Is Staking Halal? Yes, and Four Things That Are Not

Staking is halal when it means what it originally meant: committing your own coins to secure a proof-of-stake network and being paid for the validation work that results. We rate it permissible, our methodology says so, and it is one of the few places in crypto where a genuine non-riba income stream exists.

The trouble is that four other things now use the same word, and at least two of them are interest with a rebrand. If you have ever tapped a button labelled Stake and received an APY, you cannot answer this question until you know which of the five you were in.

What proof-of-stake actually pays you for

A proof-of-stake network needs someone to propose blocks, attest to other people's blocks, and stay online doing it correctly. Validators do that. Doing it requires hardware, an internet connection, monitoring, and a bonded amount of the network's token that can be destroyed if the validator misbehaves or goes offline in a correlated way.

The rewards come from two places: newly issued tokens the protocol mints for the purpose, and a share of the transaction fees users pay.

So you are being paid for a service performed, with real capital at risk while performing it. That maps cleanly onto ujrah, a fee for work, or onto ju'ala, a reward promised for an outcome. Both are established, both are permissible, and neither requires stretching anything.

The comparison people reach for is a bank deposit, and it fails in every way that matters. A bank promises you a percentage regardless of what happens next. A validator earns variable rewards for measurable work and can lose the principal for doing it badly. Those are different contracts wearing similar interfaces.

Why it is not a loan

The whole ruling turns on this, so it is worth being precise.

For staking to be riba it would have to be a qard, a loan of fungibles returned with an increase. Three things stop it being that.

Ownership never transfers. On Cosmos, Solana, Polkadot and Cardano, delegated tokens stay in your account. You have not handed them to anyone. On Ethereum they are bonded into a validator's balance on the beacon chain, still yours, still withdrawable by your credentials.

You keep the risk. In a loan the borrower guarantees your principal back in full. In staking, slashing can take part of your stake, and in a mass correlated failure on Ethereum the correlation penalty can take a great deal of it. A lender who can lose principal to the borrower's misconduct is not in a loan.

Nothing is guaranteed. The reward rate moves with the total amount staked, network activity and validator uptime. Ethereum's issuance-based yield has drifted between roughly 3% and 5% as the validator set grew. Nobody promised you a number.

Ownership retained, risk retained, return variable. That is a partnership or a service arrangement, not a debt.

Delegating without running a node

Most people will never run a validator, and delegation is what they actually do. You point your stake at someone else's validator, they run the infrastructure, they take a commission that is typically 5% to 10% of the rewards, and the rest comes to you.

This is still fine, and the reason is the same as above. You did not lend the validator anything, your coins did not move into their control, and if they get slashed your delegated stake gets slashed alongside theirs. You share the downside of their conduct, which is the clearest possible sign you are in a partnership rather than a loan.

Pick the validator on decentralisation and uptime rather than on who advertises the highest number, and avoid the ones at 0% commission, which is a marketing loss-leader that has to be recouped somewhere.

Liquid staking: real, with a caveat you should hear

Lido's stETH, Rocket Pool's rETH and Jito's jitoSOL let you stake and hold a tradeable receipt token instead of locked capital.

The underlying activity is unchanged. Validators are still validating, rewards are still ujrah, and I do not think the wrapper makes the income impermissible.

What the wrapper adds is a counterparty and a peg. Your rETH is a claim on a protocol, and a claim can trade below what it claims. In June 2022, during the Celsius and Three Arrows unwind, stETH traded down to roughly 0.94 ETH because everyone wanted out through an exit that did not exist yet. Nothing broke in the smart contracts. The market simply repriced a claim that could not be redeemed on demand.

That is a gharar consideration rather than a riba one, and gharar is a matter of degree. A large, audited, long-running liquid staking protocol whose redemption mechanism you can describe is acceptable. A new one offering a better rate on a chain you have not researched is a different risk, and gharar explained covers where the line sits.

Exchange staking, and the word "guaranteed"

Binance, Coinbase and Kraken all offer staking. They run the validators, they take a cut, you tap a button.

