The Cash Leg Crosses Two More Borders

Citi and DBS moved dollars across a weekend for the first time on Swift’s ledger, BlackRock brought its tokenised cash franchise to Hong Kong through Standard Chartered’s infrastructure, and a Bitcoin sidechain lost 95% of its reserve wallet to an exploit its operator refuses to call a bug bounty. So in short this week, institutional settlement kept advancing while a decade-old sidechain relearned an old lesson about trust.

Executive Brief

Two weeks ago the ECB put a date on settling tokenised assets in central-bank money and this week the commercial banks supplied their own version of the same argument; it arrived in a form ordinary depositors will recognise - a payment that used to wait for Monday.

On 5 September, DBS in Singapore and Citi’s New York office settled a US dollar payment over a weekend using tokenised deposits on Swift’s Digital Ledger, in minutes rather than the up to two business days a conventional correspondent-banking route would need. No stablecoin was involved - each bank issued the deposit claim on its own ledger, and Swift’s platform - a permissioned layer built on Hyperledger Besu - coordinated between them before final settlement ran through existing systems. It was only the ledger’s second live transaction. The relevant fact is not the amount (undisclosed) but that a weekend, the oldest constraint in correspondent banking, has stopped being one.

BlackRock supplied the same argument from the fund-management side. On 10 September it received Hong Kong regulatory authorisation for the BlackRock HKD Digital Liquidity Fund, its first tokenised money-market fund in Asia-Pacific, joining existing tokenised cash products in the US and Europe. Standard Chartered sits underneath it as trustee, custodian and fund administrator, and investors will subscribe and redeem using HKDAP - a Hong Kong dollar stablecoin issued by Standard Chartered-led Anchorpoint Financial - marking the bank’s first commercial role as a stablecoin distributor. One institution now supplies the custody, the administration and the settlement currency around the same product.

Set against both is a reminder of what the older architecture costs when it fails. Liquid Network, the Bitcoin sidechain Blockstream built in 2018, lost 95% of its federation reserve, roughly 4,000 of 4,200 BTC, to an exploit of the software governing its own peg mechanism. Most of it came back within days. Blockstream is refusing to call the remainder a bug bounty. The three developments together describe a market moving in one direction at two different speeds: the settlement layer advancing through banks and asset managers with methodical, almost administrative regularity, and the older infrastructure beneath a chunk of the crypto-native economy still capable of losing nearly all of its reserve to a signature it should never have accepted.

Key Developments

A. Tokenised Markets

1. Citi and DBS settle a weekend dollar payment on Swift’s ledger

What happened: DBS, from Singapore, and Citi’s New York office completed a cross-border US dollar payment on Saturday 5 September using tokenised deposits on Swift’s Digital Ledger, settling in minutes against an industry norm of up to two business days for a weekend transfer. Each bank issued its own tokenised deposit claim; Swift’s platform, a permissioned orchestration layer built on Hyperledger Besu, matched and coordinated the obligation between the two ledgers before final settlement occurred through existing banking infrastructure. No stablecoin changed hands - tokenised deposits remain claims on conventional bank deposits, governed by the same deposit, KYC and capital rules as the underlying account. Citi’s Mridula Iyer called it a demonstration that “always-on cross-border payments are already a reality.” It was the second live transaction on Swift’s ledger since the platform’s July launch with 17 pilot banks.

How to read it: A weekend payment sounds like a minor operational fix, and the modesty is exactly why it matters. Correspondent banking’s core constraint has never been technological in the interesting sense - it is that settlement follows business hours, and business hours are a calendar problem, not a cryptography problem. Removing it without introducing a new settlement asset, no stablecoin, no new instrument, just the existing deposit made programmable, is the version of tokenisation likely to scale fastest inside regulated banks, because it changes nothing about capital treatment or deposit insurance. The absence of disclosed size is a reasonable caution: a second live transaction on a new platform is a capability demonstration, and capability demonstrations are not yet volume.

B. Stablecoins & Payments

2. Switzerland’s CHF stablecoin consortium doubles its membership

What happened: The Swiss franc stablecoin initiative behind CHFD expanded to nine participants, with SIX, the market-infrastructure operator behind the Swiss stock exchange, and TWINT, the country’s dominant payments app, joining existing members UBS, PostFinance, Sygnum, Raiffeisen, Zürcher Kantonalbank, Basler Kantonalbank and Swiss Stablecoin AG to test CHFD in a controlled live sandbox.

