NEMO USDC Prime on Upshift: what sits behind a "market-neutral" 26–30% a year
Data as of 26–27.10.2026.
In short. NEMO USDC Prime is a managed vault, not a protocol. Half the money sits on one options venue, exits take up to 31 days, the track record is 4 months long and the Sharpe ratio reads 12–15. Roughly a third of TVL is shares held by a second vault from the same manager. The 26–27% a year net is real share-price growth, but the vault has not yet passed a single stress test. Verdict: not admitted for now; the natural review point is 06.01.2027, when the vault turns six months old.
What it is
An ERC-4626 USDC vault on Ethereum; the share token is nUSDC. The platform is Upshift (parent company August Digital), the strategy is run by NEMO Trading. The curator's description: "market-neutral vault… combining quantitative rate-arbitrage and options strategies… without directional crypto exposure" (Upshift vault page, 26.10.2026). Launched on Upshift on 06.07.2026; tracker data starts 26.06.2026 (tradingstrategy.ai).
- Domain: DeFi, curated vault with dollar yield (confidence 0.9). Class: actively managed hedged carry plus options — effectively a hedge fund in a vault wrapper (0.8). Stage: early, 0.3 years of history (0.9). Jurisdiction: not stated on the vault page or on nemo.trading.
- What you hold: nUSDC is a claim on the vault's NAV. NAV is made up of sub-account positions on Derive, Hyperliquid, Aave V3 and Morpho. The wrapper (Upshift's contract and policy engine) and the underlying (NEMO's trading book) are different objects. The risk lives in the book.
- Who is behind the team. NEMO Ventures (nemo.ventures) lists Lighthouse (lighthouse.one, a portfolio terminal), DegenScore and JPGmorgan as its products. The NEMO Trading page describes a "small team… operating together since 2021", with first names only. Nowhere does it say outright that NEMO Trading and NEMO Ventures are the same legal entity; the link is only indirect, through the name and branding.
Positioning
Niche: curated "market-neutral" USDC vaults (risk curators on vault marketplaces such as Upshift, Morpho, Euler). On the tracker NEMO ranks #1 on Upshift, #11 on Ethereum and #78 overall (tradingstrategy.ai, 27.10.2026); DefiLlama added NEMO under "Risk Curators" on 18.09.2026 (PR #21104); no aggregate niche market shares were found. Market: global, on-chain. Upshift deposits were above $479m on 20.08.2026 versus a peak above $550m (−13%; company's own figures, Upshift blog, updated 16.10.2026). The vault itself holds $13.43m against a $15m cap (Upshift, 26.10.2026): the cap is 90% full. Market stage: the curated-vault niche is scaling after a crisis of trust (the Stream Finance xUSD collapse on 03.11.2025, an external manager's loss of ≈$93m, Yahoo Finance), confidence 0.5; the vault itself sits on the edge between early discovery and validation, confidence 0.8: what is worth validating is what capital cannot fake — share-price growth through a stress event, not TVL. Macro stage: macro regime supportive but fragile (reading of 19.07.2026, now stale); more recent — the Fed hiked 25 bp on 16.09.2026, CPI 3.4% (August), policy room is narrowing; the phase is late cycle with fragility, where high funding pays carry but is exactly what precedes liquidation cascades (10–11.10.2025: 87% of liquidations were longs). Timing conclusion: early — four months without a single recorded stress event cannot tell skill apart from selling volatility tail risk, and a Sharpe of 12–15 is a sign of the latter; the natural point to reassess is 06.01.2027, when the vault turns six months old.
Findings
Verdict: not admitted for now. Only all of the following together would move it to "admitted with a limit": (1) ≥ 6 months of history (06.01.2027) with no share-price drawdown deeper than −1%; (2) a published breakdown of the book: the hedge on the ETH leg and the options positions on Derive, reconcilable on-chain; (3) ETH Prime's share of USDC Prime capital below 10%; (4) independent NAV attestation (Securitize Fund Services is named as the platform's administrator, but not for this vault).
What revokes any positive verdict: a daily share-price drop > 0.5% · a withdrawal queue > 31 days · a sub-account pause · a curator change · any single venue above 60% of NAV.
- The yield is real, but it is one calm stretch. Share price 1.0825 USDC on 26.10.2026 (Upshift). The tracker shows 1.08 and +8.0% since inception, 26.2% annualised CAGR net, 26.6% annualised over 1 month and 27.2% over 3 months (tradingstrategy, 27.10.2026). Cross-check: 1.0825 over ≈122 days from the 26.06 start is ≈26.8% a year — it reconciles. Mismatch: Upshift shows 30.00% APY (30D), the tracker shows 26.6% for the same month. The 3.4 pp gap is either gross vs net or different windows; the page does not explain it. Against a US Treasury bill at ≈4.1% (08.10.2026), the premium is ≈22 pp net. Before 2% + 20% fees, the book would need to earn ≈35% a year gross (26.6 / 0.8 + 2).
- A Sharpe of 12–15 on 1.6% volatility over 4 months is not proof of quality — it signals that the tail has not shown up yet. Published vault Sharpe ratios tend to be overstated, and volatility-selling strategies periodically suffer deep drawdowns. Precedent: 12–13.03.2020, realised volatility ≈250% against IV ≈75, and several vol sellers' accounts on Deribit were auto-liquidated. The model says "tail premium", the market so far says "zero losses": the tracker's max drawdown is −0.3% since inception.
