The collateral is new. The card contract is old, and the contract decides what happens when a payment is late.
Almost every stablecoin card shipped so far spends a balance. A few now borrow against one. In June, Coinbase and Cardless added a version of the Coinbase One Card that opens a credit line against USDC held on Coinbase. In August, Western Union’s Stablecard arrived as a Visa secured credit card backed by USDPT instead of a debit card on the wallet balance.
That is a design choice with consequences on both sides of the transaction. For the cardholder, a credit line creates a debt that the collateral backs, which is not the same as spending money you already have. For the issuer, it changes the economics of every swipe. Both are worth understanding before the next program announces the same structure.
What a secured line actually is
Start with the mechanics, because the marketing skips them.
On the Coinbase version, an applicant posts a USDC security deposit, and Coinbase holds it on its platform. The deposit sets the credit limit, starting at $200 and running up to $5,000. While it secures the line, that USDC cannot be traded, withdrawn, lent, or transferred. If the account defaults or closes with a balance, the deposit is applied to what is owed, and anything left goes back to the customer.
Read the last part again, because it is the part people assume away. The deposit is applied to the balance. It does not cap the obligation. A cardholder who runs up more than the collateral covers still owes the difference, the same as on any other credit card.
The card is backed by stablecoins but issued in the ordinary way: First Electronic Bank issues it, Cardless runs the program, and it clears on American Express. Rewards run from 2% to 4% in bitcoin depending on assets held at Coinbase, with the higher tiers capped at $10,000 of monthly purchases, and the card comes with a Coinbase One subscription at $49.99 a year.
Why an issuer would lend instead of debit
The first answer is interchange. Credit pays more than debit, and it is not close.
The Merchants Payments Coalition puts the average combined credit card interchange rate at 2.36% in 2025. On the debit side, the Federal Reserve’s data shows regulated debit interchange averaging well under half a percent of transaction value at covered banks, with exempt issuers averaging about 1.21%. A program that can route spending onto credit rails earns more per dollar, and the gap funds the rewards that get advertised on the front of the card.
The second answer is the balance sheet. A credit card creates a revolving asset. Even at low utilization, a portfolio of lines produces interest income that a debit program never generates, and it gives the issuer a reason to keep underwriting the customer over time.
There is a third answer specific to token funding. On a debit-style stablecoin card, the program has to check a balance, price a conversion, and approve inside the authorization window on every purchase. On a secured line, the collateral is already posted, and the limit is already set, so the spending decision at the terminal is an ordinary credit authorization and the token movement happens on repayment instead. That is a quieter piece of engineering with fewer moving parts at the moment a customer is standing at a till.
Who holds the collateral, and what that means
Three products, three answers, and the answer decides the risk.
Coinbase holds the USDC on its own platform, which means the customer is relying on Coinbase’s custody and the terms attached to it. Western Union’s secured card is backed by USDPT sitting in the Western Union wallet, issued by Anchorage Digital Bank, so the collateral and the spending balance live inside the same app. Ether.fi Cash takes a different path again: users borrow against assets they hold, with no repayment schedule and cashback by membership tier, and the company states plainly that it is not a regulated financial institution and that supplied assets and borrowings are not insured.
None of those arrangements is a bank deposit. A traditional secured card puts the deposit in an insured account at the issuing bank. A stablecoin-secured card puts it with a platform, a wallet, or a smart contract, and the protections that follow are the ones written into the terms, not the ones a bank failure regime provides. Anyone evaluating these products should read the card enabler layer underneath them as part of the credit decision, not as a technical footnote.
The rules did not change
A secured card is still a credit card, and United States credit card rules apply to it in full.
Regulation Z requires a card issuer to consider whether the consumer can make the minimum payments, based on income or assets and existing obligations, before opening an account or raising a limit. The carve-out in that rule is for home-secured plans, not for deposit-secured cards. Collateral reduces the issuer’s loss given default; it does not remove the underwriting obligation.
The rest of the credit card framework travels with it as well: periodic statements, billing rights, fee and rate disclosures, and the standard protections that come with a credit product instead of a debit one. Programs that market a collateralized card as a way around the paperwork are describing something the law does not offer.
Read more: Stablecoin corporate cards reach companies outside crypto • When a stablecoin card program fails, who gets the money back
How a lender actually takes the collateral
For issuing banks and card product teams, the interesting question is not the marketing. It is whether a security interest in a token balance is worth anything in a dispute.
That got easier this year. The 2022 amendments to the Uniform Commercial Code created a category called controllable electronic records and made control the way a lender perfects an interest in them, in place of a filing. Control means the ability to enjoy all the benefits of the asset, to stop others from enjoying them, and to pass that ability on. A security interest perfected by control outranks one perfected only by filing, even an earlier filing. New York enacted the amendments in May, effective in June, and more than 30 states have now adopted them.