How to read it: This is a broad domestic coalition for a single-currency stablecoin - a market-infrastructure operator, a retail payments network, cantonal banks and a global systemically important bank testing the same instrument together, rather than a single issuer courting adoption after the fact. Non-dollar stablecoins have mostly progressed through single-bank or single-issuer launches; a nine-member sandbox spanning infrastructure, payments and multiple bank types is a different and more durable model, because the addressable use cases (retail payments via TWINT, exchange-linked settlement via SIX) are designed in from the start rather than added later.

3. MoneyGram extends stablecoin balances onto the Visa network

What happened: MoneyGram launched a stablecoin-backed Visa card in Colombia, its first live market, letting customers hold a USDC balance, built with Rain, Crossmint and Stellar, and spend it anywhere Visa is accepted, converting to local cash at MoneyGram locations. A physical card is planned for late 2026.

How to read it: The product itself is unremarkable - a prepaid-style card funded by a stablecoin balance. The detail is in the assembly: a remittance incumbent, a card network and three separate infrastructure providers (custody and issuance via Rain, wallet infrastructure via Crossmint, settlement via Stellar) combining to make a stablecoin balance indistinguishable from a bank balance at the point of sale. This is the same pattern we have tracked in the payments layer generally - the card network supplies distribution, the stablecoin supplies the balance and the customer experiences neither.

4. Circle acquires Tazapay for $400 million

What happened: Circle agreed to acquire Tazapay, a Singapore-based cross-border payments platform with more than 60 banking and fintech partners and over 100 payout markets, for $400 million, extending the stablecoin issuer’s global banking coverage.

How to read it: Circle is buying distribution and banking rails rather than building them, which is consistent with the position we have tracked since its OCC and New York trust charters: unable to win merchant-level distribution from Visa, Circle is assembling the adjacent infrastructure it can own outright. A payments platform with 100 payout markets converts USDC from a balance sheet asset into something that can actually reach a recipient’s local currency, which is the harder half of any cross-border stablecoin proposition.

5. Coinbase brings stablecoin rails to community banks; Fasanara and Tether back a $400 million credit fund

What happened: Coinbase partnered with Moov to offer stablecoin payment acceptance, settlement, payouts and real-time funding to more than 1,000 US community banks and credit unions without those institutions building their own crypto infrastructure. Separately, Fasanara Capital and Tether launched StableFund - an evergreen private-credit vehicle anchored by $400 million in co-investment from both sponsors - deploying USDT-linked asset-backed lending across fintech platforms in more than 60 countries and targeting up to $3 billion in outside institutional capital.

How to read it: Both extend stablecoin infrastructure toward institutions that would otherwise be excluded by scale rather than by mandate. Community banks lack the resources to build stablecoin rails themselves, and Coinbase’s white-label approach is the same distribution logic we have seen with larger banks, applied further down the size curve. StableFund is the more structurally interesting of the two: a $3 billion target for USDT-linked private credit is tether’s reserve model turned outward, deploying stablecoin liquidity into real-economy lending rather than holding it in Treasuries. The credit quality of the underlying loans, not the stablecoin wrapper, is what will determine whether that target is met.

C. Regulation & Policy

6. Iran’s central bank builds a sanctions-evasion reserve in USDT

What happened: Blockchain analytics firm Elliptic said Iran’s central bank has acquired at least $507 million in USDT, describing a reserve built outside conventional banking channels to support the collapsing rial, with much of it initially routed through Nobitex, Iran’s largest exchange. Tether retains the power to freeze such holdings and has a record of exercising it, having blocked $344 million in April and a further $131 million in July.

How to read it: The two facts sit in tension and both matter. A sanctioned state building a stablecoin reserve is the clearest evidence to date that USDT functions as a genuine dollar substitute for an economy locked out of the conventional banking system - exactly the use case regulators cite when arguing stablecoins need tighter oversight. But Tether’s freezing power is the same architecture regulators are trying to legislate into every issuer: a centralised administrator that can unwind a sanctioned holding unilaterally, without a court order, at the protocol level. Iran’s reserve is real exposure, and Tether’s freeze capability is a real constraint on it. Institutions should read this as evidence for the case that centralised stablecoins are more governable than critics assume, not less, even as it is evidence that they are already being used for exactly the purpose sanctions regimes exist to prevent.

