European Banks Have No Bitcoin to Sell

• There is no European bank Bitcoin sell-off, because direct holdings are near zero. Euro area investors held €17bn in crypto products in Q4 2024, only €3.4bn of it across the entire financial sector.
• Basel’s 1,250% risk weight forces a bank to hold one euro of capital per euro of Bitcoin, with Group 2 exposure capped at 1–2% of Tier 1. The charge makes direct holdings irrational before it even binds on 1 January 2026.
• IFRS treats Bitcoin as an intangible asset under IAS 38, so gains stay invisible and losses print, which is a reporting shape no listed treasury will accept.
• Banks route demand through wrappers instead. Intesa Sanpaolo’s ~$235m is ETF and token exposure; BBVA, KBC and Société Générale sell client access; euro stablecoins now carry 82% of institutional OTC volume.
• The one live variable is the Basel Committee’s expedited review of the crypto standard, opened in November 2025, whose direction is genuinely unresolved.

Every few months a version of the same story circulates, that European banks are quietly dumping their Bitcoin, unwinding positions ahead of some rule or some crash, and it always runs into the same problem when you go looking for the trade. There is almost nothing to sell. Euro area banks hold a rounding error in crypto, and the number sits in plain sight in the ECB’s own data, so the sell-off narrative collapses the moment it meets a balance sheet.

The more useful question is not why banks are selling Bitcoin but why they were never long it in the first place, and what they are doing instead now that regulated wrappers have arrived. The answer is a story about capital math, an accounting standard written in 2003 for software licenses, and a slow migration into products that let a bank touch crypto without ever holding a coin. None of it looks like a fire sale. All of it explains why the fire sale was never going to happen.

The position that does not exist

Start with the size of the thing everyone assumes is there. In the fourth quarter of 2024, euro area investors held around €17 billion in crypto-asset-related investment products, of which roughly €3.4 billion sat with the financial sector, a bucket the ECB defines to include banks, funds, insurers and pension funds together. Households held €10 billion of that total. Direct holdings of Bitcoin on bank balance sheets, the coins themselves, are close to zero and have been the entire time. There is no large long position sitting on European bank books waiting to be liquidated, so any headline about banks selling is describing the disposal of something that was never accumulated.

In crypto, one bad click is final. Learn to verify before you lose.

The rumour tends to grow from a real signal that gets misread. On-chain analysts periodically flag large net outflows from wallets they cannot fully identify, and the leap from “big holder sold” to “a bank sold” is easy to make and impossible to support, because the entities behind those flows are usually exchanges, funds, or long-term holders rather than regulated European lenders. A bank that wanted to trim any position would move it over the counter, off any public tape, so the flows people point to as evidence of bank selling are the flows least likely to belong to a bank in the first place.

That framing corrects a lot of loose commentary at once. When a European bank appears in a story about crypto, it is almost never selling and almost never holding spot Bitcoin as a treasury asset. It is offering a client a product, custodying someone else’s coins for a fee, or building payment rails in a stablecoin. The gap between “banks are exiting Bitcoin” and “banks were never in Bitcoin” is the whole point, because the reasons they stayed out are structural, and those reasons now shape every crypto decision a European bank makes.

Capital charge ends the debate

The single fact that governs all of this is the risk weight. Under the Basel Committee’s finalised standard, unbacked crypto-assets like Bitcoin and Ether fall into the highest-risk bucket, Group 2b, and carry a 1,250% risk weight, with a bank’s total Group 2 exposure soft-capped at 1% of Tier 1 capital and hard-capped at 2%. Run the arithmetic and the intent becomes obvious. At an 8% minimum capital ratio, a 1,250% risk weight means a bank must hold one euro of its own capital against every euro of Bitcoin it owns. The asset gets treated, for capital purposes, as if it could go to zero, because the framework’s authors decided that was the prudent assumption.

A euro of Bitcoin funded entirely by capital is a euro that cannot be lent, leveraged, or put to work anywhere else, which turns even a modest position into a drag on return on equity that no treasury desk will defend. The exposure cap makes the point again from the other direction, because even a bank that wanted a meaningful Bitcoin book is legally prevented from building one above 2% of Tier 1. The standard was engineered to make direct holdings both expensive and small, and it succeeded before it even took binding effect, since the implementation date lands on 1 January 2026 and banks priced in the charge well ahead of it.

Europe has been busy turning that global standard into hard law through CRR3, and in August 2025 the European Banking Authority published its final draft technical standards on how institutions calculate and aggregate crypto exposure values, covering credit, counterparty, market and CVA risk across asset-referenced tokens, tokens referencing other crypto, and unbacked crypto. The detail worth noting is that the EBA dropped the prudent-valuation requirement for fair-valued crypto exposures after consultation and clarified the aggregation rules for the exposure limits, which is the machinery that decides how a hedged position nets against an unhedged one. For a bank, this is the difference between a workable trading book and a capital sinkhole, and it is being written now, so treasury teams are watching the rule-making far more closely than the spot price.

