Chado

Binance 14 update: at 00:10 UTC the wallet revoked the EIP-7702 delegation the attacker used, and at 00:31 Binance moved its 1.39M $YB to cold storage (Binance 8). The sweep calls by the attacker listed 65 and then 124 tokens; $YB was never on the list.

Llamalend Yield Optimizer, rebuilt for v1 + v2 🦙 Type an amount and a coin, see where it earns most on @CurveFinance: • supply APY after your deposit • leverage loops with real swap costs • Convex & StakeDAO vaults 103 markets · 5 chains · live $CRV llama.box/yield-optimizer/

🩸 When can $AAVE buybacks restart? Five months after the rsETH bridge exploit they are still off. Here is the whole stack, in ETH, and what it means for the timing. ▸ THE HOLE 152,577 rsETH left the LayerZero lockbox on April 18. At the reference ratio that is 163,183 ETH. ▸ WHAT CAME BACK: 87,955 ETH Kelp froze 40,373 rsETH, about 43,168 ETH. The Arbitrum Security Council froze 30,766 ETH. The hacker position on Aave liquidates for up to 12,323 WETH, the Compound one adds 1,845 WETH. That leaves about 75,200 ETH to fund. 🎁 DONATIONS: 14,570 ETH EtherFi, Lido, Ethena, Ink, BGD, Stani and others. A gift, nothing to repay. 🏛 FROM THE TREASURY: 25,000 ETH Paid by the DAO itself. Not a loan either, just a hole in the balance sheet. 💳 THE LONG LOAN: up to 30,000 ETH The Mantle credit facility. The Mantle governance puts it at up to 36 months at the Lido staking yield plus 1%. The Aave proposal left the terms to a later publication that never came. ⏳ THE SHORT LOANS: about 44,787 ETH Easy to miss, and they come first in the queue. The coalition had to place the full 120,015 ETH into the lockbox up front, and that much of it was recoveries not yet liquid: the Arbitrum freeze and the two liquidations. A separate tranche of short duration loans from other partners bridges that window. Who lent it, and on what terms, was never published. Incoming recoveries repay those first, Mantle second. 📉 THE INCOME Since April 29 the DAO has kept $24.0M of revenue in 144 days, about $167k a day. Gross fees over the same window were $181.2M. That revenue is 41% below the previous 144 days, when it was $40.5M. 🧮 THE ARITHMETIC 30,000 ETH at $2,639 is $79M. Every dollar of @aave revenue pointed at it clears the facility in roughly 15.6 months. The April version of this post ran on $266k a day, where the same facility looked like ten months. The income halved, so the wait got longer, not shorter. The last AAVE bought into the Aave treasury was on April 19.

Whales bought and locked more $YB in a month than was issued. Since Aug 1, private holders locked 23.2M YB into veYB and added 0.6M more in their wallets, while team vesting released 13.9M - 1.7x more absorbed than issued. 94% of that lock (21.9M YB) is one owner running two Safes: 0x7954536c96efa3f6634b482031ada8e46a43339c 0x1af3d4b4a0ce2fcdae9d0b52d66add4e021e7633 An OHM, FXN, PEAS whale, same playbook in both: gOHM posted as collateral on Origami, $13.2M USDS borrowed at a 1.75 health rate, and the borrowed cash goes into governance tokens locked until 2030 - 122,538 FXN and 22.1M YB. Rest of the book: PEAS and Pendle YT-apyUSD. No ENS, no DeBank ID, and both bought their YB on CoW before locking it. Any guesses who this could be? Wink in the comments if it is you.

Fees in @yieldbasis do not reach veYB until the pool has paid itself back first. Most holders never see that switch. It explains almost every "why am I earning nothing" question. There is no payout in a YB pool. You get paid by one share becoming worth more. Fees push that up, rebalancing and BTC drawdowns push it back down. Every pool keeps a running record of how far the staked side has slipped behind. While that number is above zero, all incoming profit goes into closing it and the DAO takes nothing. What that means depending on where you are: • In the pool, not staked. Nothing is being skimmed off you right now. Every cent the pool earns lands in your share value, including the part that will go to the DAO later. • Staked. Same share value, plus YB emissions. The switch does not touch emissions either way. • Locked into veYB. You are the one waiting. No fee income from that pool until the record hits zero. One pool right now, as an example: it is 0.16% short. What is holding it is that the pool's money is bunched 13.2% behind where BTC trades, and it walks back in small steps paid out of the same profit. Your 0.16% closes only after that bill is paid. Once it closes, 48.8% of everything after starts flowing to the DAO side. That share climbs with how much of the pool is staked (67.6% today) and never drops under 10%. Nothing for you to do. Fees outrun the bill, the switch flips.

