Author: Pink Brains; Compiler: Jiahuan, ChainCatcher

Market structure

Ask ten cryptocurrency users what a digital bank is and you'll likely get the same answer: a card that lets you spend stablecoins. Ask ten developers and the answers will quickly diverge. Some are developing non-custodial wallets with Visa payment capabilities; others are forking Aave and calling it a savings account; and a few are looking to obtain a full banking license.

Monthly transaction volume for cryptocurrency cards will grow from approximately $100 million at the beginning of 2023 to over $1.5 billion by the end of 2025 (a CAGR of 106%). The market now exceeds $18 billion on an annualized basis. In 2025, bank card consumption linked to stablecoins will reach $4.5 billion, a year-on-year increase of 673%.

But look at who is actually handling the volume. On-chain bank card data shows that RedotPay, an Asia-based custody platform, dominates 60% of the market, with transaction volume roughly four times that of the next 13 competitors combined. Those DeFi-native, self-custody digital banks pale in comparison on transaction volume charts.

The deeper story, however, is that the crypto-friendly giants have reached the same goal. Between December 2025 and March 2026:

• Coinbase applied for a national trust license

• NuBank receives conditional OCC (Office of the Comptroller of the Currency) approval from U.S. National Bank

• PayPal applies to create a PayPal bank

• Revolut has received a full UK banking license and is pursuing a US license

• Kraken becomes the first crypto company to have a master account with the Federal Reserve

• 11 companies applied for OCC trust banking licenses within 83 days, including: Circle, Ripple, BitGo, Paxos, Fidelity, Bridge, Crypto.com, Morgan Stanley, Payoneer, Zerohash, Protego

More than 50 crypto-digital banks are already online. The global digital banking market is expected to reach $552 billion by 2026 (according to The Business Research Company).

We try to map the digital banking landscape—not just who is developing what, but who has the economic model to survive.

Core Conflict

Two conflicts at the center shape the competitive landscape in digital banking.

The first conflict is economic. 76% of traditional digital banks are unprofitable. Those that are successful (Nubank, Revolut, SoFi) make money not through card purchases, but through loan books and net interest income. The handling fee is just a stepping stone, and credit is the real core business.

Now, crypto banks are joining the fray with fees and cash back, the very revenue model that caused the first generation of fintech companies to fail. Stablecoins make the situation worse: they squeeze FX margins to close to zero.

The second conflict concerns user choice. CT circles praise self-custody, DeFi yields, and non-custodial wallets. But the volume of bank card transactions on the chain tells a different story. The vast majority of crypto card purchases flow through custodial platforms, not because users don’t understand self-custody, but because a frictionless onboarding experience trumps the pursuit of financial sovereignty when you just want to buy a cup of coffee.

History has always followed this pattern: just as webmail became popular before encrypted email, Dropbox became popular before self-hosted storage, and custodial exchanges dominated the market before DeFi - it is still an open question whether crypto banks will follow the same trajectory.

Four prototypes of digital banking

Instead of using "Web2 and Web3" to divide (this can only reflect technology and has no inspiration for business models), a more useful perspective is to examine the moat, unit economic benefits and ceiling of a digital bank.

1. Crypto-friendly & banking first

The core profit point of digital banks lies in credit income and traffic conversion, rather than as a simple payment channel.

Nubank posted revenue of $15.8 billion in fiscal 2025, 85% of which came from interest income. Credit card interest contributed $4.6 billion and loan interest $4.8 billion. With monthly revenue of $15 per active user and service cost of just $0.80, that’s a 19x return. SoFi, which received a banking license in 2022, has seen quarterly net interest income grow from $94.9 million to $617 million in four years, with deposit costs 181 basis points lower than warehouse funding costs, resulting in annualized savings of about $680 million. Revolut delivered revenue of £3.1bn across five business streams in 2024, with no single business line accounting for more than 30%, and the trading/wealth business line growing 298% year-on-year.

Licensed digital banks deliberately limit stablecoins to the payment field, because the loan ledger is the source of their economic benefits. Revolut has not yet introduced stablecoin balance earnings; the UK FCA sandbox test it participated in in February 2026 used its proprietary stablecoin as a payment infrastructure, not a savings product. SoFi’s stablecoin (SoFiUSD, launching in December 2025) is a settlement channel that operates through the Mastercard network.

This restraint is based on a specific market condition: on-chain yields are currently uncompetitive. Aave v3's USDC yield recently stood at 2.6% APY, which is lower than SoFi's 3.3% savings APY and Revolut Ultra's 4.25%. However, on-chain yields are subject to cyclical compression. During periods of active DeFi activity, Aave USDC reached 8-10%, and Ethena’s funding rate-driven yield was much higher than this. This gap is cyclical, and when it widens again, the competitive dynamics will change.