The accounting standard nobody designed for this

Capital is only half the deterrent. The other half is how a bank has to carry Bitcoin on its books once it owns it, and here the problem is IAS 38. Under IFRS, a cryptocurrency held directly is an intangible asset, the same broad category as a patent or a brand, and the default treatment is cost less impairment. A bank that buys Bitcoin at 40,000 and watches it run to 90,000 books no gain, because intangibles carried at cost are not marked up through profit and loss. The same bank watching Bitcoin fall to 25,000 must recognise an impairment loss, and under the standard that impairment generally cannot be reversed through profit even if the price recovers.

The asymmetry is brutal and entirely one-directional. All of the downside hits earnings, none of the upside does, which is close to the worst possible accounting shape for a volatile asset a listed institution has to report every quarter. A corporate treasurer looking at that treatment does not need a view on Bitcoin’s long-term value to reject it, because the reporting alone converts every rally into an invisible gain and every drawdown into a printed loss. This is one of the quieter reasons US GAAP moved to fair-value measurement for crypto from 2025, and it is also why a European bank that wants price exposure reaches for a wrapper that carries fair-value treatment rather than the coin itself. The instrument solves an accounting problem as much as a custody one.

IAS 38 does allow a revaluation model where an active market exists, which in theory could let a bank carry Bitcoin at fair value, and in practice almost nobody uses it for crypto, because the revaluation route pushes gains through other comprehensive income rather than profit while still routing impairments through the income statement, so the asymmetry survives in a slightly rearranged form. Layer tax on top and the case gets no better, since a bank pays corporate tax on realised crypto gains as ordinary income across most European jurisdictions, with no unified treatment and plenty of local variation on how losses can be used. A treasurer weighing a volatile asset that is punished on capital, awkward on accounting, and unhelpful on tax does not have a hard decision to make, and the decision has been the same for years.

What the banks are doing instead

Follow the money that is moving and it lands in regulated products while the spot coin market gets left to specialists. The clearest tell came from Italy’s largest bank, Intesa Sanpaolo, whose Q1 2026 regulatory filing showed its crypto exposure more than doubling to around $235 million, spread across Bitcoin and Ether ETFs plus an XRP position, up from roughly $100 million at the end of 2025. It is worth being precise about what that money bought, because the original research muddled it. The Bitcoin exposure sits in exchange-traded funds, an $81 million position in the ARK 21Shares Bitcoin ETF and roughly $25 million in BlackRock’s iShares Bitcoin Trust, and BlackRock’s product is IBIT, while the Grayscale GBTC it gets conflated with is a different fund from a different issuer. Intesa holds shares of funds that hold Bitcoin, which sidesteps the intangible-asset accounting and hands custody to the ETF issuer, and that is exactly the shape the rules push a bank towards.

Buy bitcoin and stocks. Signup to Coinbase today

The second thing banks are doing is selling access rather than taking exposure. BBVA in Spain, KBC in Belgium and Société Générale in France have all rolled out crypto trading for retail clients over 2025 and into 2026, and the common thread is that the bank earns fees while the client carries the position and the price risk. This is crypto as a distribution business, and it fits the capital framework perfectly, because facilitating a customer’s trade consumes almost none of the balance sheet that holding the asset would. A bank can answer client demand for Bitcoin without ever putting Bitcoin in its own Group 2b bucket, and that is the whole appeal of the model.

The third and fastest-growing thing is settlement, and the instrument doing the work is the euro stablecoin. Institutional flow data from Finery Markets’ Q1 2026 review shows stablecoins now accounting for 82% of OTC trade volume, with euro-denominated EURC growing 41-fold year on year as MiCA compliance pulled European desks towards regulated euro settlement. The regulatory tailwind here is real, since a MiCA-licensed e-money token receives far gentler prudential treatment than unbacked crypto, and a euro stablecoin behaves like a payment instrument a bank can reason about, where an unbacked coin behaves like a speculative asset it has to impair. A consortium of large European banks has moved to launch a MiCA-compliant euro stablecoin of its own, and the direction of travel is telling, because the same institutions that will not touch spot Bitcoin are happy to issue and settle in a regulated euro token that sits inside the rulebook rather than outside it.

Custody is the quiet through-line under all of it. The value of crypto-asset custody services provided by euro area banks rose from around €400 million in 2023 to €4.7 billion in 2024 on the ECB’s numbers, which is a real business growing fast, and it involves holding client assets off the bank’s own balance sheet for a fee. A bank can build a crypto custody franchise, an ETF distribution desk and a euro stablecoin rail, and generate fee income across all three, without ever running the proprietary Bitcoin position that the sell-off story assumes it is unwinding.