That framework maps neatly onto the products here. When a platform holds the collateral, and the terms say it cannot be traded, withdrawn, lent, or transferred, the lender has control in the sense the statute describes. It is the legal reason a deposit made of tokens can support a credit line at all, and it is why these programs are launching in the United States now instead of two years ago.
What happens when a payment is missed
Here is the sequence worth having in mind before applying.
A missed payment triggers the same treatment as any card: late fees, a higher rate where the terms allow it, and eventually default and collections. At default, the collateral is applied to the outstanding balance. If the collateral covers it, the matter ends and the remainder is returned. If it does not, the cardholder owes the rest, and the debt behaves like any other unsecured obligation from that point.
Stablecoin collateral removes one risk that crypto-backed lending usually carries. A dollar token does not fall 30% overnight, so there is no margin call driven by price and no forced liquidation at the bottom of a chart. That is the strongest argument for using a stablecoin instead of bitcoin as the collateral asset, and it is the reason these products can price like secured cards instead of like margin loans.
What it does not remove is the counterparty question. The collateral is only as good as the entity holding it and the redemption promise behind the token, which is a different failure mode from a price drop and one that surfaces at the worst possible moment.
What the pages do not tell you
For a credit product, the disclosure set on offer is thin.
The rate is the obvious gap. None of the public pages for these cards lead with an APR range, and on a secured card aimed at applicants with thin files, the rate is the product. The fee picture is clearer: Coinbase’s card carries no foreign transaction fee and sits behind a $49.99 subscription, and the USDC posted as a deposit keeps earning at the standard rewards rate while it is locked, which is a genuine improvement on a bank deposit that pays nothing.
The other gaps are operational. How long does it take to release collateral after an account closes in good standing? Can a cardholder raise the limit by adding collateral, and how fast? What happens to the line if the platform changes its terms for the token? These are ordinary questions for a secured card, and they are the ones a product team should answer before launch rather than in a support ticket.
Does it build a credit file
This is the question the marketing pages are quietest about.
The point of a traditional secured card is that the deposit buys a credit history. Payments get reported, the file thickens, and after a year or two the customer graduates to an unsecured product and gets the deposit back. If a stablecoin-secured card does not report to the credit bureaus, it is a spending product with a lock on the collateral rather than a path to credit.
Coinbase advertises an approval check with no impact on the applicant’s credit score, and describes the deposit as a route for people who would otherwise be declined. Cardless framed the audience clearly when the card launched: people who cannot get a card through traditional channels but hold digital assets, including those “just beginning their journeys and accumulating wealth.” Whether on-time payments on that card show up on a credit report is the detail that decides whether the product is a bridge or a cul-de-sac, and it belongs in the disclosures rather than in a support article.
The DeFi version is not the same product
Ether.fi Cash and similar designs get grouped with these cards and should not be.
Borrowing against a portfolio with no repayment schedule is a collateralized loan with a card attached. There is no statement cycle, no billing rights framework, and no regulated issuer standing behind the credit. The upside is flexibility and the absence of a credit check. The downside is that every protection a cardholder assumes from using a card at a merchant depends on which entity is in the chain, and on collateral that can be liquidated when it is volatile.
Both models can be reasonable. They are not interchangeable, and a comparison table that lists them side by side under “credit cards” is doing the reader a disservice.
Who these cards are actually for
Three groups have a real case.
People with thin or no credit files who hold digital assets, which is the audience Cardless named. People who do not want to sell tokens to fund spending, since a secured line lets the collateral sit while the card handles purchases. And customers of a remittance or wallet brand who already hold a dollar balance in that app and would rather spend on credit terms than draw the balance down.
Everyone else should do the arithmetic. Locking $1,000 of USDC to unlock a $1,000 limit is expensive capital if an ordinary card is available, and the rewards have to be weighed against the subscription and the opportunity cost of collateral that cannot be moved.
What would tell us this works
Four things to watch from here.
Whether issuers publish approval rates and the pricing that sits behind these lines, since a secured card with a high rate and a locked deposit is a different offer from the one implied by a rewards headline. Whether payment history gets reported to the bureaus. Whether Rain’s Mastercard membership, which explicitly lets it offer credit and prepaid cards, produces more collateralized credit programs from other brands. And whether collateral that keeps earning while it sits, as Coinbase’s does, stays available under the rules being written for stablecoin rewards, which is a separate fight with its own timetable.
The structure itself is not exotic. Secured cards have existed for decades, and the only new part here is what the deposit is made of. That is precisely why the questions to ask are the old ones: who holds it, what unlocks it, and what happens to the rest of the debt when the collateral runs out.