Why Banks Are Finally Getting Into Crypto, and What They’re Really Building
- Banks are commercializing the parts of crypto that look like existing bank businesses. Custody, payments, tokenized securities, settlement infrastructure. They’re not broadly betting on crypto prices.
- The stablecoin market hit $317 billion in April 2026 after growing more than 50% during 2025. That number spooked banks more than any crypto price cycle ever did, because it represents real competition for payment economics they currently own.
- JPMorgan’s Kinexys is processing over $5 billion daily. Citi Token Services moved from pilot to live commercial use. BNY has active digital asset custody in the US. These are production systems.
- The 2025 US regulatory reset, plus MiCA’s full application in the EU at the end of 2024, gave banks a supervisory perimeter clear enough to build product against. That clarity is what unlocked the current wave.
- Banks win on trust, compliance infrastructure, and balance-sheet credibility. Crypto-native firms win on speed, asset coverage, and on-chain depth. The market is converging toward a partnership rather than a clean knockout for either side.
They’re Not Joining Crypto. They’re Defending Their Territory.
JPMorgan put it pretty cleanly in early 2025. The bank doesn’t offer cryptocurrencies or stablecoins as such, but does offer on-chain value-exchange products and tokenization infrastructure. The framing is intentional and it tells you everything about the strategic logic underneath the recent wave of bank announcements. Banks aren’t buying Bitcoin because someone in management suddenly fell in love with decentralization. They’re building regulated versions of things they already do well, custody, cross-border money movement, securities settlement, collateral management, and putting them on digital rails because those rails are faster, cheaper, and increasingly where institutional clients expect the work to happen.
The defensive angle is real and worth taking seriously. A stablecoin market growing more than 50% in a single year to hit $317 billion is real competition for payment flows, treasury management, and settlement services that banks currently monetize through legacy infrastructure. Moving now is partly about capturing new revenue, but it’s equally about making sure these products don’t get built entirely by non-banks while traditional institutions wait for perfect regulatory clarity that was never going to arrive on a comfortable timeline.
You can see this thinking everywhere if you read the bank press releases carefully. None of them are leading with “we believe in crypto.” They’re leading with payment volumes, T+0 settlement, programmable cash, tokenized securities issuance, and 24/7 cross-border liquidity for corporate treasury teams. The vocabulary is the vocabulary of operations and infrastructure.
What Changed
Four things converged at roughly the same time, and the combination is what pushed banks from observing to building.
Institutional demand matured into something useful. EY’s 2026 institutional digital-asset survey found that 73% of respondents planned to increase allocations, 81% preferred registered spot vehicles over direct crypto exposure, 61% used multi-custodian setups, and 68% preferred partnering with a crypto-native firm to augment traditional capabilities. These institutions don’t want unregulated access to crypto. They want regulated access wrapped in the risk management, reporting, and servicing environments they already use day to day. Banks are well-positioned to provide the wrapper, especially when paired with specialist partners for the technical layer underneath.
Stablecoins started displacing real payment flows. Federal Reserve analysis has documented why this is relevant to banks and policymakers. Stablecoins are backed by relatively safe assets and increasingly relevant for cross-border payments and monetary transmission. McKinsey’s 2025 payments analysis argues they address genuine pain points in legacy rails. Speed, cost, transparency, limited operating hours, fragmented cross-border infrastructure. When the competition is solving real problems at scale, waiting becomes a strategy with measurable costs attached.
Custody and compliance infrastructure grew up to a point where it’s genuinely buildable. BNY integrated Fireblocks and Chainalysis into its custody stack. US Bank uses NYDIG as a bitcoin sub-custodian. The market for specialist digital-asset technology is mature enough that banks don’t have to build the full stack themselves anymore, which dramatically shifts the build-versus-buy calculus inside bank technology committees.
Regulatory clarity finally arrived after years of ambiguity. The OCC confirmed in 2025 that crypto custody, certain stablecoin activities, node participation, and customer-directed execution are permissible for national banks when managed soundly. The FDIC dropped prior-approval requirements in March 2025. The Federal Reserve withdrew its non-objection expectations in April 2025. SAB 121 got rescinded. In the EU, MiCA has been fully in force since December 2024, giving banks a passportable framework across the bloc. None of this means crypto became easy for banks. It means the supervisory perimeter is finally clear enough to build product against, which was genuinely not true in 2022 or 2023 when uncertainty alone was enough to kill internal proposals.