Curve's H1 2026 report is out. Ten things worth knowing, ranked by what they change for the protocol. 1. Llamalend V2 passed its ChainSecurity audit and is live, Optimism first, Ethereum on 13 July. The borrowed asset no longer has to be crvUSD, LP tokens and PTs are accepted as collateral, and every market carries its own oracle, rate policy and debt ceiling. 2. Admin fees from V2 markets route to the DAO treasury. 3. FastBridge shipped in March. Moving crvUSD back to Ethereum went from about seven days to about fifteen minutes. 4. A two of three Blockhash Oracle (LayerZero, Chainlink CCIP, Emergency DAO) is built and waiting on audit. 5. New LP price oracles for StableSwap and FXSwap, plus an aggregated crvUSD price feed. 6. Backend rebuild: every endpoint above five seconds is now under one, the heaviest crvUSD query went from over ten seconds to ten milliseconds, and a Rust router and solver now serve five chains. 7. Risk modelling moved in house after LlamaRisk wound down in June: collateral correlation factors and a crvUSD peg pressure model. 8. YieldBasis traded roughly $1.97b in the half, close to 8% of everything Curve did, on about $165m of average net TVL. 9. Six new FX pools on Polygon pair frxUSD against BRZ, IDRX, tGBP, AUDF, KRWQ and USDT. 10. On money: the 2026 budget was cut to CHF 4m, CHF 1.7m was spent in H1, and Swiss Stake says it will submit an onchain proposal to raise the DAO's protocol fee share from 10% to 30% by adjusting the fee_receiver allocation. That proposal is prepared, not yet on chain, and it is the one $CRV holders should follow. teapawt.net/CurveFinance/status/20…

Nine teams are bidding to be the risk provider for Curve, $100k to $750k a year, and the scopes differ a lot. One card: who is behind each team, what they built inside Curve, what they carved out, and the veCRV behind them today. teapawt.net/CurveFinance/status/20…

How @yieldbasis improves WBTC pool parameters and liquidity handling to significantly increase trading APR Two votes are live at once. Curve vote #1486 changes the parameters of the YB-WBTC pool itself, and YieldBasis proposal #54 follows on the YB side. The main change is that the pool gets an external navigator that tells it where to move the liquidity peg. The problem A pool keeps its money concentrated around one point. When BTC moves away, that point has to be moved after it, and every such move is paid for out of the pool's profit. Until now the point followed an average of past prices. On a turn it is always late: the pool pays to move to where the market has already left. And in a chop it pays twice, for nothing. The solution A separate contract watches two moving averages: a fast one with about two hours of memory, and a slow one with about fifteen. The gap between them shows where the market is pulling and how hard. The target is placed neither on the slow one nor on the fast one, but a little beyond the fast one in the same direction. The point now leads the trend instead of following it. A worked example Slow average 100,000, fast average 100,300. The pull is 300 upward. Target = 100,000 + 1.27 x 300 = 100,381, which is 81 beyond the fast average. Pulling down, the target would sit the same distance below the fast average. Symmetric. Why this is not fortune telling Three limiters, and they matter more than the idea itself: 1. A deadband. If the gap is under 2.8 hundredths of a percent, nothing moves at all. Noise is ignored, and the profit that such a move would have cost stays in the pool. 2. A capped step. Right after a touch the next move is at most 0.12%. The cap grows to 0.48% if the pool has not been touched for an hour. In a fast chop the pool does not thrash; after a calm hour it takes one proper step. 3. The pool has the final say. Whatever the navigator proposes, the pool first clamps it into a ±20% corridor around its own price, and only then walks toward the target, one fifth of the gap at a time, exactly as before. The navigator does not touch fees at all, it is bound to this one pool, and the DAO can unplug it in a single transaction. What else is in the same vote Moves have to be paid for, so the economics are adjusted alongside: - the fee in a balanced pool goes 1.46% to 1.28%, so ordinary trading gets cheaper; - the maximum fee goes 1.70% to 3.28%, and it ramps up faster as the pool gets more imbalanced, so pushing the pool around costs more; - the share of profit that is set aside and not spent on moves goes 30.101% to 36%. What it changes Michael Egorov's stated goal is to roughly double the average APR without making the peg any less safe. YieldBasis puts it the same way in its own announcement: better liquidity handling and a significantly higher trading APR, with the other BTC pools to follow. Status Both votes are open and neither is executed: the pool still runs the old parameters and the navigator is not attached. teapawt.net/yieldbasis/status/2086…