The mechanism can be genuine. The question is what your return is a share of.

If the exchange shows you an estimated rate that moves with actual network rewards, you are receiving a pass-through of a permissible activity, minus a service fee, and that is fine. If the exchange guarantees you a fixed APR regardless of what its validators earned, you are no longer sharing in an outcome. You have deposited a fungible asset and been promised a percentage, and the fact that some staking happens somewhere behind the scenes does not change the contract you personally entered.

That distinction is the whole audit, and it takes about a minute to run. Look at the rate. Is it a range that changes, or a number with a promise attached?

There is a second thing custody costs you, which is not fiqh but is real. Staked assets on an exchange are on that exchange's balance sheet, and self custody and possession explains why possession is not a preference in Islamic law.

The four things that are called staking and are not

Now the part that actually catches people.

Lending pools branded as staking. A protocol invites you to "stake" USDC and pays a yield. Your USDC was lent to a borrower who is paying interest. Nothing was validated, no service was performed, no risk was shared. This is lending with the wrong label on the button, and it is why AAVE fails our screen.

Emissions farms. "Stake our token, earn 400% APY." The yield is paid in newly printed units of the same token, funded by nothing, and the rate is high because the emission is the product. There is no validation, no fee revenue and no venture. Even setting the fiqh aside, the return is arithmetic rather than income: you own a larger share of something worth proportionally less.

Protocol safety modules. Staking a governance token into a backstop that can be raided to cover a shortfall, in exchange for a cut of that protocol's revenue. Here the question is what the revenue is. If the protocol earns interest, your reward is a share of interest, and no amount of intermediate contract steps changes where it came from.

Staking a coin that already failed the screen. The most common one and the simplest. Permissible income from an impermissible asset is still an impermissible holding. Every haram coin and what breaks it covers what to check before you get as far as the staking question. Start with the asset, then look at the yield.

The reservation I actually have

I have never seen this raised in an English-language discussion of staking, and I think it deserves an airing rather than a confident answer.

Where a staking reward is paid purely from new issuance, and the network's supply is inflating as a result, the staker's gain has a source. It is dilution borne by everyone holding the token who did not stake. The pie did not grow. Your slice grew and theirs shrank by the same amount.

You can defend it, and I lean toward the defence. The service is real, the security the network gets is real, and every holder benefits from a secured chain whether they staked or not, so the issuance is arguably payment out of a common benefit rather than a transfer taken from bystanders. Fee-funded rewards, like Ethereum's after the fee-burn mechanism, sidestep the objection entirely because the money is coming from users paying for a service.

But on a high-inflation chain where fee revenue is negligible and rewards are almost entirely printed, I am less comfortable than the standard position suggests I should be, and I have not seen a scholar engage with it at that level of specificity. I do not think it flips the verdict. I do think anyone claiming this question is fully settled has not looked at the tokenomics chain by chain.

Lock-ups, unbonding and zakat

Staked coins are not always available on demand, and the periods are published rather than uncertain.

Cosmos unbonds over 21 days. Polkadot takes 28. Ethereum has an exit queue whose length depends on how many validators are leaving at once. Solana releases at the end of the current epoch, which is a couple of days.

A known term is not gharar. Gharar is about not being able to describe what you are in, and "28 days, stated in the documentation" is a description. Plan around it and do not stake anything you might need next week.

Zakat is the part people get wrong. Locked and staked tokens are still your property and still zakatable, and the rewards accrued during the year form part of the calculation. Zakat on staked, locked and airdropped tokens has the working.

A five-minute check on your own positions

Open whatever you are staking through and answer four questions.

Is the coin itself screened. Is the reward variable or guaranteed. Where does the reward come from, validation and fees, or borrowers, or new issuance. And can you state the unbonding period without looking it up twice.

Variable rewards, from validation, on a screened asset, with a term you can describe. That combination is halal, and I would rather you get there by running the check than by taking my word for it.

If you want the chain-by-chain view of which networks pass and what breaks the ones that do not, Layer 1 blockchains ranked by Shariah compliance does exactly that, and staking design is one of the things it weighs.