2. Business and social super applications

MercadoPago, Grab, WeChat, Alipay. They were not originally intended to be banks, but rather to embed finance into business applications. The moat is not the product itself, but distribution channels and behavioral data – which allows them to conduct more precise credit assessments than any bank.

MercadoPago's credit revenue will grow from $246 million in 2020 to $5.9 billion in 2025, a 24x increase in five years. Grab's loan portfolio will grow from $185 million in 2022 to $1.3 billion by the end of 2025, with financial services revenue reaching $348 million in fiscal 2025.

Both have explored stablecoins. MercadoPago launched Meli Dólar (MUSD) in Brazil and expanded to Chile and Mexico, but MUSD has a float market cap of only $65 million, less than 0.4% of its $19 billion AUM. Grab has partnered with StraitsX for stablecoin settlement, allowing tourists to achieve instant settlement in Singapore dollars via the XSGD stablecoin at GrabPay merchants in Singapore.

Neither has explored stablecoin returns. This is the breakthrough point for crypto-native players, and it is also the most underestimated gap in this field.

Whop deserves attention here. It’s not a digital bank yet, but a marketplace for creators. But after a $200 million investment in Tether (valuing it at $1.6 billion), creators can accept USDT, hold stablecoins, and settle without a bank. Plasma’s integration with Aave provides stablecoin earnings targeting an audience of 18.4 million users and $3 billion in annual creator revenue. MercadoPago was not a digital bank in 2003 either, just a third-party guarantee for the marketplace—the financial relationship that comes with commerce. Whop is currently in that same infancy and has been built on an encrypted network from day one.

The implication for digital banks with bank cards at the core is that the most lasting financial relationship may not start from finance at all, but from e-commerce.

3. Transaction priority

Robinhood, Coinbase, Binance, Kraken, Bybit, OKX – started with centralized cryptocurrency trading and expanded to crypto banking. Every platform in this group is explicitly building out the banking layer to generate revenue that doesn’t require reliance on bull markets.

Robinhood is the most complete representation of this: Total platform assets grew about 70% year-over-year to $324 billion, and net deposits hit a record $68 billion. Coinbase is making a big push into digital banking: its own L2 blockchain (Base), a wallet with swipe capabilities, crypto-backed loans based on Morpho, Bitcoin-collateralized loans in partnership with Better, and a pending application for a trust license. Kraken has both a trust license and a main account with the Federal Reserve.

These platforms start from transaction revenue that already has economies of scale, and superimpose banking services on top of it. Stablecoin digital banks go in the opposite direction: start with a meager fee and hope to stack everything on top of that – that’s much harder.

4. Stablecoin priority (encryption native)

Ether.fi, Gnosis Pay, RedotPay, KAST, Holyheld, Bleap, Ready, Tria, Cypher, Payy and dozens more. These platforms leverage the low operating costs of stablecoins and the composability of DeFi as back-end product infrastructure. The value proposition is clear: self-custody; DeFi yields (5-15% APY in active markets, 3-4% on traditional savings); near-instant cross-border payments based on stablecoins with extremely low FX fees; global portability without geographic restrictions.

Stablecoin-first digital banks have the most obvious structural advantages in emerging markets and cross-border use cases. But the weaknesses are just as real: None has broken through in scaling unsecured lending; they compete on the thinnest profit layer (fees) while subsidizing user acquisition with token-funded cashback; and stablecoins make it worse—squeezing FX margins and settlement fees to near zero, eroding the revenue streams that kept early digital banks afloat.

Infrastructure

Most crypto banks are just front-ends built on shared infrastructure. Understanding this technology stack is critical to assessing its moat.

Card Networks (Visa, Mastercard): Despite being almost even in terms of number of projects (more than 130 each), Visa accounts for over 90% of on-chain card transaction volume through early collaboration with crypto-native infrastructure providers. This is where the industry-wide risk lies in a single point of failure — if Visa changes its policy on crypto projects, slows expansion, or raises rates, the entire industry’s economic ledger could change overnight.

Card issuers (Rain, Reap, Baanx, StraitsX): regulated bridges between the on-chain world and traditional finance. The most important structural development is the emergence of full-stack card issuers – companies that hold primary membership directly with Visa/Mastercard, bypassing traditional sponsoring banks.

Most crypto banks share the same backend. Rain powers Ether.fi, RedotPay and Avalanche Card. If Rain had technical glitches, regulatory issues or a strategic shift, the entire industry would be affected. A report by Solus Partners analyzing 19 platforms points to infrastructure concentration and vendor dependence as systemic risks – exactly what Synapse risks are for crypto-digital banks.

