Skip to content
PROTOCOL RESEARCH

Neutrl risk

A young (public mainnet Nov 2025) CeFi-dependent synthetic dollar earning basis and OTC-discount yield, with real named audits but a February 2026 integration exploit, a March 2026 DNS/frontend hijack, whitelisted KYC-gated redemption behind a cooldown, and admin-key/timelock opacity.
AvoidResearched Jul 4, 2026

Scorecard

watch: Audits & contractsReal audits (Cantina/Spearbit Jul 8-15 2025: 0 crit, 0 high, 1 med, 3 low, all fixed; Sherlock contest, 118k USDC prize, Aug 2025), but a Feb 2026 exploit drained the Pendle SY-NUSD wrapper integration.
weak: Admin controlCeFi-dependent governance; issuer admin freeze/denylist controls confirmed; whether mint/denylist keys sit behind a timelock or multisig is not confirmed this run.
strong: OracleChainlink pricing via the Euler ERC-7726 standard, on-chain settlement, Hypernative real-time pauser.
weak: Liquidity & exitDirect redemption is whitelisted/KYC-gated with a ~10-day cooldown (an instant buffer path exists); retail exit is via a thin Curve pool (~$5M), not open par arbitrage.
strong: Yield (real vs emission)Yield is real (basis funding + OTC discounts + collateral yield), no token emissions; sNUSD ~8.7% and compressing from beta highs.
watch: Holder concentrationSize differs by slice (~$70M NUSD circ mcap, ~$119M sNUSD TVL, ~$170M protocol TVL); holder concentration and locked-token backing share not verified this run.
weak: Track recordPublic mainnet Nov 2025 (~8 months); already hit a Feb 2026 integration exploit and a March 2026 DNS/frontend hijack; TVL down ~24% in a month.
🟢 strong🟡 watch / caveat🔴 weak / fund-loss risk
Verdict is a gate (worst flaw wins), not an average. Our read, not financial advice.

Live facts

auto-sourced now
TVL—
30dflat
Auditsn/a
Last hackamount undisclosed · Mar 2026

DeFiLlama + our exploits feed. Cross-check the dated report against today.

TL;DR

Neutrl issues NUSD, a synthetic dollar that earns yield from a delta-neutral basis trade plus OTC-discounted altcoin purchases. The reserve is mostly liquid stablecoins (about 93.5% per Pharos), with a smaller sleeve of OTC-locked altcoins hedged on perps, so it is less of a pure altcoin bet than the pitch implies. But the structure leans on centralized custody and off-exchange settlement, direct redemption is a whitelisted KYC queue rather than open arbitrage, and issuer admin/freeze controls sit over minting and the denylist. Contracts were audited by Cantina/Spearbit and a Sherlock contest. Two real 2026 incidents land it in avoid territory: a February integration exploit and a March DNS hijack.

Checklist

Audits & contracts. Cantina/Spearbit reviewed the Neutrl contracts July 8-15, 2025 (0 critical, 0 high, 1 medium, 3 low, 4 informational; the medium and lows fixed), and a Sherlock audit contest ran with a 118,000 USDC prize and an August 2025 report; Hypernative provides real-time monitoring. That is a credible audit stack. The caveat is that on February 4, 2026 an access-control flaw in the Pendle SY-NUSD wrapper (a burnFromInternalBalance flag any caller could set) was drained for about 45.9k USDC via a Morpho flash loan. Neutrl's core mint/redeem was not the flawed contract, but NUSD/USDC LPs and the SY reserve took the loss, so NUSD holders were exposed through an integration, not through their own choices. Admin control. Pharos classifies governance as CeFi-dependent, and issuer admin/freeze controls over the restricted-staker denylist are on record. Whether a timelock or multisig sits on the mint and denylist keys is not confirmed this run; the docs describe "policy-based approvals" and a plan to progressively decentralize, but no on-chain timelock/multisig is evidenced. For a fund-backed dollar this is the central weakness.

Freeze / seizure. On the NUSD contract itself DENYLIST_MANAGER_ROLE can flag an address so that every transfer to or from it reverts, and REDEEMER_ROLE can burn a holder's balance outright - individual-holder freeze and seizure, not a protocol-wide pause; sNUSD carries a comparable blacklist that can seize and redistribute a staker's balance, neither power has any use on record, and no timelock or multisig in front of them is confirmed. Oracle. Chainlink pricing consumed through the Euler ERC-7726 standard, with on-chain settlement and a Hypernative pauser. This is a genuinely strong part of the design. Liquidity & exit. Direct redemption is a permissioned, whitelisted process gated by KYC/KYB, with a reported 10-day unstake cooldown (one doc reference says 30-day, so the exact figure varies); an AssetReserve liquid buffer serves smaller redemptions instantly. Retail without whitelist status exits through the Curve secondary pool, which is thin ($5M). You cannot rely on open-market arbitrage to exit at par during stress; you are either in a KYC queue or hitting a shallow pool. Yield: real vs emission. The yield is real: perp funding on the hedge, the discount on OTC-locked tokens, and yield on the stablecoin collateral. No token emissions prop it up. sNUSD has compressed from beta double digits (30-50% reported) toward roughly 8.7% as TVL scaled and OTC supply tightened. Separately, points and Founder Locks farming are live, which attracts mercenary, flighty deposits. Holder concentration. Size depends on the slice: CoinGecko shows ~$70.3M NUSD circulating market cap, Stablewatch puts sNUSD (staked) TVL near $119M, and Pharos/Messari cite ~$170M total protocol TVL (down from ~$225M a month earlier). These are consistent different measures, not a contradiction, but holder concentration and the exact share of illiquid locked-token collateral are not verified this run. Track record. Founded early 2025 (a $5M seed in April 2025), pre-deposit vaults opened Oct 15, 2025, and public mainnet went live Nov 10, 2025 - so roughly eight months of live public history. In that window it has taken an integration exploit (Feb 2026) and a DNS/frontend hijack that forced a contract pause and domain migration (March 2026), and TVL fell about 24% in a single month. Short history, two incidents, shrinking book.

