An ordinary transaction executes at the first opportunity and still costs gas if it fails. A conditional one waits until chain state matches the stated conditions, and otherwise never lands in a block. We walk through the mechanism using Base's Cobalt upgrade, and its limits.
Abracadabra is winding down: it has about $900,000 of collateral against $21 million of bad debt. Here is where bad debt comes from, why stablecoin holders absorb it and how a protocol liquidation works.
On September 30 the token of Concrete launched; the protocol holds $1.3 billion in vaults built on ERC-4626. Here is what you receive for a deposit and why the share price moves.
On September 25 the SEC staff said a token buyback announcement does not make a token a security if the network is functional, and on September 28 added a "no central party" condition. How buybacks work at Aave and Uniswap, how a treasury differs from a burn, and why a buyback does not guarantee a higher price.
ZetaChain holders voted 99.4% in favour of closing their own chain and moving ZETA to Solana 1:1. How token migrations work, what happened to Kadena after its team left, and the holder's risks: deadlines, exchanges, bridged versions and phishing.
About $10 billion in DeFi lending sits in vaults where a curator makes decisions for depositors. We look at the roles inside such a vault, which curator decisions affect depositors' money, and what protects them — timelocks and exit rights.
On September 24 around $2.3 million of unbacked wrapped MTRG was minted through the Meter Passport bridge and sold on a DEX; the price fell about 72% in a day. We explain why an unbacked mint hurts every holder and liquidity pool, not just the bridge.
Coinbase launched fixed-rate bitcoin-backed loans via Morpho Midnight. We explain how variable rates follow pool utilization, where a fixed rate comes from without a formula, and why fixing the rate does not protect against liquidation, maturity risk or bad debt.
Over 30 days Uniswap users paid about $210 million in fees, but the protocol kept $16 million. Morpho generated $19 million in fees and zero revenue. We explain the difference between fees and revenue, what a take rate is, and why these numbers need context.
In one week in mid-September, Solana DEXs processed 208 million trades versus 189–190 million on the NYSE. But trade count is not market size: the average Solana trade is worth tens of dollars. What this metric measures and which indicators tell you more.
On September 17, 2026, NSTR's price was pushed up about 8,000-fold and roughly $3.5M was borrowed against it on Nostra. How illiquid collateral turns into bad debt, why TWAP only partly helps, and which limits cap the damage.
A fee-on-transfer token withholds part of an amount inside its own logic during a buy, sale, or ordinary transfer. A router may therefore send the quoted input while the pool actually receives less. A normal quote that ignores this behavior becomes inaccurate, increases slippage, or causes the transaction to revert.
