HostHostDeFi › Library › Is sDAI safe on Ethereum?
Is sDAI safe on Ethereum? The mintable vault
sDAI on Ethereum grades B (75) on our September 26, 2026 read — one bad flag, a live mint, on the savings wrapper where minting is the product.
sDAI — Savings Dai, the yield-bearing wrapper for DAI deposits — on Ethereum grades B (75) on our September 26, 2026 read. One bad flag: the contract can mint new supply. The rest of the surface is clean: no freeze, verified source, creator under 0.01%. Market snapshot: ~$1.34B reported cap, ~$5.4M of tracked liquidity, ~3,634 holders, 932 days live, ~$1.8M of daily volume.
Of every mintable contract in the series, this is the one where the flag is least sinister and most necessary: a savings wrapper is mintable because it works.
A mint that is the product
sDAI is a vault share: deposit DAI into the DAI Savings Rate contract, receive sDAI, redeem later for DAI plus accrued yield. Minting is how deposits become shares — the power isn't a backdoor around the design, it is the design. Every savings-share token in DeFi carries the same signature: ERC-4626-style mint-on-deposit, burn-on-redeem.
The checklist flags it anyway, and rightly: the scanner reads the contract's capability surface, not its intent. "Can mint" is literally true — a holder's dilution protection doesn't come from the function being absent but from the mint being bound to deposits. The B-75 rather than a D is the flags' way of saying powered, but only the one power, on a verified contract.
932 days: the longest-lived B in this batch
Two and a half years on-chain makes sDAI the elder of the powered cohort. The age does real work here: the mint has been exercised constantly (every deposit is a mint event) and only ever in the shape the design specifies. A bad-actor mint would have shown up as supply inflation detached from deposits; 932 days of operation without that divergence is the strongest "hasn't" a snapshot can accumulate.
~3,634 holders is the other honest number: smaller than the token's reputation suggests, because sDAI's usage is largely programmatic — protocols hold it as treasury collateral, aggregators hold it inside vaults, and the index counts wallets, not positions nested inside positions.
$5.4M of pool on a savings share
The warn is thin-for-cap liquidity (~$5.4M on ~$1.34B ≈ 0.4%), and on a savings wrapper the warn deserves a footnote: the exit that matters for sDAI isn't the pool at all — it's the redeem function. A holder exiting converts shares back to DAI at the protocol rate; the DEX pool is the secondary market, sized for convenience rather than redemption volume.
That doesn't make the warn wrong — a holder who needs the on-chain door still finds a thin one — but it does make it differently-shaped: on sDAI the checklist's pool number measures the fast exit, while the slow exit (redemption at the rate) sits behind it, contract-guaranteed. ~$1.8M of daily volume says the fast door gets used anyway.
The powered cohort, honestly tiered
The batch this page belongs to is a survey of how a live mint distributes risk: sUSDe-Ethereum at E (mint plus freeze plus balance-writes plus hidden owner), CRV-OP at D (mint plus writes), SKY and sDAI here at B-75 (mint alone, verified, exercised through protocol machinery). The checklist's ordering is doing what it should: the same bad flag is cheap on a contract where it's the only power, and compounding on a contract where it's three of four.
sDAI's slot — the least-alarming member of the mintable set — is also the most instructive: it's proof that the checklist doesn't punish the flag for existing, only for existing alongside the other verbs. A savings wrapper that can't freeze, can't rewrite balances, and mints only against deposits is a B, not because the mint is safe, but because the mint is bounded.
Where the risk actually lives on sDAI
One paragraph of honesty the flags can't carry: sDAI's real risk surface isn't mint-or-freeze — it's the DAI Savings Rate's solvency and Sky's governance of the rate. A holder's dilution protection is contract-level (mint is deposit-bound); their yield protection is policy-level (the rate is set by governance). The B-75 grades the first; the second is a system question the series keeps naming on every savings instrument it covers.
The hidden holder graph
Worth dwelling once on why ~3,634 holders undercounts the instrument: sDAI lives inside other contracts — Spark/Sky protocol positions, yield aggregators, treasuries, L2 vault strategies — so the wallet count measures direct custody, not exposure. A protocol holding sDAI for ten thousand users shows up here as one row. It's the same blind spot as PYUSD's treasury concentration read the other way: the number is true, the interpretation is incomplete.
For the page's purposes that's mostly color — the flags don't depend on holder count here — but it keeps the small number honest: a savings share that most of DeFi holds through something else will always read fewer holders than it has beneficiaries.
The clean-verb checklist, one more time
The full surface on this read: cannot freeze, cannot blacklist, cannot rewrite balances, owner unattributed-but-inert (creator under 0.01%), source verified. The only live verb is the mint, and the mint is bounded to deposits. It's the closest thing the series has to a template for "how to carry a powered function honestly": one flag, visibly necessary, on a verified contract with a two-and-a-half-year record.
Every other powered page in this batch is the counter-example — powers beyond need, powers unbounded, powers hidden. sDAI is what the same flag looks like when the contract gives it nowhere to hide.
What the B-75 is buying you
For a holder the practical content of this grade: your shares can't be frozen, your balance can't be edited, the contract's source is public, and the only live power is the one that turned your DAI into the share you hold. The residual risks — the rate, the solvency, the governance — are the same risks any savings instrument carries, on-chain or off. The checklist's verdict is narrow by design and this page keeps it that way: B-75 means "the contract layer is bounded," full stop.
Why it isn't an A anyway
The question a sympathetic reader asks — if the mint is structurally necessary and bounded, why isn't this an A? — has a two-part answer. First, the checklist grades capability, and the capability is genuinely live: the deposit-bounded property is enforced by the minting path, not by the mint function being absent. Second, honest grading doesn't read design intent; an A means "no powers to misuse," and "bounded power" isn't "no power." The B-75 is the compromise the letter scale makes with ERC-4626-shaped reality: good design earns the high-B, the flag keeps the flag.
Holding it, honestly
If you hold sDAI on Ethereum you hold a savings share whose mint is live because minting is what deposits are — one powered verb, bounded by design, on a verified contract 932 days into operation. The B-75 is the checklist's honest middle: the flag is real, the power is load-bearing, and the two-and-a-half-year record is what the word "bounded" is standing on.
Frequently asked
Is sDAI safe on Ethereum?
B (75) on our Sept 26, 2026 read: live mint (structural — deposits mint shares), no freeze, verified source, creator under 0.01%. ~3,634 holders, ~$5.4M liquidity, 932 days.
Why does sDAI have a mint function?
It's a vault share — minting is how DAI deposits become sDAI shares (ERC-4626 mechanics). The flag is real but the power is bounded to deposits.
How do I exit sDAI — pool or redeem?
Both work: the DEX pool (~$5.4M tracked) is the fast secondary market; the redeem function converts shares back to DAI at the protocol rate — the deep exit for size.
Is sDAI's grade worse than USDS's?
Comparable band — both are the Sky system's tokens with live issuance paths (USDS B/80 series page exists separately). sDAI's mint is the narrowest of the powered cohort.
What risk does the grade not cover?
The DAI Savings Rate's solvency and rate policy — system-level questions set by Sky governance. The B-75 grades the contract's powers; the yield sits one layer up.