MARKET OUTLOOK
DeFi Yield Market Outlook: What Four Months Did to 35 Strategies
Four months of DeFi yield, in numbers
- 35 strategies tracked; 6 (17%) stopped being viable
- 2 protocol exploits among them — ~$31M extracted (Resolv, Summer.fi)
- Mid-cap synthetic dollars contracted 50–75% in supply
- Flagship stablecoin yields compressed from ~8–12% to ~5%
- Blue chips (Sky, Lido, ether.fi) — flat; the one clear grower: Maple (+81% AUM YoY)
"Before" figures date from our reviews of late 2025 – March 2026; "after" figures were verified July 27, 2026. This article is a dated snapshot — for live numbers, every strategy's report page has daily-updated statistics.
The shape of the market
The last four months rewarded exactly one thing: being established. Everything else — points-driven deposits, single-key admin setups, small pools, euro-denominated experiments — was punished in proportion to its novelty. The October 2025 deleveraging started the contraction; the points-season endings of early 2026 accelerated it; and two exploits finished off protocols that might otherwise have merely shrunk.
The result is a barbell. On one end, casualties. On the other, incumbents that ended the period almost exactly where they started. The middle thinned out dramatically.
The casualties
Six of the 35 strategies no longer exist in investable form. Two died violently, the rest quietly. All six have been retired from our leaderboards.
| Strategy | Before | After (July 27, 2026) |
|---|---|---|
| Resolv — USD0++ Pool & USR Staking | USR: a $1.00 delta-neutral synthetic dollar, ~$238M market cap, stUSR yielding ~8%, RLP junior tranche as insurance | Exploited March 22, 2026: a privileged minting key (single EOA, no mint caps) minted ~80M unbacked USR from $100K USDC — ~$25M extracted. USR crashed to $0.025 and never repegged (~$0.16 in July). Protocol paused. |
| Summer.fi — USDC Lower Risk | Automated "lower-risk" USDC vault; Lazy Summer Protocol peaked near $182M TVL (Sep 2025), Block Analitica as risk manager | Exploited July 6, 2026: a flash-loan attack on the USDC vaults' accounting extracted $6.04M. Summer.fi announced full shutdown; the app runs only until August 31, 2026. |
| Asymmetry — USDaf Yearn Vault | Immutable Liquity-v2-fork stablecoin, ~$6.3M USDaf supply | No exploit — the contracts still run. Supply collapsed ~95% to ~$289K, protocol TVL ~$436K. At that scale any real deposit becomes the whale in the pool. |
| Ripe — sGREEN Stability Pool | Triple-yield stability pool, ~$535K TVL (already small at review time) | TVL of ~$5.9K — roughly the price of a used car. |
| Beefy — stEUR–EURe LP Vault | Euro LP auto-compounder pairing Angle's yield-bearing stEUR with Monerium's EURe | A second-order casualty: Angle voted to wind down EURA/stEUR (redemption open until March 1, 2027) and the LP's liquidity evaporated to ~$41K in the nearest related pool. A strategy can die because someone else's protocol retires. |
Both exploits are recorded in our DeFi hacks tracker, and both fit the same pattern: the yield strategy itself was fine — the scaffolding failed. Resolv lost to a privileged key; Summer.fi to vault share accounting. Neither loss came from the delta-neutral trade or the lending book the depositors thought they were exposed to.
The middle of the market didn't die — it deflated. Mid-cap synthetic dollars gave back most of their points-era growth: Cap's cUSD went from ~$360M to ~$73M, Neutrl from ~$227M to ~$59M, InfiniFi from ~$190M to ~$61M, and Falcon — the largest of the group — from a $2.2B peak to ~$1.26B, with its flagship sUSDf yield compressing from ~12% to ~5%. Even Ethena, the category's blue chip, slid from a $6.5B peak to ~$4.0B and ceded its #5 stablecoin rank to World Liberty's USD1. Functioning protocols, holding pegs — just half the size their marketing pages remember.
What held up — and what grew
The other end of the barbell barely registered the same four months. Size, age, and boring collateral won.
| Strategy | Before | After (July 27, 2026) |
|---|---|---|
| Maple — syrupUSDC | $5.1B originated; an ongoing Core Foundation lawsuit hanging over expansion plans | The stand-out grower: $5.4B originated, $4.6B AUM (+81% YoY), now powering Robinhood's Earn product. The lawsuit settled in May 2026 |
| Ether.fi Liquid — ETH Yield | Vault ~70% eETH/weETH with a 15% Pendle PT book; 2% platform fee | Still #1 in liquid restaking — but the vault quietly became ~two-thirds an Aave V3 lending position, Pendle under 1%, and the platform fee dropped 2% → 0.5%. Same wrapper, different machine inside |
| Sky (USDS/DAI) | ~$11.25B combined, #3 stablecoin | ~$11.5B, still #3. Effectively unchanged |
| Lido | ~$18.8B staked, #1 liquid staking | ~$18.3B, still #1 with ~23% of staked ETH |
| Avant — avUSD | ~$110M supply | ~$115M — one of the only mid-caps that grew, and it rode out October 2025 without visible damage |
| Liquity v1 / v2, OpenEden USDO | Small, conservative, fully collateralized | Small, conservative, fully collateralized. Pegs within half a cent |
The ether.fi row is the subtlest lesson in the table: nothing broke, TVL held, and yet the vault's internal composition changed almost completely — from a restaking-and-Pendle story to a lending book. A depositor who read the launch-era description would be wrong about what their money actually does today. That class of drift is invisible unless you re-check primary sources.
Takeaways
- The base rate of decay is real: 17% of tracked strategies stopped being viable in roughly four months. Diversification across strategies is not optional in DeFi yield.
- Exploits hit scaffolding, not strategies: both 2026 casualties lost depositor money through admin keys and share accounting — not through the trades that generated the yield. Audit coverage of the boring parts matters most.
- Points-driven TVL is mercenary capital: the supply charts of Cap, Neutrl, and InfiniFi all inflected exactly where their incentive seasons ended. Yield that exists to attract deposits leaves with the deposits.
- Flight to quality favors scale: the protocols that ended the period unchanged — Sky, Lido, ether.fi — and the one that grew — Maple — are the largest and most institutional names we track.
- Watch list for next quarter: Gearbox (TVL down from a ~$400M peak to ~$19M, though never exploited), Avantis (~$80M → ~$21.5M), and Angle's stEUR redemption deadline of March 1, 2027.
For live, daily-updated numbers on every surviving strategy, start from the leaderboards; for the exploit history behind the casualties, see the hacks tracker.