7 min read · 12 sources

Bitcoin is pressing into the most contested price band of its 2026 recovery. Eight straight days of ETF inflows totalling $2.8 billion have pushed spot back toward $81,000, where four overlapping forces are stacked against the rally: long-term holders sitting near breakeven, self-custody accumulation starting around $80,800, options dealer hedges activating near $82,300, and dense short liquidation clusters between $82,000 and $86,000. The move was kick-started by the US Treasury’s buyback expansion announcement and then turbocharged by the largest short liquidation event since 2019 on August 19, which compressed futures open interest 11% in coin terms. Analysts are calling $83,300 the line that matters: a daily close above it on continued ETF inflows would suggest the overhead supply is being absorbed rather than defending. A rejection is uglier. A retreat from this band exposes $62,900, the level that would wipe out the entire rally.

A $62,900 print would erase the entire recovery from Bitcoin’s August 19 short squeeze, the largest since 2019.

Moonwell drained for $8.7M on Base

Source: theblock.co ↗

Lending protocol Moonwell is investigating what looks like a textbook illiquid-collateral exploit. Security firms estimate an $8.7M loss after an attacker manipulated the price of the MAMO token, the illiquid asset used as collateral in some Moonwell markets, then borrowed real cbBTC against the inflated value. The pattern is familiar: a thin oracle, a manipulable curve, a collateral asset nobody is actively marking, and an attacker who knows exactly which markets will accept the printed price.

Moonwell’s response was blunt. Borrowing and supply on Base Core Markets have been capped to 1 wei, effectively pausing new activity. The market reacted fast: WELL is down about 13% and MAMO down about 9% over 24 hours. For protocol teams, the lesson is the same one lending markets keep relearning: any collateral asset whose price can be moved by a single trader in a single venue is a bug, not a feature. If your oracle can be pushed, your liquidation engine will eventually be.

Revolut ships EURR, a euro stablecoin on Ethereum

Source: threadreaderapp.com ↗

Revolut, the European neobank with 80 million customers, announced EURR, a euro-denominated stablecoin deployed on Ethereum mainnet and aimed at users in the European Economic Area. The launch plants a high-distribution regulated fintech into Ethereum’s euro stablecoin segment, which has historically lagged USD-pegged peers in adoption and onchain liquidity. The shape of the play mirrors PayPal’s PYUSD: a regulated issuer capturing payment flows onchain rather than waiting for crypto-native issuers to win the corridor.

The timing is not accidental. MiCA’s stablecoin provisions are giving EU issuers a clearer path, and Revolut’s existing euro user base gives EURR a distribution advantage no crypto-native euro stablecoin can match. If EURR gets wired into Revolut’s core app for transfers and card spending, it will likely reset euro stablecoin market share on Ethereum. The interesting question for builders is which rails get reused: the EURR smart contract, the redemption backend, or both.

LayerZero's institutional exchange lands this fall

Source: threadreaderapp.com ↗

LayerZero is launching a blockchain-based exchange for financial institutions this autumn, with a backing slate that reads like a who’s who of traditional market plumbing: Citadel Securities for primary market-making, DTCC for central US securities clearing, and Intercontinental Exchange for exchange operations. The pitch is that LayerZero’s cross-chain messaging layer becomes a settlement backbone for institutional capital markets, not just a bridge for tokens.

If this ships as described, it is one of the more concrete signals that TradFi post-trade is being rebuilt on top of crypto infrastructure rather than the other way around. The relevant detail for engineers is not the exchange UI but the settlement model: who is the canonical source of truth on a fill, who can censor it, and how do you reconcile an onchain event with DTCC’s books. Citadel and ICE do not sign up to play in sandboxes.

Uniswap v3's many pools are functionally one pool

Source: ethresear.ch ↗

A new analysis of 293,273 capacity scenarios across 78 Ethereum token families lands a hard number on something DEX routers have long suspected. The cell-weighted median HHI is 0.930, the top pool in each family captures a median 96.3% of capacity, and only 0.67% of scenarios produce an effective pool count above 2. Multi-pool routing buys you about 3.2% extra capacity on average. In other words, the multiplicity of fee tiers in Uniswap v3 is mostly an illusion at execution sizes that matter.