What the Major Banks Are Building
JPMorgan Chase: Kinexys is the clearest example of the strategy in action. Over $5 billion processed daily through the platform. A USD deposit token launched on a public blockchain in 2025. A cross-chain delivery-versus-payment test completed with Chainlink and Ondo Finance. This is bank-controlled digital cash and tokenized debt infrastructure rather than a crypto exchange business. JPMorgan isn’t trying to compete with Coinbase. It’s trying to make sure that when institutional payments and settlement move to programmable rails, JPMorgan owns the rails its clients use.
BNY Mellon: BNY’s approach is unapologetically infrastructure-layer. Live bitcoin and ether custody since October 2022. Fireblocks for wallet infrastructure underneath, Chainalysis for compliance monitoring layered on top. The goal is to become the multi-asset recordkeeper across both traditional and digital assets, leveraging the same custody and administration franchise it has built and refined over decades. The bet is that custody is custody regardless of asset class, and the institution that runs the most reliable custody at scale captures the recurring fee economics.
Citi: Citi Token Services moved from pilot to live commercial use, with tokenized cash enabling 24/7 cross-border liquidity for corporate clients without forcing them to manage tokens directly themselves. Citi extended the same thinking into private-market tokenization through SDX, where the bank acts as custodian and tokenization agent. The logic is straightforward enough. If custody, settlement, and issuance all move onto programmable rails, a global bank with relationships across all of those functions can capture more of the workflow than any single-product fintech could on its own.
HSBC, UBS, and DBS: These three illustrate how the addressable market has widened beyond custody alone into something more comprehensive. HSBC offers tokenized gold and is building tokenized securities custody. UBS launched a tokenized warrant on Ethereum, a tokenized investment fund, and a blockchain-based multi-currency payment pilot under the UBS Digital Cash brand. DBS has tokenized transaction banking as a live product and tokenized structured notes on Ethereum, with clients trading over $1 billion in crypto options and structured notes through DBS in just the first half of 2025. Standard Chartered is acting as digital-asset custodian and settlement agent for TP ICAP’s institutional venue. State Street became the first third-party custodian on JPMorgan’s Digital Debt Service, enabling T+0 settlement for blockchain-based debt securities.
The revenue thesis across all of them looks remarkably similar when you strip away the branding. Custody and administration fees, cross-border payment economics, tokenized issuance fees, and compliance and reporting infrastructure that clients will pay for rather than build themselves. Banks have always made money from being the trusted middle. They’re rebuilding that role for the next iteration of financial infrastructure.
Where Regulation Stands
The regulatory picture varies enough by jurisdiction that bank strategy genuinely differs based on where a business is headquartered.
| Jurisdiction | Current regime | What it means for banks |
| US | OCC, FDIC, Federal Reserve all reset guidance in 2025, SAB 121 rescinded, stablecoin legislation advancing | Materially easier than 2022, still safety-and-soundness driven |
| EU | MiCA fully in force from December 2024 | Passportable framework across the bloc, major advantage for scaling digital-asset services |
| UK | Legislation finalized late 2025, FCA authorizations from September 2026, full regime October 2027 | Clearer perimeter, still a staged build rather than a live harmonized system |
| Singapore | Expanded Payment Services Act (2024), DTSP regime (2025), demanding AML and consumer-protection requirements | Strong commercial infrastructure with serious compliance expectations |
| Switzerland | FINMA custody guidance (2026), SNB extended Project Helvetia | One of the most explicitly bank-friendly environments for custody and tokenization |
Regulatory clarity doesn’t translate to light-touch regulation. It means the supervisory perimeter is defined enough to build product against without getting blindsided. FINMA emphasized in its 2026 guidance that technical expertise, robust infrastructure, and bankruptcy protection are required, especially when custody involves providers based abroad. US agencies have been consistent that digital-asset custody must meet the same safety-and-soundness standards as any other custody activity, which is what most bank compliance teams wanted to hear in the first place.