D. Market Structure / Infrastructure

7. A Bitcoin sidechain loses 95% of its reserve to a federation exploit

What happened: On 6 September, attackers withdrew roughly 4,000 of the 4,200 BTC, around $320 million, held in the federation wallet backing Liquid Network, Blockstream’s Bitcoin sidechain. They exploited an inflation bug in Elements, the sidechain’s underlying software, to mint L-BTC that should not have existed and redeem it for real Bitcoin through the network’s peg-out mechanism. The transaction carried valid signatures from 11 of the federation’s 15 members, so the federation’s own security infrastructure approved it. The attackers left an onchain message claiming to be white-hat researchers and returned roughly 3,400 BTC over the following days, but around 598 BTC, worth close to $47 million, remains outstanding. Blockstream has refused to treat the remainder as a bug-bounty payment, stating publicly it will not pay a ransom and calling the retention of funds theft rather than responsible disclosure. Ledger’s chief technology officer, Charles Guillemet, said keeping roughly 600 BTC “did not fit a genuine bug-bounty arrangement.” Block production has resumed, but peg operations, the mechanism that lets BTC move in and out of the sidechain, remain shut.

How to read it: The federation model is the point. Liquid’s security rests on a defined group of members whose signatures the network trusts by construction, a design that predates the permissioned-ledger architectures now being built by Swift and the tokenised-deposit banks above, and this incident is what happens when that trust boundary contains a software bug rather than a malicious member. Eleven valid signatures approved a withdrawal that should have been impossible, which means the federation’s defence was never really eleven independent checks, it was one shared piece of software with eleven copies of the same flaw. The distinction between this and the partial-return “white-hat” framing matters for anyone assessing custody risk in permissioned or federated systems generally: a group that negotiates over what it keeps after an unauthorised withdrawal is engaging in extortion with better manners, not security research, and the correct read of a federation’s safety is the quality of its code, not the number of its members.

Focus Signal: The fund is BlackRock; the position is Standard Chartered.

Analyse the stack, not the product. BlackRock’s Hong Kong authorisation is being reported, correctly, as a BlackRock story - the world’s largest asset manager bringing its tokenised money-market franchise to a third region after the US and Europe. That is the headline layer, and it is real. It is also the layer that tells you least about where the durable economics of this deal actually sit.

Standard Chartered holds every function around the fund that is not the fund itself. It is trustee, custodian and administrator. It also owns, through Anchorpoint Financial, the stablecoin - HKDAP - that investors will use to subscribe and redeem; one of only two Hong Kong dollar stablecoin licences the Hong Kong Monetary Authority has granted, selected from 36 applicants. This launch is Standard Chartered’s first commercial deployment as a bank distributor of that stablecoin. So the same institution supplies the custody rail, the administrative infrastructure and the settlement currency that moves cash into and out of BlackRock’s product. BlackRock’s name sells the fund, but Standard Chartered’s infrastructure is what the fund actually runs on, and that infrastructure does not disappear if the fund underperforms, changes manager or is joined by a second, third and fourth tokenised fund using the same rails.

This is worth reading as a single example of the argument we return to across the tokenisation stack. Layer 3, the fund itself, the ETF-style access product, is where the brand sits and where the coverage concentrates, and it is a real and growing part of the market. But Layers 1, 2 and 4, origination, registry, and settlement, are where a position compounds independent of any single product’s success. Standard Chartered has now assembled a comparable position across three separate developments this year: institutional Bitcoin and Ether trading infrastructure in the UAE, tokenised-deposit settlement through Swift’s ledger (discussed above) and now custody, administration and the settlement stablecoin for BlackRock’s Hong Kong fund. None of these announcements individually reads as a strategy but collectively across custody, trading and tokenised cash, in three separate jurisdictions, they do.

Yield & Staking Watch

A standing section tracking where institutional yield is actually being built.

