Are Crypto Savings Accounts Halal? Centralized Yield and Hidden Riba

The pitch is always some version of the same sentence: park your BTC, ETH or USDC with us and collect 4% to 9% a year, paid weekly, no lockup. It sounds like a savings account because it is designed to sound like a savings account. And that resemblance is exactly the problem when you run it through a Shariah lens, because the thing a savings account most resembles in fiqh is a qard, an interest-free loan, and a qard that pays you extra is the textbook shape of riba.

So: are crypto savings accounts halal? Centralized yield products are one of the rare crypto questions where the scholarly answer is unusually tight rather than unusually contested. The disagreement in Islamic finance about crypto is mostly about the asset (is a token mal, is it currency, is Bitcoin permissible to hold at all). Centralized lending yield is about the contract, and the contract analysis is much older than crypto.

What a crypto savings account actually does with your coins

Strip the branding and a centralized earn product does three things in sequence. You transfer coins to the platform's wallet. Legal title to those coins usually moves too, which matters enormously and we will come back to it. The platform then deploys them, and pays you a slice of whatever it makes.

That deployment is where the APY number comes from, and it is worth knowing which engine you are actually attached to, because a single advertised rate can be sourced from very different activities.

The three yield engines behind one APY number

Institutional lending. The platform lends your coins, usually overcollateralized, to trading firms, market makers and hedge funds who want leverage or need to borrow to short. The borrower pays interest. You get part of it. This was the core Celsius and BlockFi model and it is still the core of most CeFi earn desks.

Reserve interest pass-through. For stablecoins specifically, the money frequently comes from short-term US Treasuries. Circle holds USDC reserves in cash and short-dated Treasuries and does not itself pass that interest to holders, so a platform paying you "USDC rewards" is often funding it from the T-bill yield on reserves it controls, or from its own balance sheet as a customer acquisition cost.

Routed DeFi. Increasingly the front end is centralized while the engine is a lending protocol underneath. Coinbase's USDC lending vaults, for example, route deposits into Morpho markets where borrowers post collateral and pay a floating rate. The user experience looks like a savings account, and the underlying activity is still collateralized interest-bearing lending.

Notice what is absent from all three. None of them is a profit-and-loss partnership. None involves you taking equity risk in a venture. In each case you hand over a fungible asset, someone else uses it, and you are promised the same amount back plus a defined increment.

Celsius and BlockFi: read what the paperwork actually said

The 2022 collapses did Muslim investors an accidental favor by forcing the legal structure into open court, where the contract stopped being marketing copy and became evidence.

Celsius filed for Chapter 11 in July 2022. Its Earn program covered roughly 600,000 accounts holding about $4.2 billion at the petition date. The Terms of Use transferred to Celsius "all right and title" to deposited crypto, including ownership rights and the right to pledge, repledge, hypothecate, rehypothecate, sell, lend or otherwise transfer the assets. In January 2023, Judge Martin Glenn of the Southern District of New York bankruptcy court held that those terms meant what they said: the coins belonged to the estate, and Earn depositors stood in line as unsecured creditors.

BlockFi's Interest Account had already drawn a separate kind of scrutiny. In February 2022 BlockFi settled with the SEC and a coalition of state regulators for $100 million over the BIA product, with regulators treating it as an unregistered securities offering rather than a deposit. BlockFi filed for bankruptcy that November.

For a Shariah analysis, the courtroom finding is the useful part. If title passes to the platform, if the platform can lend and rehypothecate freely, and if you hold a personal claim for the return of an equivalent amount plus a rate, then you are a lender. That is the fiqh definition of qard whatever the product page calls it. It also explains why the "your funds are safe, it is just a savings account" framing was legally hollow: depositors had no proprietary claim to specific coins.

The riba analysis, and what is doctrine versus inference

Here the distinction matters, because these two layers carry different weight.

Doctrine. The prohibition of riba al-nasiah, the increment stipulated on a deferred obligation, is explicit in the Quran (2:275 to 2:279), which permits trade and forbids riba, and closes with the principle that the lender is entitled to his principal sum, neither wronging nor being wronged. The classical fiqh maxim built on that, cited across the schools, is that every loan which draws a benefit to the lender is riba. AAOIFI's Shariah Standard No. 19 on Qard codifies the same rule for modern institutions: a loan must be repaid at par, and any stipulated benefit to the lender, in cash or kind, is prohibited. This part is not a matter of opinion among Sunni or Shia jurists.

Inference. The reasoned judgment is the classification step, deciding that a Celsius-style Earn account is a qard and not something else. That inference rests on the transfer of title, the fungibility of the asset, the guarantee of the principal, and the pre-agreed or platform-set rate. Where any of those four features is genuinely absent, the analysis has to be redone from scratch rather than assumed. Some products are structured as custody plus a discretionary bonus, and a few Islamic-window products in the Gulf attempt a wakala or mudarabah wrapper where returns genuinely float with realized performance and loss is possible. Those need to be judged on their own documents.

