A discount to ETH is not necessarily missing ETH. Liquid staking exposes investors to three different numbers: secondary-market prices, protocol redemption claims and the amount of stake actually available after penalties.
What Does “Backing” Mean for a Liquid Staking Token?
Liquid staking tokens are often displayed next to ETH with a reassuring reference price. That interface conceals a three-way distinction. There is the number of ETH validators and protocol contracts ultimately control; there is the amount of ETH contract terms and accounting assign to each receipt token; and there is the price another trader will pay for the token right now. All three can diverge even when no attacker has stolen a single ether.
The first separation is structural. Ethereum itself recognizes validators and their withdrawal credentials, not the investor’s stETH, rETH or cbETH balance. An LST represents a relationship with a staking pool, custodian or smart-contract system. Ethereum.org identifies smart-contract, slashing, market-liquidity and governance risks as additions to the underlying economics of ETH staking. When the receipt trades freely, its price can fluctuate independently of the underlying withdrawal claim. [1]
A second separation is accounting. Lido’s stETH rebases: holders’ nominal balances usually rise as earnings accrue. Wrapped stETH (wstETH) keeps the token quantity constant and increases the exchange rate per token. Coinbase’s cbETH also uses an exchange-rate model. Comparing an exchange-rate LST at, say, 1.15 ETH per token with a rebasing token at roughly 1 ETH is not evidence that one is overvalued or unbacked. The relevant benchmark is each contract’s ETH redemption or conversion rate, not a universal “one token equals one ETH” label. [2][3]
Market Depeg: A Liquidity Problem, Not Necessarily a Solvency Problem
Suppose a holder owns 100 stETH and the protocol’s expected claim, before processing conditions, is 100 ETH. In a stressed trading venue the best executable secondary-market quote is 0.94 ETH per stETH. Selling immediately returns approximately 94 ETH before spread, slippage and gas, a haircut of 6 ETH against the assumed 100 ETH claim. That difference is a liquidity discount. It does not by itself prove that the protocol possesses only 94 ETH.
The investor’s alternative is to request withdrawal through the protocol. Lido’s withdrawal-queue contract issues an NFT representing a claim and finalizes requests as ETH becomes available from its buffer and validator withdrawals. Pending request holders no longer earn staking rewards while queued, and significant protocol losses can change the finalized amount. Thus “redeem at par” is conditional on the protocol’s actual rules and withdrawal state; it is not equivalent to an instant guaranteed bank redemption. [4]
The spread can persist even with arbitrageurs. An arbitrageur purchasing discounted stETH needs capital, a valid withdrawal route, acceptable queue duration, predictable accounting and confidence in future liquidity. During an exodus, multiple actors compete for the same funding and execution capacity. Lido’s 2026 public risk disclosure explicitly warns that secondary-market slippage and protocol-redemption queues are different risk exposures. [5]
FIGURE 1 | Protocol Coverage and Market Quotes Are Different. Source: Dave Finances calculations / cited primary disclosures; see Methodology.
Price, Liability and Actual Asset Loss Are Separate Ledgers
An analytical balance sheet makes the distinctions clearer. In Scenario A, a protocol accounts for 1,000,000 ETH supporting 1,000,000 stETH-equivalent claims and the market quote is 0.96 ETH. The protocol may be fully backed in ETH terms even though the market value of the circulating receipt is 960,000 ETH. The 40,000-ETH difference is a market discount, not a verified balance-sheet hole.
In Scenario B, the same starting ledger suffers 20,000 ETH in validator penalties, leaving 980,000 ETH for 1,000,000 equivalent claims before recovery or other arrangements. The economically relevant backing rate falls to 0.98 ETH per claim. A secondary quote of 0.94 would now combine an approximately 2% claim impairment with an additional discount against the revised recovery value. Investors should not count the whole six-cent gap as a pure liquidity opportunity.
In Scenario C, no ETH is stolen or slashed, but withdrawals are delayed, a pricing oracle stops updating, or collateral liquidation overwhelms DEX liquidity. The token may trade down without any decline in recorded underlying ETH. It is the same distinction our stablecoin depeg and recovery analysis makes between the price paid by a secondary-market buyer and what a holder can legally or operationally redeem—but staking receipts carry validator and protocol risks rather than a fiat reserve promise.
