, and the privacy clause is doing work. “Configurable privacy” is Visa’s phrase, repeated by EPI. It is not a description of which fields were hidden, from whom, or for how long. A node Lloyds runs itself is a different operational fact from a privacy setting. They ran a node. They also say the network can be configured for privacy. I will not collapse those into “the transfer was private.” Private from the public mempool, private from Visa, private from a regulator: those are three claims, and none of them is printed.
The Block adds a second Lloyds fact that is easy to mix into this one. Lloyds was among six major UK banks that began a live pilot of tokenized sterling deposits last year. That is a 2025 deposit pilot, in sterling, with other banks. This week’s pilot is dollar USDC, one bank, one card network, Jersey to the United States. We wrote the Canadian banks’ tokenized-deposit plan as a deposit story. Same family of idea, different instrument. A tokenized deposit and a purchased USDC are not the same liability. The Block does not say the sterling pilot and the USDC pilot share a platform. I will not say it either.
The $20 billion figure is a different denominator
The Block’s Oct. 1 piece says Visa’s stablecoin settlement volume surpassed a $20 billion annualized run rate last month, and it points at a Sept. 8 story for that number. I am not relitigating September. I am using the Oct. 1 article as the place a reader of this week’s news will meet the larger number, and I am refusing the slide from one to the other.
$750,000 over seven days is the live obligation this pilot settled. A $20 billion annualized run rate is a pace applied to a much wider book, if the earlier story holds. You cannot divide one into the other and get a market share. You also cannot treat a successful weekend test of three-quarters of a million dollars as evidence that the run rate is real. The run rate was not measured in Jersey this week. It was cited.
Rob Cameron’s job title shows up in PYMNTS and not in the chunks of the Visa release I am working from. PYMNTS calls him group country manager, UK and Ireland, at Visa, and says the line is from the release. His quoted point is that the pilot shows stablecoins can work alongside existing banking infrastructure, so institutions have more choice over how and when they settle. Visa’s role, in his sentence, is to make these forms of money practical, interoperable, and trusted at scale. PYMNTS also has him calling the pilot a potential solution for businesses that operate across borders and around the clock.
“Alongside” is the word I would keep. This is not a replacement announcement. Nothing in the four pieces says correspondent banking is being switched off, or that a Lloyds corporate client can now ask for USDC settlement as a product. The choice Cameron describes is a choice for financial institutions about settlement, which matches the scope sentence The Block is careful to include: the pilot dealt with settlement between the institutions rather than the payments made by customers.
That distinction is the one a product manager will try to erase in the next slide. A cardholder did not receive USDC. A merchant did not get paid in USDC, at least not in any sentence these articles print. Visa and Lloyds settled an obligation between themselves. If you already think of Visa as the network that moves the consumer payment, this is the layer underneath that, the layer where the institutions square up. Faster squaring-up is a treasury fact. It is not a new checkout button.
What these pages do not contain
I keep a list of the blanks, because the blanks are where a rewrite usually cheats.
No fee. No gas figure. No FX spread on the Archax purchase. No count of transactions inside the $750,000. No name for Visa’s public chain. No production date. No customer cohort. No statement that the under-an-hour result was the worst case, the median, or the best case. No independent observer named. Canton Network’s X post, as quoted, does not add a number the release did not already have.
Lloyds’ “first” is narrow, and it should stay narrow. The Visa release and The Block both carry the description: the first stablecoin settlement trial between Visa and a major UK banking group. That is not “the first stablecoin settlement,” and it is not “the first Visa stablecoin pilot.” The Block’s own April reference, nine blockchains in the settlement pilot, is the reason the wider “first” would be false even if you wanted it. First with this counterparty class is a claim about the UK bank relationship. I can repeat it. I cannot promote it.
EPI’s headline frame, faster cross-border flows, is the companies’ frame. The project examined whether stablecoins could support faster, more transparent, and more flexible cross-border transactions, “the British financial institution said in a statement.” Examined whether is the right tense. Completed a pilot is the status. Scaled the corridor is not a status anyone has earned this week.
If you only need one operational picture, it is this. A Jersey booking branch, USDC bought via a UK-regulated exchange, a Canton node on the Lloyds side, an unnamed public chain on the Visa side, $750,000 of institutional settlement, arrival in under an hour on a clock that included a weekend, and a comparison case of a day or more when the old process starts after hours. Everything else in the releases is a sentence about what that picture might mean later. Later is not a date. I am not filling one in.
