BIS compared thirty years of dollar deposits with eight years of stablecoin flows across 184 countries, and the capital controls that cut deposit dollarization by up to 32 points don't touch stablecoins at all.

The answer sits in the gap between two numbers: 25 to 32 percentage points on one side, and roughly zero on the other.

Rated against the paper's own dataset:

Three of four rows are measurement. The fourth is the question every central bank in the dataset now answers on its own.

A saver in a capital-controlled economy wanting dollar exposure has two paths available.

• The perimeter is the variable. Capital controls are enforced at the licensed-institution boundary. Stablecoins move peer to peer around that boundary, not through it.
• The comparison isn't hypothetical. Both datasets measure the same behavior, holding dollars instead of local currency, across the same country set.

Capital controls attach to a chokepoint a regulator can supervise. A foreign-currency bank account has one: the licensed institution that opens it. A dollar stablecoin held in a self-custodied wallet has no equivalent chokepoint, which is exactly why the 25-to-32-point gap in deposit data disappears in stablecoin flow data.

This is the diagnosis tracee's own client base is already working from. Central banks and regulators building CBDC pilots or tokenized-deposit rails in dollarized economies have treated capital controls as one lever among several against stablecoin substitution. BIS's data says that specific lever does not apply to the stablecoin channel, which narrows the real policy menu to building a domestic alternative competitive enough to be chosen, not restricting the dollar option away.

None of this is new in kind, physical dollar cash has always moved around capital controls too. What's new is that BIS now has the dataset to show stablecoins replicate that evasion at internet speed, with no border to cross and no queue at a currency exchange.

• Correlation, not a natural experiment. The comparison holds capital-control classification against dollarization outcomes across countries; it does not isolate capital controls from the other differences between countries that impose them and countries that don't.
• The two datasets don't share a clock. Deposit dollarization data runs 1990 to 2019; stablecoin flow data runs 2017 to 2024. The overlap is five years, not thirty, so this is closer to a cross-era comparison than a single synchronized panel.
• On-chain volume isn't proof of household savings. Chainalysis inflows capture wallet-to-wallet activity, which mixes exchange and institutional flows with the retail saver the deposit-dollarization literature is built around.
• No country-level breakdown, no policy recommendation. The public findings name an aggregate gap, not which economies drive it, and stop at diagnosis. Central banks get the finding, not the fix.

tracee's briefing on Isabel Schnabel's June 2026 remarks at the Bank of Korea conference covered the same worry from the ECB's side: stablecoins cement dollar dominance and carry money-market-fund-style run risk if redemption ever gaps. That briefing described the concern. This paper gives it a number.

It also sits next to BIS's own Project Agorá, which tracee covered in May, where the same institution tests tokenized central-bank reserves and commercial deposits as the rail regulators would rather savers use instead. The commercial side of the same gap is already being built for: Trace Finance's $32M raise, which tracee covered in June, is a regulated bank-rail bet on exactly the corridor this paper describes, Brazil and other emerging markets where dollar stablecoins already move money capital controls can't stop.

Three institutions, one shared premise: the dollarization the BIS paper measures is not a temporary gap in enforcement. It is a structural feature of a bearer-like instrument, and the response has to be a better rail, not a tighter perimeter.

BIS just took away one argument central banks have used for capital controls in dollarized economies, that restricting FX-account access slows dollarization. It still does, for deposits. It does nothing, per this dataset, for the dollar stablecoins increasingly doing the same job. That reframes the question tracee's central-bank and regulator clients are already asking. It is no longer whether to restrict the stablecoin channel, the data says that lever doesn't work there, it is whether the CBDC or tokenized-deposit alternative on the table is competitive enough that a saver chooses it over a dollar stablecoin wallet.

Watch three things:

• Whether an emerging-market central bank cites this paper directly in a capital-control review or a CBDC business case over the next two quarters.
• Whether the IMF follows with its own cross-check. Its April 2026 tokenization note already flagged reserve and settlement fragmentation risk from stablecoins; this paper hands it a dollarization dataset to extend that argument.
• Whether Chainalysis's next flow update shows the gap widening, as stablecoin market capitalization, already past $300 billion and still concentrated in USDT and USDC, keeps growing faster than the regulated alternatives meant to compete with it.