Congress Is Weighing. The Basel Committee Already Answered.

The operative word in "Congress weighs letting banks hold crypto, issue stablecoins" is not "crypto." It is "weighs." Weighing is what a legislature does instead of acting. It signals intent, absorbs lobbying, and produces headlines. It commits no capital. The article carrying this headline yields four claims: one fact — a legislative body is deliberating — and three projections, each assuming the deliberation becomes law, the law becomes bank balance sheets, and the balance sheets become spot bids. A four-step causal chain presented as a single event. No bill number. No sponsor. No committee stage. No named regulator. The absence of those identifiers is the most informative data point in the piece. I have audited enough pre-launch documentation to recognize the shape. The pitch describes the destination and omits the terrain. The terrain is four years old. The United States has been weighing bank crypto custody since at least 2020, when the OCC issued interpretive letters permitting national banks to provide crypto custody. That permission never resolved into scale, because in March 2022 the SEC published Staff Accounting Bulletin 121. SAB 121 requires a bank holding crypto for clients to recognize the asset and a corresponding liability on its own balance sheet. Custody is supposed to be an off-balance-sheet service. SAB 121 converts it into a balance-sheet event, with capital attached. The capital side is governed by the Basel Committee's cryptoasset exposure standard. Assets are sorted into groups. Group 2b — unbacked cryptoassets such as bitcoin, with no effective hedge — carries a 1,250% risk weight. Run the arithmetic. A bank custodying $1 billion in BTC for clients generates $12.5 billion in risk-weighted assets. At an 8% minimum CET1 ratio, that consumes roughly $1 billion of regulatory capital. Custody fees on the same $1 billion run in the low single-digit basis points annually. The capital charge exceeds the revenue by orders of magnitude. Permission without a capital rule change is permission to lose money. The legislative record confirms the pattern. The Clarity for Payment Stablecoins Act cleared the House Financial Services Committee in 2023. FIT21 passed the full House in May 2024, then stopped. The Senate has not moved a comprehensive market-structure bill. The EU's MiCA framework phased in across 2024, making the United States the laggard rather than the leader. "Weighs" has been the steady state for four years. There is a mechanical reason these headlines recur. "Congress weighs" is a low-cost sentence to publish and a high-engagement sentence to read. It requires no bill text to verify and no vote to falsify. Each cycle it resets the same expectation: that regulatory permission is the binding constraint on institutional capital. Based on my audit experience, the binding constraint is almost never permission. In 2017 I modeled 0x's advertised liquidity depth against testnet fills and found roughly 40% of it was wash-traded. The number that is published is rarely the number that clears. The same holds here: the adoption figure that circulates is rarely the adoption figure that settles. The transmission from legislative permission to deployed capital passes through three gates. Each removes most of the hypothetical demand. Gate one is capital treatment, and it is arithmetic, not ideology. A bank's willingness to hold an asset is a function of risk-weighted capital cost against risk-adjusted return. Under Group 2b treatment, crypto is the most capital-expensive asset a bank can touch — costlier than equity, costlier than venture exposure, costlier than anything on a standard trading book. Until the Basel treatment changes, or the US writes a differentiated domestic rule, no risk committee signs off. "Banks may hold crypto" and "banks may hold crypto at a 1,250% risk weight" are the same sentence with the operating clause deleted. Gate two is legal isolation. Custody is a bankruptcy question dressed as a technology question. When a custodian fails, client assets must be identifiable and segregated so they fall outside the estate. Banks are engineered around deposit insurance and the bankruptcy code's treatment of deposits. Crypto is neither a deposit nor a security in the classic sense. The 2022–2023 failures made the gap concrete: Silvergate and Signature collapsed with custody relationships no one had cleanly mapped to a legal wrapper. In 2021 I flagged the Terra stability mechanism as mathematically unsound because its redemption path assumed conditions that would not hold under stress. Chaos reveals itself only when the noise stops. A custody statute that does not specify the isolation mechanism is a promise with no address. Gate three is client demand. "Banks will hold crypto" is not "bank clients will ask banks to hold crypto." The demand already exists and already has a route. Spot bitcoin ETFs absorbed the allocators who wanted regulated exposure inside a familiar wrapper. The marginal bank entrant arrives at a market already served, competing against a product with no custody-capital penalty. If the ETF already clears the demand, then bank custody is a margin business, not a growth business. Then the stablecoin clause. A bank-issued stablecoin is a money-market fund with a token interface. Reserves are cash and short-dated Treasuries. Value capture accrues to the issuing bank as net interest margin, not to holders. That is a legitimate product and a fundamentally different object from an algorithmic stablecoin; it should be modeled as a payment rail, not a speculative asset. But it also means the token carries no upside narrative. Utility is the vacuum where hype goes to die. The offshore layer complicates it further. The stablecoin market is split between an offshore incumbent holding the majority of circulating supply and a compliant challenger. Bank issuance could compress the offshore share — but only if the US reserve and capital rules are competitive. If domestic rules mandate 100% cash and T-bill reserves while prohibiting yield pass-through to holders, an offshore issuer that holds T-bills and rebates the carry retains a structural edge. Regulation does not automatically win; it wins when the economics are better. The article does not mention reserve composition, yield treatment, or the depository-insurance question — the three parameters that determine whether a bank stablecoin is a product or a compliance artifact. The DeFi transmission is two-directional and under-modeled. Compliant bank stablecoins could deepen institutional liquidity inside permissioned venues. The same compliance layer raises the KYC floor and can push permissionless DeFi toward the margin, where its stablecoin supply thins. The article treats "stable digital asset markets" as an unambiguous good. Stable markets and deep markets are not synonyms. The bulls are right about the direction, and it is worth stating plainly. Institutional adoption is measurable, not rhetorical: spot ETF flows are on the record, and the trend does not reverse on a single committee vote. Regulatory clarity — even partial — reduces the tail risk of enforcement-driven de-banking, a real cost crypto businesses have paid. And bank custody would hand institutional allocators a counterparty their compliance committee already recognizes. That is genuine value, not narrative. Where the bulls overreach is time. History repeats, but the code changes the syntax. The 2023–2024 sequence ran committee passage, headline, stall — three times. The marginal reaction to each iteration decayed. Code executes exactly as written, not as intended; legislation does the same. The text that clears a committee is not the text that becomes a rule, and the rule is not the product. Pricing the endpoint while ignoring the pipeline is how you buy the top of a headline. There is a second blind spot. The bulls treat "permission" and "adoption" as adjacent. They are separated by a risk committee, a capital model, and a client conversation — none of which appear in a press release. Track the identifier, not the headline. A bill number, a markup date, and a line in the Basel implementation or an SAB 121 repeal are the only signals that move this from deliberation to structure. Until those appear, the article is a temperature reading. Temperature is not a thesis. The question is not whether Congress is weighing. It is whether anyone is counting the weight.