The Custody and Compliance Layer
Most bank digital-asset strategies succeed or fail at the operational layer. Building custody that meets regulatory expectations means legal segregation of customer assets, bankruptcy-remoteness analysis, dual authorization for key access, cold storage or hardware security modules, multi-party computation or multisig structures, transaction monitoring, sanctions screening, and incident response procedures. That’s a significant build even when you’re sourcing components from specialist vendors, and it’s why the early movers have been institutions that already had mature operational infrastructure to extend.
The fact that 61% of institutions in EY’s survey prefer multi-custodian arrangements reflects a genuine market dynamic. Concentration risk in digital asset custody is a real concern, and that’s good news for established custody banks with the infrastructure to support multiple clients simultaneously without the operational fragility that single-product custodians sometimes have.
On DeFi interoperability, the pattern across major banks is controlled gateways rather than full permissionless exposure. JPMorgan’s Chainlink and Ondo test, UBS’s Ethereum-hosted warrant, both show regulated banks connecting regulated assets to interoperable networks where governance can still be maintained and counterparty oversight is possible. This is selective and deliberate engagement rather than open-ended embrace, and it’s probably the right model for institutions with fiduciary obligations.
Banks vs Crypto-Native: An Honest Assessment
Banks have balance sheets, fiduciary credibility, existing institutional relationships, decades of compliance infrastructure, and the regulatory standing to offer custody for the assets that large institutions are most willing to entrust them with. Coinbase, Anchorage, BitGo, and Fireblocks have broader asset coverage, faster development cycles, richer on-chain tooling, and cleaner API-first architecture. These two sets of capabilities aren’t converging into one player who has both. They’re becoming complementary in a market where institutions want both at once.
The EY finding that 68% of institutions prefer partnering with a crypto-native firm to augment traditional capabilities essentially describes the end-state. Banks provide the regulated wrapper and the institutional credibility. Crypto-native firms provide specialist on-chain depth and product velocity. The remaining question is which institutions capture the most economic value in that partnership arrangement, and whether the white-label or sub-custody dynamics tilt in favor of the bank or the crypto-native firm over time.
McKinsey estimates roughly $2 trillion in tokenized assets by 2030, excluding crypto and stablecoins from that number. State Street’s research suggests many institutions expect 10 to 24% of their investments to be tokenized by 2030. These are large enough numbers that banks can justify continued infrastructure investment even if the realistic tokenization timeline slips by a few years against the optimistic forecasts.
Frequently Asked Questions (FAQ)
Are banks really buying Bitcoin?
Mostly not in any meaningful proprietary sense. The primary activity is custody, tokenized cash, payment infrastructure, and securities services. Some banks support client-directed bitcoin custody or registered exposure products, but the strategic framing is almost universally about infrastructure rather than price exposure.
Why did this accelerate in 2025?
Regulatory constraints eased across multiple major jurisdictions simultaneously, the stablecoin market scaled to a size that made banks pay attention as a competitive threat, and institutional demand for registered digital-asset access stayed consistent through volatility cycles.
What are banks saying yes to first?
Custody, tokenized deposits, blockchain-based payments, tokenized bond and fund issuance, and digitally serviced securities. All of these map onto existing bank revenue lines with a reasonably understood compliance framework, which is exactly why they came first.
Are tokenized deposits the same as stablecoins?
No. Tokenized deposits are bank liabilities represented on blockchain rails, preserving traditional bank-money relationships and deposit protections. Stablecoins are separately issued tokens backed by reserve assets, typically from non-banks. Regulators and central banks treat the distinction as economically significant for good reasons.
Which region offers the clearest framework for banks right now?
The EU offers the most harmonized passportable framework through MiCA. Switzerland is arguably the most explicitly bank-friendly environment for custody and tokenization. The US improved materially in 2025 but still requires bank-by-bank supervisory engagement. The UK has the framework in place but isn’t fully live yet.
Are banks connecting to DeFi?
Selectively and cautiously, through controlled gateways with compliance checkpoints. JPMorgan’s Chainlink and Ondo test and UBS’s Ethereum warrant are examples of the controlled approach. Full permissionless participation isn’t the model and probably never will be for institutions with fiduciary obligations.
