Stani Kulechov

Stani Kulechov has rebuilt the same market four times, from ETHLend's peer-to-peer matching in 2017 to Aave V4's central hub extending credit lines to independent spoke markets. Every version trades the same two quantities against each other: the cost of coordinating liquidity, and the risk of holding many assets in one pool.¹

Kulechov is a Finnish entrepreneur, based in Switzerland, who holds a law degree from the University of Helsinki and is the founder and chief executive of Aave, one of the largest lending protocols built on Ethereum. He founded the project as ETHLend in 2017, then rebranded it to Aave and rebuilt it around pooled liquidity in 2020.

Kulechov's central intellectual contribution is a framework he calls liquidity coordination cost: the idea that every design for an onchain lending market trades off the cost of matching borrowers and lenders against the risk created when many assets share exposure inside the same pool.¹ Peer-to-peer markets like the original ETHLend minimize shared risk but carry high coordination costs, since every loan has to be individually matched. Pooling many assets into one market, the design Aave moved to as its V3 version, sharply lowers coordination costs but aggregates all of that risk into a single profile, so that one bad collateral asset can threaten the entire pool.

His answer, laid out in Aave's fourth version, is a Hub and Spoke architecture: a central hub holds liquidity and extends credit lines to separate spoke markets, each of which sets its own risk parameters, so that several hubs, such as Aave's Prime, Core, and Plus markets, can run different risk profiles at the same time without fragmenting the underlying liquidity.¹ He frames this as the modular resolution to a tradeoff with no single correct answer, and contrasts it with Morpho, a rival protocol built around permissionless, isolated markets that put curation in the hands of the market rather than a managing hub.² What the architecture amounts to is an explicit taxonomy for a problem most competitors solve one product at a time.