What Redemption Waiting Costs in Practice
A 6% discount may look like free money if the investor can simply buy stETH for 0.94 ETH and ultimately receive roughly 1 ETH. But the return is earned by committing capital to a withdrawal process with time and loss uncertainty. Ignoring all costs, converting 0.94 ETH into 1 ETH after 90 days is a 6.38% holding-period gain, equivalent to approximately 28.5% per year if the same outcome could be repeated and compounded. At 30 days the same hypothetical trade annualizes to more than 110%. Those large annualized numbers reveal how dramatically the time assumption matters; they are not yield forecasts.
The real comparison is the investor’s alternative return and the risks during the queue. Lido notes that users do not earn ongoing stETH rewards on requests held in its queue. A rational arbitrageur also deducts execution fees, gas, borrowing costs if leveraged, expected penalties and the probability that redemption is delayed or suspended. If the expected net redemption value is only 0.98 ETH rather than 1, the 0.94 entry offers a 4.26% gross gain, not 6.38%.
There is a second constraint: not every token has the same redemption contract. Coinbase Exchange documentation notes that cbETH exchange rates and conversion rates can differ and that redemption can take up to 12 days through its described workflow. That information belongs in a comparison, not flattened into a universal “ETH withdrawals take X days” rule. [6]
FIGURE 2 | The Cost of Waiting Depends on the Queue. Source: Dave Finances calculations / cited primary disclosures; see Methodology.
How DeFi Collateral Turns a Discount Into a Liquidation
Collateralized lending can convert a temporary market discount into a permanent realized loss. Assume a leveraged investor holds 100 LST units, each initially valued at 1 ETH, ETH trades at a fixed illustrative $2,000, and the investor has borrowed $120,000. Initial collateral value is $200,000 and loan-to-value is 60%. If the lending protocol liquidates at a 65% effective LTV threshold, the trigger occurs when collateral valuation falls below about $184,615, or 0.9231 ETH per LST under these assumptions.
At an LST market price of 0.94 ETH, collateral value is $188,000 and effective LTV rises to 63.83%, uncomfortably close to the line despite no assumed loss of ETH in the staking protocol. At 0.92 ETH, effective LTV reaches 65.22%; liquidation may commence depending on the lending system’s own oracle, threshold, health-factor formula and liquidation logic. No real protocol is being claimed to use the hypothetical 65% level.
Oracle design is decisive. A lending venue that prices the token at protocol exchange rate can show a different position from a venue that uses a market-price oracle. Using a backing-based feed can smooth temporary dislocations but may also delay the recognition of true impairment; using a spot feed is more sensitive to liquidity shocks. Lido’s integration documentation discusses exchange-rate feeds for wstETH and how application developers combine them with ETH prices. Retail investors should read the actual collateral and liquidation rules rather than assuming a generic “1:1 backed” token receives identical treatment everywhere. [2]
What Does Each LST “Depeg” Actually Mean?
| Observed change | Backing conclusion | Retail exposure |
| DEX price drops; redemption stable | No verified asset shortage | Immediate exit haircut and margin risk |
| ETH-per-claim conversion falls | Underlying asset impairment possible | Lower eventual ETH redemption |
| Redemption queue expands | Not by itself evidence of lost ETH | Capital tied up without instant exit |
| Collateral oracle marks spot lower | Does not prove backing shortfall | Forced liquidation before redemption |
How Backing Can Be Impaired Without a Theft
Validator slashing is the clearest example: Ethereum removes part of a validator’s staked ETH for certain slashable misconduct. No thief needs to receive those coins. The pool’s economic assets simply decline. Coinbase’s cbETH white paper expressly says that slashing can reduce the ETH behind cbETH and cause its conversion rate to decrease; Lido’s accounting documents similarly describe negative rebases as a possible consequence of penalties, even though its integration guide says they had not occurred at the time of that document. [7][2]
Theft is not the only route through software either. Contract bugs, accounting mismatches and governance failures can prevent claims from being fulfilled, while operational withdrawal delays produce market discounts even if the asset ledger remains sound. A malicious oracle report or stale oracle may misstate a conversion rate temporarily, causing secondary-market trades or liquidations before a corrected ledger becomes visible. Such events are not interchangeable: a false price can be reversed or arbitraged away, while a contract shortfall may be permanent.
