The Tria crypto neobank: banking ergonomics, no banking protections

Revolut, N26 and Monzo are neobanks because they hold banking or e-money permissions somewhere and operate under a supervisor. That is what makes an IBAN work, what puts your money inside a deposit guarantee scheme up to the local limit, and what gives you a complaints route with teeth.

Tria uses the word to describe the experience rather than the legal status: a self-custodial app for trading, earning and spending, with card rails attached. Its own framing — bank-grade security and compliance combined with on-chain freedom — is a claim about engineering and partners, not a claim to hold a banking licence itself.

This distinction is not pedantry, it is the whole risk profile. In a neobank, the institution owes you your balance. In a self-custodial app, nobody owes you anything, because nobody has it. Those are opposite arrangements dressed in the same visual language, and the second one is only better if you actually want the responsibility.

Deposit insurance. In the EU, bank deposits are covered to €100,000 per depositor per institution; comparable schemes exist elsewhere. A self-custodial balance has no equivalent, and it does not need one in the same way — there is no institution that can fail with your money — but it also means nothing compensates you for your own mistake.

Payment reversal. A bank can recall a mistaken transfer and force a merchant chargeback. On-chain settlement is final. Card payments retain scheme-level dispute rights, which is a genuine partial protection, but a wrong-address crypto transfer has no analogue at all.

Credit and overdraft. Neobanks lend. A self-custodial app can offer collateralised borrowing at best, which is the opposite of an overdraft: you must already have assets to borrow against them.

Accountability. If a bank freezes your account unfairly there is an ombudsman. If you lose your key material there is nobody — and if a smart contract holding your yield position is exploited, your claim is against an attacker you cannot identify.

• Kept Card scheme dispute rights on card purchases, via the issuer.
• Lost Deposit guarantees, transfer recall, overdrafts, an ombudsman for your balance.
• Gained Nobody can freeze the balance, and no institution failure can consume it.

What functions now: holding assets across many chains, swapping between them with routing handled by Tria’s BestPath system, spending through the Visa card programme, and on-chain yield products. Tria also markets self-custodial cross-chain perpetual futures, which is a trading feature rather than a banking one and carries a completely different risk profile.

What is marketed as forthcoming: US bank accounts for citizens, and instant fiat on and off ramps in 100+ countries. Both are described on Tria’s site as coming soon. Treat “coming soon” as a roadmap statement, not a capability — build your plans on the features that work today, in your country, on your account.

The practical gap this creates is the fiat boundary. Until local ramps exist for you, converting between your local currency and crypto still happens somewhere else: an exchange, a broker, a payment provider. That is not a flaw in the app so much as a reminder that a self-custodial neobank sits on top of the traditional system rather than replacing it.

On-chain yield comes from a small number of real sources: lending demand, market-making spreads, staking rewards, and incentive emissions paid in a protocol’s own token. The first three are sustainable at modest rates. The fourth is a marketing budget, and it stops when the budget stops. A double-digit stablecoin yield in 2026 is usually a blend, weighted toward the fourth.

The risks are equally concrete. Smart-contract risk is the possibility that the strategy contract is exploited — “audited” narrows this without eliminating it, and an audit applies to a specific contract version on a specific date. Liquidity risk is the possibility that exiting takes longer than the interface implies during stress. Counterparty and oracle risk arrive when a strategy depends on external price feeds or venues.

None of that makes yield products unusable. It makes them investments rather than savings accounts, and the honest way to size them is as a portion of capital you can afford to see impaired — not as the place your rent money waits.

Europe’s MiCA framework brought crypto-asset service providers and stablecoin issuers inside a licensing regime, with reserve requirements for issuers, disclosure obligations, and passporting across member states. It deliberately does not license software that merely lets you control your own keys: providing a self-custodial wallet is not custody, so a purely self-custodial app is largely outside the perimeter.

That is why the compliance picture in an app like this is layered. The wallet is unlicensed and does not need to be. The card is issued by a regulated issuer. Fiat conversion involves a licensed payment or crypto-asset service provider. Yield products may or may not be offered by a regulated entity depending on structure and jurisdiction.

For you, the useful question is not “is this app regulated?” but “which regulated entity stands behind the specific function I am about to use, and in which country?” Terms of service normally name them. If they do not, that is information too.

The structure that works is boring and layered. Keep long-term holdings on a hardware wallet you connect rarely. Keep one to two months of spending in the app, funded with stablecoins on a low-cost network. Keep a genuine bank account for salary, rent and anything with a direct debit, because a self-custodial app cannot host a standing order to a landlord.

Then treat the app as what it is best at: converting between assets cheaply, spending on a card without a withdrawal cycle, and holding a modest yield position you have consciously sized. That is a real improvement over the previous decade of moving money between an exchange and a bank every time you wanted to buy something.

What does not work is treating it as your only account. Not because the app is untrustworthy, but because single points of failure are a bad idea regardless of who runs them — and because your electricity provider will not accept USDC.

• Bank Salary, rent, direct debits, anything requiring recourse.
• App One to two months of spending, conversions, card payments.
• Hardware wallet Savings you intend to still own in three years.

First, your own mistake: a wrong-network transfer, a lost recovery factor, an approval granted to a malicious contract. This dwarfs everything else in frequency, which is why every page on this site keeps returning to it.

Second, a card programme interruption. Issuers suspend regions, migrate portfolios and re-issue cards. This has happened repeatedly across the industry and it usually arrives without warning. Keeping a spending float rather than a balance behind the card converts it into an inconvenience.

Third, a smart-contract failure in a yield product. Lower probability than the first two, higher severity, and largely uninsurable for retail users.

Fourth, company failure. In a genuinely self-custodial design this should be survivable — your assets remain on-chain — provided you can still sign transactions without the app. Verify that in advance, because it is the one scenario where preparation must precede the event.