(Note: Synapse is a financial technology infrastructure company in the United States. In 2024, it filed for bankruptcy due to a fund isolation gap, which resulted in the funds of dozens of cooperative platform users who relied on its services being frozen, becoming a landmark warning case in the industry about the risk of infrastructure concentration.)

The threat of wallet-native stablecoins

An often overlooked competitive dynamic: major wallet providers are issuing their own stablecoins specifically to fund bank card purchases, building closed-loop ecosystems to capture the value that would otherwise be earned by independent digital banks.

At the end of Q3 2025, MetaMask launched mUSD and Phantom launched CASH, both designed as funding mechanisms for their respective debit card products. These wallets do not rely on users to hold USDC or USDT, but instead build a closed loop - users convert assets into wallet-native stablecoins, which are then used to pay for card purchases. Early data shows very different trajectories: Phantom’s CASH grew steadily from about $25 million in September to about $100 million in late December; MetaMask’s mUSD fell to about $25 million, a 75% loss, after peaking at nearly $100 million in early October.

If a wallet succeeds in making its native stablecoin the default funding source, channel fees, FX spreads and reserve earnings will be withheld by the wallet platform itself. MetaMask, Phantom, and Coinbase Wallet already have user relationships in place—adding digital banking functionality is an extension of the product line, not a new product. Independent crypto banks may lose a significant part of their value proposition as a result.

Economics Difficulties

76% of traditional digital banks are unprofitable. Crypto-native digital banks inherit the same broken model — and stablecoins make it worse, not better. The lesson for bank-first players is clear: payments are just a distribution channel, not the core business. This is evidenced by Nubank's 85% share of interest income; and net interest margin (NIM) expansion driven by SoFi's license. An encrypted digital bank that uses bank card consumption as its core revenue engine is tantamount to building a building on sand.

The sustainable model is to use bank cards as a channel for user acquisition while monetizing through higher-margin on-chain finance: DeFi earnings, exchange transactions, structured products, and lending.

Five things that change the rules of the game

1. On-chain credit score

In the crypto world, wallet transaction history, DeFi usage patterns, repayment behavior on lending protocols, pledge duration, transaction frequency, and protocol diversity can all be used as inputs for credit assessment. No crypto bank has yet achieved this at scale. Whoever cracks it first will be copying Nubank’s approach on a permissionless on-chain infrastructure.

2. Crypto-native companies obtain full banking licenses

Not a trust license (custody only), but a full license that allows taking deposits and making loans. This will allow crypto-native businesses to build loan ledgers funded by stablecoin deposits at a lower cost than warehousing financing facilities.

3. Supervise and clarify the legality of income

The regulatory direction of major markets is consistent: stablecoin issuers are not allowed to pay earnings. The United States and Europe have clearly drawn this red line, and major Asian markets such as Japan, Singapore, and Hong Kong have also adopted similar conservative stances.

4. Agent-driven finance

AI agents perform financial operations on behalf of users - rebalancing portfolios, optimizing returns, managing payments, executing cross-protocol strategies. Mastercard had six crypto partners in 2024, expanding to more than 25 in 2025. Visa has launched Smart Commerce Connect, which enables AI agents to make purchases on behalf of users at merchants around the world. Whoever can build the best agency infrastructure on top of the stablecoin network will win the next wave of traffic in the e-commerce field.

5. Make on-chain operations feelless

Encrypted digital banks still rely on bank card networks, but payment terminal technologies that bypass them already exist: QR code payment for stable currency settlement, NFC touch payment without the involvement of Apple Pay or Google Pay, and physical card swiping for on-chain settlement. Projects like OpenPasskey (based on Base) are proving this path is possible: have your own ISO-issued IIN, P-256 cryptography, a completely non-custodial crypto card – three payment methods, no Visa required, no bank required.

Who will win?

We don't know yet. But the key variables of victory and defeat are already there.

Licensed digital banks have proven their economic model, and they have an advantage where credit drives user relationships, which is most developed markets. Stablecoin-first digital banks bring a globally portable U.S. dollar, local stablecoins for emerging markets, access to DeFi yields, and retroactive rewards, but current data shows that users still prefer simplicity and ease of use over crypto-native ideas. Commercially embedded players may have the deepest moats because they already control distribution channels—but layering cryptographic capabilities on top of mature infrastructure is costly, requires user education, and is highly dependent on regulatory clarity.

The value captured by the infrastructure layer (card issuer, custody, fiat deposits and withdrawals, core banking system, blockchain settlement, KYC/AML) is destined to exceed that of any consumer brand. More than 40 stablecoin cards are rolling in on token-subsidized cashback, have no real commercial moats, and share the same infrastructure — and most of them won’t survive the next two years.

The landscape of encrypted digital banks is at an inflection point.