Banks across the US, the UK, and Europe finally have a legal path to issue stablecoins, custody Bitcoin, and settle tokenized funds, yet the capital rulebook that governs it all still treats a Bitcoin position as something close to a guaranteed loss.
Under the Basel Committee’s cryptoasset standardwhich has been live in member jurisdictions since January 1, 2026, unbacked crypto is in the most punitive bucket in the whole framework, carrying a 1,250% risk weight. Once you push that through Basel’s 8% minimum, you end up with a bank holding capital equal to its full exposure, a dollar of equity set aside for every dollar of Bitcoin on the books.
That gap between permission and capital cost is the part of crypto regulation almost nobody’s paying attention to, even though it’s the thing that’s going to decide how much digital-asset business actually ends up inside regulated banks.
The standard was created in a different time, back when supervisors were mostly trying to keep crypto out of the banking system altogether, and it was shaped by everything that was going wrong then: the opacity around stablecoin reserves, the exchange collapses, the contagion that ran through FTX and Celsius.
The phase banks are walking into now is very different, because tokenized deposits, stablecoin reserve management, custody, and on-chain settlement are now part of regulated balance sheets. You can already see it in JPMorgan’s JPMD deposit token, Citi’s Token Services, and the tokenized deposit work underway at HSBC.
The Committee itself can tell the fit has loosened, which is why it opened an expedited review of targeted parts of the standard back in November 2025, noted progress through February and May of 2026, and has promised an update later this year.
The capital math that prices Bitcoin like a certain write-down
Basel itself doesn’t write law in any single country, but it sets the template that national regulators in the US, EU, UK, Canada, Japan, Singapore, and Hong Kong use to decide how much equity a bank must hold against any given asset.
The cryptoasset chapter, known as SCO60takes everything a bank might touch and sorts it into tiers, and the logic is fairly intuitive. Group 1a is for tokenized versions of traditional assets; Group 1b is for stablecoins that pass strict reserve and redemption tests; and both can be treated more or less like their conventional equivalents. Group 2 catches everything that fails those conditions, splitting into Group 2a for assets liquid enough to hedge and Group 2b for the rest.
The weight attached to each of those tiers is really where the business case lives or dies. A low capital charge allows a bank to hold or finance an asset cheaply, whereas a high one forces it to set aside equity that could be working much harder elsewhere. At the very top of the scale, the charge gets high enough that the whole activity stops making economic sense.
That’s what the 1,250% figure on Group 2b does in practice, so a $100 million Bitcoin position ends up eating roughly $100 million of capital, and because there’s no netting of long and short exposures, the real bill usually runs higher still once you stack buffers and supervisory add-ons on top.
On top of all that, SCO60 layers on an exposure cap that has no real equivalent anywhere else in the Basel framework, which is to say a bank’s total Group 2 holdings are supposed to stay under 1% of its Tier 1 capital, and the moment it crosses 2% every single Group 2 position gets dragged into the punitive 2b treatment at once, with hedging recognition stripped away entirely.
This is what industry has pushed back on the hardest, and bodies like ISDA and the GFMA told the Committee back in August 2025 that whole sections of the standard were overly conservative and punitive, pressing for a recalibration before it ever reached full adoption.
To be fair to the Committee, all of that caution made complete sense at the moment they finalized the rules, because supervisors were staring at frozen client funds, weak offshore controls, reserve assets nobody could actually verify, and tokens that would routinely fall 70% to 80% in a single drawdown. Basel’s whole mandate is to stop banks from importing those kinds of losses into the deposit base.
The strain you’re seeing now is that the bucket they labeled as crypto exposure has stretched to cover wildly different things: a tokenized US Treasury funda fully reserved payment stablecoin, a custodied client coin, and a straightforward Bitcoin trade have almost nothing in common once you look at the real risk underneath.
We also have the problem of scale, as tokenized real-world assets on public chains have already surpassed $16 billion, with government securities making up the largest share.
This means that a tokenized Treasury bond on a public blockchain can fail the Group 1 conditions on a technicality and drop straight into Group 2b, where Basel has filed all purely speculative tokens.
What’s the cost for crypto if the capital math holds?
Probably the best sign that these categories are buckling is that the world’s biggest economies have simply stopped agreeing on them.
The Trump administration outright rejected SCO60, with Executive Order 14178 and the July 2025 digital-asset report describing that fixed 1,250% weight as anti-innovation and anti-competitive, and pointing US regulators toward a risk-based approach tied to how these markets actually behave.
Europe is going the other way and holding the cautious linefolding the Basel treatment into its CRR3 capital rules and the technical standards that its banking authority is still drafting.
And because Basel rules only ever take effect through national adoption, you can end up with the same tokenized asset carrying a heavier capital charge in Frankfurt than it does in New York, and a global bank having to build separate digital-asset products for separate jurisdictions just to deal with it.
That fragmentation cuts both ways for a bank trying to figure out where to commit, because loose rules let crypto risk seep into the deposit base while punitive ones just push the activity toward firms sitting outside the bank perimeter.
The thing people sometimes miss is that most of what banks actually want here is fee-based and light on the balance sheet, things like custody, fund administration, stablecoin reserve management, tokenized-deposit settlement, collateral services, and market making in regulated products. The capital treatment determines which of those lines meet an internal return hurdle, since a heavy charge on inventory or financing can close off the ones that need a balance sheet to run in the first place.
Stablecoins are really where all this pressure concentrates, because a fully reserved payment token, a bank’s own tokenized deposit, and a tokenized money-market fund each carry different legal claims and sit on the balance sheet in different ways. This means Basel has to price redemption, reserve, liquidity, and enforceability risk separately for every one of them.
The US has already leaned hard into that split, with GENIUS keeping tokenized deposits under ordinary deposit treatment while payment stablecoins are subject to a dedicated regime of their own.
When you remember that the stablecoin market is now somewhere around $320 billion and almost entirely dollar-denominated, you start to see why this classification carries so much weight. It effectively determines how much of the settlement layer banks get to hold themselves and how much continues to flow through nonbank issuers. It’s essentially the same deposit-flight worry that sits behind the US banking lobby’s warning about trillions potentially migrating out of insured accounts.
And those two paths (a harsh capital regime versus a more risk-sensitive one) lead to two very different markets. If the charge remains punitive, regulated issuers lean even harder on nonbank infrastructure, tokenized markets keep scaling outside traditional banking channels, and crypto-native firms hold on to a larger share of settlement for themselves.
If the treatment turns more risk-sensitive, tokenized deposits become a credible rival to payment stablecoins, tokenized Treasuries start reaching investors through bank distribution channels, and much of that activity drifts back toward the regulated core, where supervisors would rather have it.
Most of the time, crypto regulation reaches people through court fights, enforcement actions, and licensing bills. But banks answer to a much slower and heavier rulebook, and for them the deciding factor really comes down to the capital cost, the cold calculation of whether a given line of business still clears its return hurdle once you count the equity charge against it.
The Basel review isn’t going to settle all of that in one stroke, and it’s happening because the old dividing line between speculative tokens and regulated settlement has worn through. Until somebody redraws that line, the banks best equipped to bring crypto inside the regulated system are going to have every reason to keep working from its edge.