The right diligence test therefore has four stages: reconstruct the validator/underlying asset balance; reconcile receipt claims and exchange-rate methodology; validate available withdrawals and queue rules; and price the token’s current sellable liquidity. Traditional proof-of-reserves alone misses those settlement and contingent-loss mechanics. The limitations resemble those documented in our analysis of exchange proof-of-reserves reports, though staking ledgers have distinct on-chain observability.
FIGURE 3 | A Collateral Discount Can Trigger Forced Liquidation. Source: Dave Finances calculations / cited primary disclosures; see Methodology.
A Retail Checklist: What to Measure Before Buying the Discount
First identify whether the token rebases or appreciates through its ETH-per-token conversion rate. A nominal quote above 1 ETH can be normal for an accumulating receipt. Next consult the provider’s published redemption route, whether an individual holder is eligible, any queue or cooldown, the amount actually claimable under the withdrawal contract and the evidence for reported assets. The existence of a DEX pool is not a substitute for redeemability.
For execution, size the trade against *executable* liquidity, not only a chart’s midpoint. A $20,000 order moving price by 0.5% and a $2 million order moving it by 5% represent very different routes even when the screenshot shows the same price. Compare net ETH from selling now with net ETH from redemption after estimated waiting, fees and potential slashing. Review your lending protocol’s oracle and liquidation policy if the receipt is pledged as collateral; the intersection of redemption delay and forced selling is the greatest risk to a leveraged holder.
Finally, assess hidden links. Bridged LSTs create another contract and bridge dependency; using an LST as collateral to borrow a stablecoin can add a liquidation chain on top of staking risk. As our analysis of sandwich attacks on swap prices shows, the advertised exchange rate is not identical to an investor’s realized swap proceeds. These layers can magnify a discount without changing the number of ETH controlled by the original validators.
How to Stress-Test a Backing Claim Using Two Ratios
Analysts should publish both the protocol claim ratio and the market realization ratio. Define backing coverage as redeemable ETH assets divided by ETH-denominated claims. Define market realization as executable net ETH sale proceeds divided by the claim amount. For a hypothetical LST with 500,000 ETH of attributable assets, 500,000 ETH of claims and secondary execution at 0.95 ETH for each unit of claim, coverage is 100% and immediate realization is 95%. If a confirmed 10,000-ETH loss lowers the asset ledger to 490,000 ETH, coverage becomes 98% even if the market later quotes 0.99 ETH. Price recovery does not repair the underlying deficit.
For the individual holder this produces four possible states. Full backing and a market discount is mainly a liquidity and patience problem; full backing and an above-par quote is a convenience premium that might disappear; impaired backing plus a discount combines solvency and market risk; impaired backing with a market premium means the buyer is paying more than the ultimate underlying claim might justify. The framework also highlights that quoted order-book prices are not necessarily executable for the desired size, so the denominator and numerator must be measured on compatible units.
Apply the test to 10,000 units in a hypothetical pool. If redemption rights equal one ETH each and the best realizable market quote net of all trading costs is 0.94 ETH, the owner can take 9,400 ETH immediately or submit a claim that might ultimately yield 10,000 ETH. Now assume a 1.5% haircut on the redemption due to verified protocol losses: projected protocol proceeds become 9,850 ETH. The incremental benefit of waiting rather than selling is only 450 ETH before waiting costs and residual risk. The initial 600-ETH headline discount exaggerated the economically achievable spread by one quarter.
These ratios demand honest reporting about estimates. Validator balances, pending rewards, withdrawal receivables and liabilities may update at different cadences. Auditors should disclose the data timestamp and whether counted assets are withdrawable or simply attributable. The ratio is a stress-testing device, not a ready-made credit rating or guarantee of priority in insolvency.