Strive adds to its Bitcoin treasury. Strive acquired 1,375 Bitcoin for roughly $109 million at an average price of $79,281, bringing total holdings to 24,531 Bitcoin, with CEO Matt Cole noting that notional outstanding on the company’s SATA preferred is approaching $1 billion. The purchase itself is unremarkable against the size of the holding; the preferred-stock obligation approaching $1 billion is the figure worth tracking, because it is the fixed liability against which any future forced-sale dynamic, of the kind we have flagged across this cohort, would eventually be tested.

The Scorecard

We keep score on our own calls, monitoring events and developments against our house view:

• Standard Chartered assembling an infrastructure position across custody, trading and settlement(a pattern first noted with its Euroclear and Swift work in August, and its HKDAP distribution role from 24 August): Strengthening, and now spanning three distinct functions in three jurisdictions. This week’s Hong Kong fund adds custody, trusteeship and stablecoin distribution to a UAE trading build-out already under way. See Focus Signal.
• Value accrues at the settlement and collateral layer(our central structural thesis): Strengthening. Citi and DBS removing the weekend constraint from cross-border dollar settlement is a second consecutive fortnight of evidence at the plumbing layer, following the ECB’s Pontes timeline - banks acting on the same logic central banks are now stating explicitly.
• Ethereum, Solana and Canton pulling ahead in the institutional layer-one contest(a standing position, not touched by this week’s news): Holding, no fresh evidence either way. Worth stating plainly rather than silently repeating a stronger verdict: nothing this week moved institutional issuance or settlement volume toward or away from these chains. The position stands on the evidence already accumulated, not on anything new.
• Stablecoins as a genuine dollar substitute, with a governability caveat that cuts both ways(our stablecoins thesis, first flagged with a reserve-margin friction in August): Tension, not simple strengthening. Iran’s central bank building a $507 million USDT reserve outside conventional banking is strong evidence for the thesis in its purest form - a sanctioned economy treating a dollar stablecoin as usable money. But Tether’s demonstrated freezing power over that same reserve is evidence the other way: the instrument central to the growth thesis is also more centrally governable, and therefore more exposed to enforcement action, than the market narrative around stablecoins usually allows. Both readings are correct at once, and neither should be dropped in favour of the other.
• Federated and permissioned custody models carry a distinct trust-boundary risk(a sub-thesis of our settlement-layer work, sharpened this week): New evidence, cautionary. Liquid Network’s federation model failed not because a member acted maliciously but because a shared software flaw let a minority-viable signature set approve an invalid withdrawal. As tokenised-deposit and permissioned-ledger architectures scale - the same category of design, differently implemented - the quality of the underlying code, not the number of participants, is the relevant safety measure.

What We’re Watching

The specific future events that would confirm or challenge our standing positions:

• Whether the first Bitcoin treasury company is forced through a complete sell-then-rebuy cycle - still unobserved, and the genuine stress test of whether the leveraged-treasury model has a stable equilibrium or a reflexive doom loop
• Whether institutional tokenised issuance and settlement activity spreads meaningfully beyond Ethereum and Solana to the likes of Avalanche, Aptos or Sui - the break condition for our view that the two chains pull away from the rest of the public chain field
• Whether privacy-preserving activity on public chains, permissioned rollups or Canton-to-public bridges grows materially - the break condition for treating institutional settlement as diverging by design rather than converging on one architecture
• Whether a second sovereign issuer frames tokenised settlement as a monetary-sovereignty question in the way the ECB has - the test of whether Frankfurt’s position was a single instance or the start of a pattern among central banks
• Whether any regulated crypto product without a yield or staking component visibly loses share to one that has it - the direct evidence our yield-as-competitive-frontier thesis currently lacks
• Whether the remaining 598 BTC from the Liquid Network exploit is recovered, and whether Blockstream’s post-mortem confirms the Elements inflation bug as the sole cause

The Institutional Digital Assets Weekly is produced by Giltspur Research from our proprietary research engine and intelligence platform, drawing on verified regulatory, institutional and market sources.

Giltspur is a London-based research and advisory firm specialising in the institutional adoption of digital assets and distributed ledger technology. We provide independent advice, research and education to financial institutions at every stage of their digital asset journey. For enquiries: info@giltspur.co.uk

If this email was forwarded to you, subscribe free to receive The Institutional Digital Assets Weekly.

Originally published on Substack