The authors also reconstructed daily pool states and compared them to observed trading. Actual trading concentration exceeded the already-concentrated modeled capacity on 67.6% of matched family-days, with a Spearman correlation of 0.841 between observed and modeled distributions. Two caveats: the study covers liquidity-layer capacity only, and excludes gas, routing friction, and MEV, so the measured concentration is a floor, not a ceiling. For routing engineers and solver teams, the operational implication is straightforward: for most pairs, sending the order to the single deepest pool is the right default, and the extra hops you are paying for are not buying you what you think.

The social trading thesis, from Vector to Fomo

Source: x.com ↗

The co-founder of Vector, the Solana social trading app acquired by Coinbase at the end of 2025, pulled together the post-mortem. Vector hit about $20M in peak daily volume and roughly $1B all-time before stalling when it pivoted toward pro desktop tooling for a base of memecoin traders that turned out to be concentrated in about 5% of users. Fomo, the inverse bet, is going after TikTok and Instagram audiences new to onchain trading and has just crossed $100M in daily spot volume.

The argument is that retention on Vector looked like a social network, not a trading app, which is why pivoting to pro tooling broke the flywheel. Crypto Twitter was the beachhead that validated the social-signal-plus-execution thesis, but it is a fraction of the addressable market. As equities, RWAs, and prediction markets migrate onchain, and as AI sharpens signal discovery, the bet is that a social trading graph paired with execution rails becomes a category that scales to a trillion dollars. For builders, the design lesson is that the social mechanics are the product; the trading UI is just the conversion surface.

Ethlabs emerges from its first offsite with four bets

Source: x.com ↗

Ethlabs, a nine-week-old nonprofit R&D lab for Ethereum, ran its first in-person offsite in Hunter, NY, and came out with four concrete priorities. EIP-8198, called Quick Slots, makes slot time configurable so the 12-second block time can be brought down in stages and is advancing in the Hegotá fork discussion. A major bridge is preparing to adopt the Fast Confirmation Rule in production, which promises up to 32x faster deposit confirmations in best-case scenarios. On interop, the team set a 1-second cross-chain movement target and split the work into problems needing protocol changes versus problems solvable with ecosystem coordination alone. Finally, the lab is committing to substantive work on ETH as an asset, a topic protocol organisations have historically sidestepped in favour of the Ethereum-as-platform framing.

The takeaway for engineers working across the stack is that the next 12 months of Ethereum client work is going to be shaped by specific, named proposals, not vague scaling narratives. If you operate a bridge, the FCR change is one to track closely.

Smaller beats

Source: threadreaderapp.com ↗

A dozen-plus banks are moving forward with a joint stablecoin venture after a year spent lobbying in favour of tokenised deposits. The pivot is notable: the institutions that called stablecoins a threat to monetary policy a year ago are now building one. A Seoul court ruled in Bithumb’s favour in one of four lawsuits tied to February’s promotion that mistakenly credited 620,000 BTC, about $43B at the time, to users. The exchange credited the windfall, not the error, which is the only legal theory that makes sense given the number.

DeFi is absorbing tokenised stocks through lending markets like Resupply, which accept FraxLend and LamaLend LP positions as collateral so users can keep earning LP fees while borrowing against them. The pattern is the same one Aave and Compound ran on crypto collateral: wrap the position, let the wrapper accrue yield, and let users borrow against the wrapper. The risk is identical too, and concentrated in stablecoin-depeg scenarios where the underlying LP and the borrowed stable diverge in the wrong direction.

KAST’s new business card ships 3% uncapped cashback on the free tier, 4% on the paid tier up to $1M and 3% uncapped after, up to 12% APY on idle cash, and 5% ad-spend cashback for the first 100 approved businesses. All bonus rates step down after a 2026 launch window. The interesting numbers for anyone building a neobank are the unit economics in the 14 revenue streams breakdown that ran alongside the card news: interchange is roughly $15k per $1M in spend, on-ramp spreads run 50-70 bps on the customer side against 5-50 bps of provider cost, and FX markup sits around 2-3% on a Visa base of about 1%. The author’s call is that on-ramp spreads head to zero in 18-24 months, which means the durable moat for neobanks is everything except the cheapest, most commoditised surface.