The discount was never a backing failure. It was a liquidity event in an asset that could not be redeemed at the time.
Nadia OkoroPolicy & Consumer Editor · Regulation, tax, stablecoins and the products retail users hold
A liquid staking token represents staked ETH plus rewards, redeemable for ETH through the network's exit queue. Before withdrawals were enabled in April 2023, that redemption did not exist, so the token's price depended entirely on secondary market liquidity — and in June 2022 forced selling by leveraged holders pushed stETH several per cent below ETH. Withdrawals now provide an arbitrage anchor, but discounts can still open when exit queues lengthen or leveraged positions unwind.
In June 2022, stETH traded roughly 5% below ETH, and a large part of the market read it as a solvency question about Lido. It was not. Every stETH was backed by staked ETH the entire time. What did not exist was a way to get that ETH out.
A liquid staking token is a claim on staked ETH plus accrued rewards. Its exchange rate against ETH is set by the protocol from the underlying balance — it grows as rewards accrue, and it is not a peg in the stablecoin sense.
The market price is a separate thing. It is whatever someone will pay right now, and it converges to the exchange rate only if there is a reliable way to convert one into the other. That mechanism is redemption, and its availability is the whole story.
Ethereum's transition to proof of stake had happened conceptually but withdrawals were not yet live. Staked ETH was locked with no exit date, so stETH holders wanting ETH had exactly one option: sell in the secondary market.
At the same time, a large share of stETH was held in leveraged loops — borrow ETH against stETH, buy more stETH, repeat. When prices fell, those positions faced liquidation, and unwinding required selling stETH into a market where the only buyers were people willing to hold an illiquid claim indefinitely. Celsius held a very large stETH position it could not exit; Three Arrows and others were unwinding simultaneously.
The result was a classic forced-seller discount: a solvent asset trading below its intrinsic value because the marginal seller had no choice and the marginal buyer had no deadline.
The Shapella upgrade in April 2023 enabled staked ETH withdrawals. This changed the structure permanently, because an arbitrage now exists: if stETH trades below ETH, someone can buy stETH, redeem it through the exit queue, and receive ETH worth more than they paid.
That arbitrage is not instant — it takes as long as the exit queue takes — so the discount does not close immediately. It becomes a function of the queue: a two-day queue supports a discount of a few basis points, a three-week queue supports considerably more. The mechanics of that queue are covered in withdrawal queues.
Three scenarios, all of which have appeared since.
**Queue congestion.** Mass exits lengthen the wait, widening the discount the arbitrage can tolerate. This is self-reinforcing: a discount encourages more people to sell rather than queue, and heavy exit demand lengthens the queue further.
**Leverage unwinds.** The loop trade never went away. It is now a standard strategy on major lending markets, and it still means that a fall in ETH forces stETH selling into a market that widens exactly when it is needed most.
**Smaller LSTs.** The arbitrage requires liquidity to execute and a redemption path that works. A minor liquid staking token with thin pools can trade at a persistent discount because nobody is doing the trade, and one with an operator-controlled redemption process may not have the path at all.
It is a liquidity signal, not a backing signal. To check backing, look at the protocol's reported total staked ETH against the token supply and at the exchange rate's continued accrual — both are on-chain and verifiable.
The genuine solvency risks in an LST are different: validator slashing reducing the underlying balance, node operator concentration, the withdrawal credential setup, and the governance of the contracts. Those risks would show up as a change in the exchange rate, not as a market discount, and they are what the liquid staking ratings weight most heavily.
First, the LST discount is priced by the exit path, so know what yours is: redemption available on demand, redemption through a queue of some length, or no redemption at all.
Second, do not lever a position whose collateral is the asset you would need to sell in a stress event. The loop trade is profitable most of the time and mathematically designed to fail at the moment everyone needs to exit.
Third, protocols pricing LST collateral at the underlying exchange rate rather than the market price protect borrowers from exactly this scenario — and that oracle choice, listed in the protocol documentation, is worth checking before you borrow against one.
Fourth, if you are holding an LST for staking yield rather than for liquidity, a discount is an opportunity rather than an event. The redemption still pays the full underlying amount whenever the queue clears.
How liquid staking tokens let holders earn network rewards while keeping their capital free to lend, trade, or post as collateral.
A tokenized bond is only as good as the custodian holding it, the legal entity issuing it, and your ability to actually redeem it. A framework for the risks that live off-chain in every RWA tokenization deal.
The fee comparison everyone quotes is the wrong one. What separates these two is what you give up, and the answer changes with how much you stake.