- Concentration on one options venue: 51% of NAV. Allocation on 26.10.2026 (Upshift): Derive $6.87m (51.2%) · Hyperliquid $2.08m (15.5%) · Aave V3 $1.85m net (13.8%: $5.59m wstETH against $3.74m USDC debt, LTV ≈67%) · Morpho $1.27m (9.4%: cbBTC, WETH, WHYPE against USDC debt) · idle $0.73m (5.4%). Mismatch: the components add up to $12.80m against a $13.43m NAV — $0.63m (4.7%) does not appear in the breakdown. Through the wrapper, a depositor holds concentrated Derive risk. The $5.59m long wstETH leg has to be hedged with a short on Hyperliquid or Derive; the page does not show it, so "no directional exposure" is the curator's claim, not verified on-chain.
- ≈37% of capital is another vault from the same curator. DefiLlama, PR of 17.09.2026: "ETH Prime holds USDC Prime receipts… about $5.0M of internal duplication". 5.0 / 13.43 ≈ 37%, if the share has not changed. An ETH Prime exit is a third of TVL leaving on one decision by the same manager, while only 5.4% sits idle for instant withdrawals. This is not fraud but a circular dependency: two NEMO products holding each other.
- Exit takes up to 31 days; instant exit carries an unstated fee and only draws on a ≈$0.73m buffer (Upshift, 26.10.2026). The platform: "Every vault processes claimable redemptions daily… instant redemption for a fee, subject to available liquidity in the buffer", "every vault can lose value". There is no in-house insurance; the 11 audits by 6 firms cover Upshift's contracts, not NEMO's trading book (Upshift blog, updated 16.10.2026). The tracker rates "Protocol technical risk: Severe".
On trusting the team. A well-known team behind a brand is trust in people, not in the book. Past product success colours the assessment of a new one (the halo effect); Stream Finance was also run by a "well-known" external manager. Reputation supplies no number for the analysis: no hedge NAV, no loss limit, no exit terms. The worst plausible scenario for this class is a Stream-style break: xUSD went 1 → 0.23 → 0.08 dollars, i.e. −77…−92% of share value.
What kills it, and at what number
- A volatility tail on the options leg. An event like 12.03.2020 or 10–11.10.2025: a sharp ETH/BTC move against short options on Derive. Number: a daily nUSDC share-price drop > 0.5%, or a tracker drawdown deeper than −1% (currently −0.3% since inception).
- The circular capital unwinds. ETH Prime pulls its ≈$5m. Number: USDC Prime TVL below $9m with the cap unchanged, or a withdrawal queue above 31 days.
- Marketing outruns reality. Number: 30-day realised share-price yield more than 3 pp below Upshift's displayed 30D APY for a second month running. At that point the premium should be measured off the realised share-price curve, not the shop window.
What we don't know
- Vault admin rights: who administers the contract, whether there is a timelock on changing the curator and the protocol whitelist, who signs the owner multisig. The "curator withdraws the funds" scenario depends on it (Upshift blog: "only the owner multisig can withdraw… only in an emergency"). The data is on-chain but has not been read yet.
- The hedge and the options book: whether there is an ETH short on Hyperliquid against the $5.59m of wstETH, and what exactly sits on Derive — short or long options, delta, margin. Both venues are on-chain; the data is public.
- The share-price curve from the contract: nUSDC convertToAssets history by block over 30 and 90 days. The yield figures here come from the tracker and the vault page, not from the contract.
- The instant-exit fee: the Upshift page says "subject to… fee" but gives no figure.
- Legal entity, jurisdiction, full names of the team, an independent audit of the trading book — searched nemo.trading, nemo.ventures, X @NEMOtradingHQ, LinkedIn, Coinlaunch; none found.
- Incidents with Upshift vaults — none found in the Stream/xUSD coverage (November 2025) or on the Upshift blog. That is "not found", not "did not happen".
What would change the conclusion (by impact): a breakdown of the hedge and options book that reconciles on-chain · a stress event survived (a daily ETH move ≥ 10%) without a share-price drawdown > 1% · ETH Prime's share below 10%.
Check calendar
| Date | Object | What to capture | Where | Forecast |
|---|---|---|---|---|
| 27.11.2026 | NEMO USDC Prime nUSDC | share price on 27.11 vs 1.0825 on 26.10, and the 30D APY on the page | app.upshift.finance and tradingstrategy.ai | Annualised realised share-price yield over 27.10–27.11.2026 will be below 30%, i.e. below the 30D APY Upshift showed on 26.10 |
| 27.11.2026 | NEMO USDC Prime allocation | Derive share of NAV on the vault page | app.upshift.finance | On 27.11.2026 Derive will hold more than 40% of vault NAV |
| 06.01.2027 | NEMO Trading | independent NAV attestation or trading-book audit, hedge breakdown | nemo.trading, X NEMOtradingHQ, Upshift blog | By 06.01.2027 NEMO will not have published an independent NAV attestation for the USDC Prime vault |
| 06.01.2027 | NEMO USDC Prime | max drawdown since inception and TVL | tradingstrategy.ai, DefiLlama NEMO Trading | On 06.01.2027 the tracker's max share-price drawdown will be no deeper than −1% |
Sources: Upshift — NEMO USDC Prime · tradingstrategy.ai — NEMO USDC Yield · tradingstrategy.ai — NEMO curator · DefiLlama Adapters PR #21104 · Upshift — What Upshift does and doesn't do · The Block 22.04.2026 — Upshift and Securitize · NEMO Ventures · Yahoo Finance — Stream Finance $93M · Decrypt — Stream Finance
Research for information only. Not investment advice.