None of this restraint reflects a small or shrinking market around them. Europe is one of the world’s largest crypto economies, with regional transaction volume peaking at roughly $234 billion in December 2024 on Chainalysis figures, and the number of regulated crypto investment products euro area investors can buy climbed from 215 in late 2023 to 294 a year later on the ECB’s count. Demand is large, growing, and increasingly served by compliant wrappers, and the banks have positioned themselves as the distributors and custodians of that flow rather than as principals taking the price risk. A retail client in Madrid or Milan can get Bitcoin exposure through their bank far more easily than two years ago, and the bank booking that relationship still carries none of the coin.

Where a real reduction would come from

Suppose a European bank genuinely did cut crypto exposure, the question becomes where that would show up and what would drive it. The honest answer is that most of the plausible reductions are small, opportunistic, and nearly impossible to see from outside, and the confident sell-off headlines never cite a trade for that same reason.

The branch-versus-parent distinction does some work here that gets overlooked. A branch is not a separate legal entity, so its assets consolidate straight into the parent, and a branch cannot hold its own capital buffer or set its own risk appetite. A US or UK parent that wanted to trim crypto held through a European branch would see the effect land on group capital, since a branch has no ring-fenced balance sheet of its own, and it would likely route the sale through OTC channels that never print to a public tape. If any bank-linked selling is happening at the margin, this is the shape it takes, invisible and small, which is precisely why nobody can point to it.

Branches also inherit a double compliance burden that discourages the activity in the first place. The EBA has been clear that a crypto establishment in a member state, even something as minor as a single machine, pulls the operator under local AML and counter-terrorist-financing rules on top of its home-country regime, and MiCA’s travel-rule obligations on crypto transfers add another layer of monitoring cost. A branch weighing whether to run a crypto desk is looking at dual supervision, full travel-rule KYC, and a capital charge, and the rational answer for most of them is to not bother, or to route clients into the group’s central crypto unit instead of building anything locally.

There is one scenario where a bank would sell crypto quickly, and it is worth naming precisely because it is the closest thing to a real sell-off trigger. In a liquidity squeeze, a treasury reaches first for whatever is most volatile, most capital-hungry, and easiest to exit, and a small Bitcoin or Bitcoin-ETF line ticks all three boxes, so it would go early in any genuine stress event. The important qualifier is scale, because a position that starts at a rounding error cannot produce a market-moving sale on the way out, and a forced liquidation of a €200 million ETF book by one Italian bank is not the same event as European banking abandoning Bitcoin. The stress case explains why banks keep these positions small and liquid in the first place, which is the opposite of the accumulation the sell-off story needs to be true.

Macro conditions have leaned the same way. Through 2024 and 2025 European policy rates sat high enough that conventional assets paid a competitive yield, which drained some of the speculative appeal out of a non-yielding, capital-punitive asset like Bitcoin. When cash and short-dated paper pay you to hold them and Bitcoin costs you a euro of capital per euro of exposure, the relative-value case for a bank tilts hard towards the boring option, and it stays there until either rates fall or the capital treatment changes.

The future of finance is here. Trade crypto and more

The Basel review that could move the whole board

The one genuine open question is whether the capital regime itself softens, and that debate is live. In November 2025 the Basel Committee met in Mexico City and agreed to expedite a targeted review of its cryptoasset standard, driven by the explosive growth of stablecoins and by the awkward fact that major jurisdictions including the United States and the United Kingdom have declined to adopt the framework as written. A global standard that Washington and London ignore is not much of a global standard, and the Committee knows it.

The direction of that review is genuinely contested, and it would be dishonest to pretend otherwise. Some of the consultation language points towards tightening specific elements, particularly around stablecoins on public blockchains, while the broader pressure from non-adoption and market growth pushes towards recalibration that would make certain exposures cheaper to hold. By February 2026 the Committee said only that the work was progressing and that an update would come later in the year, so the honest position today is that the framework is being reopened without a settled outcome. If the review lowers the charge on hedged or collateralised positions, expect banks to inch back towards selective crypto activity. If it tightens the stablecoin treatment, expect them to lean harder into ETFs and licensed e-money tokens instead. Either way, the spot Bitcoin book stays small.

What to watch

The sell-off framing gets the story backwards, because there is no meaningful long position on European bank balance sheets to unwind, and the reasons for that absence are the same reasons the banks are unlikely to build one soon. A 1,250% risk weight, a capped exposure limit, an accounting standard that prints every loss and hides every gain, and a compliance load that doubles for anyone operating through a branch, together they make holding Bitcoin a decision no European treasury desk can justify to its board.

What is worth watching is not whether banks sell Bitcoin but which regulated channel they route the demand through, and the current answer splits three ways, into ETF distribution, custody, and euro stablecoin settlement. Intesa’s ETF book, the retail trading desks at BBVA and Société Générale, and the euro stablecoin rails showing up in institutional OTC flow are all the same trade in different clothes, a way to earn from crypto while the coin itself, and its capital charge, stays with somebody else. The Basel review is the wildcard that could widen or narrow those channels, and until it reports, the picture holds. European banks are not dumping Bitcoin. They arranged their affairs so they would never be holding enough of it to dump.