Institutional vault infrastructure whose flagship Monad product, earnAUSD, has fallen roughly 80% from its January 2026 peak — a concrete number worth knowing before treating it as steady.
Upshift positions itself as infrastructure for institutional on-chain asset management — letting professional managers design and launch ERC-4626 vaults spanning DeFi and CeFi strategies across many chains. It's a spinout of August, an on-chain prime brokerage co-founded by Alexandre Elkrief (August's CEO, previously ran a trading book at a DeFi hedge fund) and Aya Kantorovich (August's co-CEO, a founding team member at institutional crypto brokerage FalconX, where she built a desk serving 700+ clients). August itself raised a $10M Series A led by Dragonfly Ventures, citing roughly $7B in monthly facilitated volume at the time. Depositors put a single asset into a vault; the vault's manager allocates it across underlying venues (lending markets, DEX liquidity, basis trades) within a constrained "policy engine." Vaults also connect into August's separate prime-brokerage layer that lets KYC'd institutional depositors pledge deposits above $500,000 as collateral, cross-margining DeFi and CeFi positions together from one account — a capability historically reserved for institutional TradFi clients.
earnAUSD's roughly 80% decline from its January 2026 peak needs a more careful read than "something went wrong." No evidence was found that earnAUSD's own contracts, or the underlying AUSD lending markets it allocates into, were hacked or suffered bad debt — the one confirmed 2026 security event touching an allocation venue, Curvance's May 2026 eBTC exploit, was contained to an isolated market and doesn't appear to have reached AUSD pools specifically. earnAUSD's per-share value has stayed close to $1 throughout 2026 in every snapshot found, meaning this looks like a TVL and redemption exodus rather than a loss of principal. The best-supported explanation, though not an official Upshift postmortem: a large share of Monad's broader DeFi TVL is incentive-farmed rather than organic (one estimate attributed roughly half of a $277.5M Monad TVL snapshot to capital chasing ecosystem incentives specifically), and MON's price fell roughly 50% from its own peak over the same period — both plausible reasons for yield-chasing capital to exit, compounded by sector-wide risk aversion after a wave of large 2026 DeFi hacks elsewhere. Treat this as the best available inference from public data, not a confirmed cause.
Against Yuzu Money, which explicitly tranches risk (a junior first-loss layer ahead of senior holders) and names third-party insurance and proof-of-reserve vendors, Upshift's earnAUSD has no tranching or named insurer found — every depositor holds the same pari passu claim, so a loss anywhere in its underlying Curvance, Euler, Neverland or Morpho allocations would in principle hit all depositors proportionally rather than being absorbed by a buffer first. Against Beefy's fixed, coded, bot-executed strategies, Upshift is the structural opposite — active, discretionary allocation by a named manager who can move capital between protocols within a mandate, adding a curator-judgment risk layer Beefy's approach doesn't carry. Against Backpack's centralized custody, Upshift is non-custodial by contract design: Gamma Research cannot withdraw funds to an external address, only move them between whitelisted strategy contracts within the policy engine's guardrails.
Upshift's flagship Monad vault, earnAUSD, launched November 24, 2025 with an initial deposit cap and grew quickly — reaching a peak on Monad around mid-January 2026 — but has since fallen roughly 80% from that peak. Gamma Research manages the vault's allocation across Curvance, Euler, Neverland and Morpho Blue lending markets on Monad, though Gamma Research itself has no independently verifiable public team roster or track record on other mandates found anywhere — a real transparency gap relative to Upshift's other, better-documented curators like MEV Capital and K3 Capital active on its non-Monad vaults. A newer, much smaller native-MON vault, earnMON, has also launched.
The steep decline in Monad TVL from its January peak is a concrete data point worth taking seriously before treating Upshift as a stable or growing Monad presence, even though the best available evidence points to a redemption exodus rather than a loss event — the exact cause hasn't been officially confirmed by Upshift or Gamma Research. Depositors also carry a layered risk stack: exposure to whichever underlying Monad lending markets the vault allocates into (including Neverland, which carries its own single-audit history noted separately), to Gamma Research's allocation decisions from a curator with no independently verifiable track record, and to the separate prime-brokerage infrastructure used for cross-margining. No evidence of insurance, a first-loss tranche, or proof-of-reserve attestations was found for Upshift vaults.
Note: No confirmed governance token yet — Upshift runs a pre-token points program that third-party trackers speculate precedes a future airdrop, unconfirmed by the team.
No direct hack of earnAUSD itself was found. The one confirmed 2026 incident touching an allocation venue (Curvance, via the May 2026 Echo Protocol eBTC exploit) was contained to an isolated market and doesn't appear to have touched AUSD pools. The vault's per-share value has stayed near $1, so this looks like a redemption exodus tied to fading Monad incentive-farming and broader risk-off sentiment, not a loss of principal — though Upshift hasn't published an official cause.
No evidence of insurance, a first-loss tranche, or proof-of-reserve attestations was found for Upshift vaults, unlike Yuzu Money which names both. Protection is limited to audits and architectural guardrails, neither of which guarantees against loss.
Structurally, losses inside an isolated lending market are absorbed by that market's suppliers — earnAUSD depositors pro rata — not by Gamma Research or Upshift out of pocket. No compensation fund or liability commitment was found; risk sits with depositors, mitigated only by ongoing curator diligence and whatever isolation the underlying protocol provides.
Sources: Upshift Documentation, Agora and Upshift partner to deploy AUSD and earnAUSD on Monad — Metaverse Post, Upshift exits stealth — Blockworks, Is Monad's record $477M TVL organic or incentive-driven — AMBCrypto, Echo Protocol suffers $76M exploit in eBTC minting attack on Monad — CoinDesk
Last reviewed 2026-07-08. More Monad research.
Monad DeFi board · Markets · Swap on NullTerminal