There is also a second defect in most CeFi earn products that is independent of riba. The borrower on the other side is frequently a trading firm using your coins for leveraged directional bets or shorting. Even if you somehow solved the interest problem, income traced to that activity raises gharar and impermissible-use concerns.

Where scholars land, and where they split

The near-consensus among contemporary Shariah boards reviewing digital-asset products is that fixed-rate custodial earn accounts fail. Shariyah Review Bureau, one of the more active advisories on crypto structures, draws the line between reward for genuine risk-bearing work and a guaranteed return on a transferred balance, and centralized lending yield sits on the wrong side of it. Islamic fintech screens generally treat exchange "earn" and "flexible savings" products as riba-bearing regardless of the token involved.

The real split in crypto fiqh sits elsewhere, and it is worth not confusing the two debates. Mufti Taqi Usmani and the Darul Uloom Karachi position holds that cryptocurrencies are not valid mal or thaman at all, which makes the yield question moot because the underlying asset already fails. Malaysia's Securities Commission Shariah Advisory Council took the permissive route, recognizing digital assets as recognized property (mal) and therefore tradeable, subject to the usual conditions. But even under the permissive Malaysian line, a lending contract with a stipulated increment is still riba. The permissive camp opens the door to holding and trading tokens, and it does not open a door to fixed-yield lending.

The genuinely contested territory is native protocol staking. Delegating ETH to a validator, bearing slashing risk, and receiving a floating protocol reward tied to real network activity is treated by several Shariah boards as compensation for a service and for assumed risk, which is a different structure from a loan. Others remain uneasy because the staked amount is effectively guaranteed in ordinary conditions. Either way, the label "staking" on an exchange is unreliable, because many platforms use that word for what is functionally a lending desk.

Where the other faith frameworks land

Christian BRI screening focuses on the six categories (abortion, pornography, anti-family entertainment, alcohol, gambling, tobacco, plus human rights and lifestyle concerns depending on the implementation) and does not treat interest as intrinsically prohibited, so a USDC earn account raises no BRI category flag on its own. The USCCB socially responsible guidelines likewise do not exclude interest income, though the Catholic tradition's usury teaching bites on exploitative rates rather than on interest as such.

Jewish halakha is the closest analogue. Ribbis rules apply between Jews, and Bais HaVaad and other contemporary poskim address exactly this two-tier structure, with the heter iska device converting what looks like an interest-bearing loan into a profit-sharing joint venture on paper. A crypto earn product between a Jewish depositor and a Jewish-owned platform would need that treatment; the same product from a non-Jewish issuer generally would not trigger the prohibition.

The LDS framing is less about the contract and more about the behavior. Dallin H. Oaks warned in 1971 against speculation and get-rich-quick reasoning, and a double-digit yield on a volatile token, offered by an unregulated intermediary, is precisely the sort of thing that warning was aimed at.

What to actually do with this

Read the terms of use for the title clause before anything else. Search the document for "rehypothecate" and for "right, title and interest." If ownership transfers, you are a lender and an unsecured creditor, and both the fiqh and the bankruptcy risk follow from that single clause.

Ask where the yield comes from and refuse products that will not say. A platform that cannot describe its yield engine in one sentence is one you cannot screen.

Distinguish custody from earning. Holding a screened token in custody, or in self-custody, carries none of this analysis. The riba problem is created entirely by opting into the earn feature, and opting out is usually one toggle.

Treat "staking" claims skeptically. Check whether the reward floats with protocol issuance and whether you carry slashing exposure. If the rate is fixed and the principal is guaranteed, the word on the button does not change the contract underneath.

If you want dollar-denominated stability, holding a fully reserved stablecoin without the yield product avoids the contract problem, though it leaves open the separate scholarly discussion about fiat-backed stablecoins generally.

How FaithScreener handles it

Our crypto screening covers more than 3,300 tokens, and the screen runs on the asset and its protocol rather than on the wrapper a platform sells around it. A token can pass the asset-level screen while the earn product built on top of it fails, which is why the yield mechanism is assessed separately from the coin. Staking treatment follows the reward structure: protocol-native, risk-bearing, floating rewards are handled differently from custodial fixed-rate programs. If you want to see how the same token reads under Islamic, BRI, USCCB, halakhic and LDS lenses side by side, the framework comparison shows where they diverge, and the screening methodology documents the thresholds and the sources behind each judgment.

The Bottom Line

A centralized crypto savings account paying a fixed or platform-set rate on coins whose title you transferred is a loan with a stipulated increment, and that structure is riba under every major Sunni and Shia school regardless of where a scholar stands on whether Bitcoin is valid property in the first place. The one thing to carry away is that the token and the yield product are two separate rulings, and Celsius proved it in court: the terms of use, not the marketing, tell you which contract you signed.

This is educational research rather than a fatwa or personalized investment advice, so confirm your own situation with a qualified scholar or advisor before acting on it.