Why the Same Discount Affects Three Investors Differently
An unleveraged investor who can hold a receipt until redemption faces liquidity and future claim risk. A leveraged investor who needs the receipt to maintain collateral value faces liquidation risk on a shorter clock. A trader providing liquidity to an LST/ETH pool faces inventory risk: as the token cheapens, an automated market maker can leave that provider owning more discounted LST and less immediately usable ETH. All three may hold the same economic asset but realize very different outcomes from a temporary market shock.
Consider a 100-unit LST position at 1 ETH of protocol claim each. Investor A can wait through a queue and might ultimately recover about 100 ETH, assuming no future loss. Investor B has borrowed against those tokens and faces a collateral threshold near a price of 0.923 under the illustration used earlier. A secondary quote of 0.90 can force a sale before redemptions complete, making the mark-to-market discount an irreversible loss. Investor C supplied equal-value ETH and LST liquidity and may suffer adverse inventory rebalancing as traders sell the LST into the pool. Neither the liquidity provider nor the borrower can eliminate this risk by citing the original 1:1 claim.
The distinction becomes still sharper across chains. A wrapped staked token may sit behind a bridge, and the local market can price bridge security, local depth and redemption-friction risk in addition to staking risk. The receipt may remain sound on Ethereum while the representation used as collateral on a different chain loses liquidity or market confidence. A cross-chain discount therefore should not be automatically attributed to validator losses.
This is why an analysis based only on “ETH was never stolen” is incomplete. Losses can arise through forced execution, waiting, financing and collateral liquidation without any permanent underlying shortage. Conversely, a real backing shortfall can exist even when a shallow market prints a price close to par. The decisive evidence is the combination of contract claims, recoverability, market depth and the holder’s constraints.
What Would Prove the Market-Only Thesis Wrong?
A full redemption backlog is inconvenient but does not establish insolvency. Evidence against a “temporary liquidity discount” thesis would include a verified decline in ETH assets per redeemable claim, substantial realized validator penalties, withdrawals finalized materially below the expected accounting rate, unreconciled shares, or governance/contract action that permanently impairs claims. By contrast, shrinking queues, successful redemptions at the advertised claim rate and deepening arbitrage liquidity would support the case for a market-driven discount.
The bottom line is that an LST can trade below ETH without any ETH theft because the holder is selling a delayed, conditional claim, not instantaneous transferable ETH. It can also experience real backing erosion without theft through slashing or contract accounting losses. Investors need a two-axis analysis—market discount and underlying claim impairment—before deciding whether a low exchange price is an opportunity or a warning.
Methodology
Figures use ETH-denominated examples, not current market quotations. “Backing coverage” is attributed ETH assets divided by ETH claims; the “market realization ratio” is executable market proceeds divided by the claim. The waiting-return chart annualizes a 0.94-to-1.00 gross conversion over specified days assuming no fees, changes in backing or interruptions. The loan scenario assumes 100 LST, $2,000/ETH, $120,000 debt, a fixed ETH price and an illustrative 65% liquidation LTV; real lending contracts set their own rules.
Sources
[1] Ethereum.org, pooled and liquid staking overview
[2] Lido, stETH/wstETH integration and accounting guide
[3] Coinbase, cbETH white paper
[4] Lido, WithdrawalQueueERC721 contract and claim rules
[5] Lido, public risk disclosure (2026)
[6] Coinbase Exchange, cbETH redemption procedure
[7] Ethereum Launchpad, validator slashing FAQ
[8] Ethereum.org, validator withdrawal process
[9] Lido, protocol levers, pauses and bunker mode
[10] Lido, Liquid Staking Risk Disclosures
[11] Coinbase, cbETH market and conversion disclosures
[12] Ethereum EIP-7002, execution-layer validator exits
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape. You can reach out to him via his social media accounts:
Linkedin: https://www.linkedin.com/in/johan-shamshad-742851262/
X: https://x.com/Yasmine_FX
Investing: https://www.investing.com/members/contributors/279781574