Basel's cryptoasset rules run on a classification test. Every exposure a bank holds gets sorted into one of two groups, and which group it lands in decides the capital charge. Group 1 is the pass. Group 2 is the fail, and failing is expensive.
The chapter is SCO60, "Cryptoasset exposures", part of the consolidated Basel Framework and aimed at internationally active banks. Its text as in force from 1 January 2026 sets out the whole scheme.
Read it at the level of the paragraph numbers, though, because almost every popular summary collapses four categories into two.
There are four. SCO60.6 splits Group 1 into 1a, tokenized traditional assets that meet the classification conditions in SCO60.8 to SCO60.19, and 1b, cryptoassets with effective stabilization mechanisms that meet those same conditions. Group 2 is everything that fails them, and it splits too: 2a covers assets that fail the conditions but pass the Group 2a hedging recognition criteria, and 2b is all the rest.
Bitcoin sits in Group 2 because it has no stabilization mechanism to test. A stablecoin can sit there as well, if its mechanism does not hold up under the conditions.
Group 2b carries a 1250% risk weight. SCO60.84 applies it to the larger of a bank's aggregate long or aggregate short position in each cryptoasset, and SCO60.86 states what the number is engineered to do: ensure banks hold minimum risk-based capital at least equal in value to the Group 2b exposure. Hold the asset, hold its full value in capital. That is the mechanism behind the line you see repeated everywhere about direct Bitcoin being capital-expensive for a regulated bank, and it is a design choice with a stated purpose.
The same paragraph is candid about where the approach runs out. The formula applies the 1250% weight to short positions too, for simplicity, and the Committee acknowledges that shorts and certain other exposures could in theory produce unlimited losses, so the capital required could in some circumstances be insufficient. Supervisors are told to consider a Pillar 1 add-on where a bank has material short or derivative exposures, calibrated by running the market risk framework and the BA-CVA framework and taking the higher result.
Then there is a second constraint that most explanations leave out, and it does more work than the risk weight.
Under SCO60.117 a bank's total Group 2 exposure should not generally be higher than 1% of its Tier 1 capital, and must not exceed 2%. SCO60.118 turns those two numbers into a ratchet. Breach 1%, and the excess above the threshold gets Group 2b treatment. Breach 2%, and every Group 2 exposure the bank has gets Group 2b treatment, including the 2a positions that had passed the hedging criteria. A bank that drifts over the second line does not pay a little more on the overage. It just repriced its whole book.
Group 1 is no free pass either. SCO60.34 starts from the observation that a Group 1b cryptoasset has to be redeemable, and if the entity performing the redemption fails, the asset may become worthless, so the bank's capital treatment turns on what kind of claim it holds on that redeemer. Where a Group 1b asset references a pool of traditional assets, SCO60.33 sends the bank to the equity-investments-in-funds rules in CRE60, with look-through and mandate-based approaches available and a 1250% fall-back if neither applies. The punitive weight is reachable from inside Group 1.
One historical detail is worth getting right, since it is the kind of thing summaries garble. The Committee's oversight body endorsed the finalized standard on 16 December 2022, describing the treatment for unbacked cryptoassets and stablecoins with ineffective stabilization mechanisms as conservative, and members agreed to implement it by 1 January 2025. The publication itself carries that same 2025 date. The text sitting in the framework carries an in-force date of 1 January 2026, and the framework's own change log lists SCO60 among the chapters that came into effect then. Cite the 2022 document for what binds a bank, and you are quoting a date that moved.
None of this is a verdict on any network or token. It is a capital standard describing what a bank must set aside, and a bank finding an exposure expensive is a statement about bank balance sheets.
Classification is also not a one-time event. SCO60.6 requires cryptoassets to be screened on an ongoing basis, so an asset that qualifies for Group 1 can fail the conditions later and move, taking its capital treatment with it.
Disclosure: I work in digital asset advisory and hold digital assets. Nothing here is investment, legal, or tax advice.
No comments:
Post a Comment