Worst case

A basis trade on OTC-locked altcoins gaps against the hedge while the locked collateral cannot be sold, the peg slips, and non-whitelisted holders are left with a thin Curve pool instead of par redemption. Because minting and the denylist sit behind issuer admin controls inside a CeFi-dependent structure with no confirmed timelock/multisig, a key compromise or custodian failure is a direct path to loss, and the March DNS hijack shows the attack surface is not only on-chain. The February exploit already demonstrated that an approved integration can bleed NUSD backing without touching the core contracts. There is no confirmed protocol-level insurance fund large enough to absorb a real basis blowup (Ceffu custody carries an English-law trust structure, but that protects custody assets, not a strategy loss).

Bottom line

Avoid for the showcase. The engineering is better than the average young synthetic dollar (Chainlink/Euler oracle, named audits with a clean severity profile, on-chain settlement, a real-time pauser), and the yield is genuine rather than emission-fueled. But three fund-loss surfaces are red at once: issuer admin controls under CeFi-dependent governance with no confirmed timelock/multisig, exit gated behind a KYC redemption queue with thin secondary liquidity, and an ~eight-month public track record already marked by an integration exploit and a frontend hijack. Worst-flaw-wins puts this at avoid until custody, the admin-key controls, and a reserve/insurance backstop are transparently verifiable.

Data appendix

  • TVL / supply: ~$170M total protocol TVL, down from ~$225M a month prior, a ~24% drop (Pharos/Messari). sNUSD (staked) TVL ~$119.3M (Stablewatch). CoinGecko shows NUSD circulating market cap ~$70.3M at ~$0.999. DeFiLlama has no "neutrl" protocol page as of this run (verified against the full protocols list). The slices are consistent (circulating NUSD < staked sNUSD < total TVL); treat the headline size as ~$170M. 90d/ATH TVL history: refresh.
  • Audits: Cantina/Spearbit review of "Neutrl Contracts" July 8-15, 2025 (0 critical, 0 high, 1 medium, 3 low, 4 info; medium + lows fixed, researchers Anurag Jain and ladboy233); Sherlock audit contest (118,000 USDC prize) with an August 2025 report; Hypernative real-time monitoring/pauser (Cantina portfolio page; Sherlock contest page; WebSearch). Sherlock final severity counts: refresh.
  • Admin/governance: CeFi-dependent (Pharos), issuer admin/freeze + denylist controls on record; "policy-based approvals," progressive-decentralization plan stated. Timelock/multisig on mint/denylist keys: refresh.
  • Oracle: Chainlink pricing via Euler ERC-7726 standard, on-chain settlement (Hypernative blog; WebSearch of Neutrl docs).
  • Collateral & custody: ~93.5% liquid stablecoins plus a smaller OTC-locked-altcoin basis sleeve hedged on perps (Pharos). Collateral routes to segregated vaults at institutional custodians (Fireblocks, Copper, Ceffu) with OES via Copper Clearloop / Ceffu MirrorX; Ceffu adds an English-law trust structure (Neutrl docs / Messari, corroborated this run - upgraded from X-only). A specific ~$4.8M reserve-fund figure remains X-sourced: refresh.
  • Peg history: Held near $1 through recent stress; shallow sub-200bp blips on record. ATH $1.03 (Jan 31, 2026) / ATL $0.9754 (Nov 20, 2025) (CoinGecko).
  • Exit terms: Direct redemption permissioned/whitelisted with KYC/KYB and a 10-day cooldown (one doc says 30-day), plus an instant AssetReserve-buffer path; non-whitelisted retail exits via a thin Curve pool ($5M) (Neutrl docs; Pharos).
  • Yield: sNUSD ~8.7% (Pharos/Stablewatch, June 2026), compressed from 30-50% private-beta highs; no emissions; points/Founder Locks farming live.
  • Incidents: Feb 4, 2026 exploit of the Pendle SY-NUSD wrapper (burnFromInternalBalance access-control flaw, SY-NUSD proxy 0x29ac34026c369d21fe3b2c7735ec986e2880b347), ~45.9k USDC drained via Morpho flash loan, borne by NUSD/USDC LPs and the SY reserve, core mint/redeem unaffected (DARKNAVY). March 2026 DNS/frontend hijack, contracts paused, domain migrated to a new DNS provider, no contract compromise; part of a coordinated registrar-social-engineering campaign (Phemex; CryptoTimes). Maintained monthly. Methodology: DeFi Research Instruction v2.

← all protocols