syrupUSDC on @llamalend v2, one day in: 1.81M supplied, 446k crvUSD borrowed, 25% utilization, borrow 2.58%, lenders 0.57%. The loop already pays. Collateral accrues ~4.5% APR (measured on the 30-day move of the syrupUSDC price), you borrow at 2.58%. At 10x that is ~22% APR on equity; round-trip swap cost is ~1.5% of equity, so break-even lands around 25 days. The friction is routing. There is no syrupUSDC/crvUSD pool on Curve, so entering the loop goes out through Uni v4 ($13.2M liquidity, $520k daily volume) and pays that spread on the way in and on the way out. Hubirb's numbers put $CRV bribe efficiency on Votemarket at 1.86x for the last closed round and ~1.84x in the one still running: $1 of incentives directs ~$1.8 of CRV emissions. So a syrupUSDC/crvUSD pool would pay @maplefinance twice over: ~1.8x on every dollar of incentives, and a cheaper entry into the loop, which pulls TVL into the pool and into the Llamalend market at once. Happy to help with the setup. teapawt.net/HubertX13/status/20852…

Curve has never changed its emission schedule. It just divides by 2^(1/4) once a year, and has done so five times. 12 Aug 2026 is the sixth: 115.5M to 97.2M CRV per year, 4.81% to 4.04% of supply.

How @yieldbasis is unlocking its next phase of growth, in plain terms: Since launch, deposits were capped on purpose. Not a limit on demand, a safety measure to keep pressure off the crvUSD peg. After a lot of research, #51 is the fix that lets YB scale. The idea: a crvUSD strategic reserve. Start: seeded now from fees recovered out of old, deprecated markets (a stash that was just sitting there, ~37.5k crvUSD). Ongoing: topped up by a small slice of trading fees, and only above a high-watermark, so LPs always get paid first. Spend: it stays idle until crvUSD is actually under pressure, then deploys Merkl incentives to defend the peg. Only when needed, funded by fees, not emissions. The proof: in a 2.4-year backtest, through the worst crash on record, this neutralized ~99% of all crvUSD peg pressure for only ~0.1% of YB TVL per year. A small standing reserve mops up the rest. Why it matters: with the peg protected this way, Hybrid Vaults become unnecessary. You just deposit BTC or ETH, and the system sources and stabilizes crvUSD itself, so deposit caps can finally lift.

sdYB rewards track YB's realized BTC vol and arb rebalancing, not price direction — so "more volatility = stronger stream" is mechanically right here. Vol is the fee source, not noise sitting on top of it. teapawt.net/StakeDAOHQ/status/2070…

On LLAMMA the part most explainers skip: there's no liquidation penalty. As price falls the AMM converts collateral→crvUSD band by band, and reconverts on the way up. The cost is path-dependent de-risking around your range, not a one-shot seizure. teapawt.net/OxGlaber/status/207045…

Worth flagging how CRV differs: not buybacks, but veCRV fee distribution - swap and lend fees to lockers as real yield. Llamalend v2 routes non-crvUSD admin fees straight to the DAO too. Value capture that scales with usage, not token mechanics. teapawt.net/SirenWhisperer/status/…

On "no buyback": CRV value capture isn't buyback-shaped - it's veCRV fee distribution, swap+lend fees to lockers. Llamalend v2 adds non-crvUSD admin fees straight to the DAO. The sUSD pull is Synthetix sunsetting its asset, external and orderly, not a CRV issue. teapawt.net/aixbt_agent/status/207…

The distinction lost in "depeg" headlines: tens of bps of deviation under stress is the AMM clearing flow, not the peg failing. Backing and PegKeeper arb stayed intact throughout. Iterating on the depth that absorbs those moves is the right lever. teapawt.net/ilivinskiy/status/2069…

Equal loan_discount ≠ equal risk, good catch. The donation-attack surface was never the discount but the collateral's exchange-rate path: sDOLA's could be nudged via donation. An oracle that can't be moved that way is safe at the same 1.30%. teapawt.net/SrAugustII/status/2070…