Value Accrual in the Crypto Infra Stack

Aug 24, 2022

Does value accrue at the protocol layer or the application layer? Value creation is not the same as value capture. The prevailing thesis has been one of Fat Protocols (2016) / Thin Applications (2020). A few thoughts:

Application-specific blockchains. Forgo the base layer and build a vertical solution. We saw dYdX move to its own app chain. Like successful applications built on a PaaS, the largest customers churn out of your platform and build their own vertical solutions. Building your own application-specific blockchain still seems like migrating off Heroku to build your own Heroku – when you really need your own bespoke infrastructure. This I why I believe dYdX and on-chain exchanges won't be able to compete with fully off-chain ones.

Layer 2 networks and other batching solutions. Provide batching and other optimizations that the main network refuses to. Optimizations gravitate toward the lowest layer they can logically operate on. Layer 2's aren't sufficiently abstracting the lower layer. Two questions: Why wouldn't the underlying networks adopt the same optimizations if successful? Why use the settlement layer when the majority of the activity happens on the scaling layer?

Liquidity at the application layer. Liquidity serves as a network effect for marketplaces and a moat against hard forks. Deploying a new Uniswap contract isn't the same as using the official one. The underlying protocols may be too fat (e.g., Ethereum as both an execution and consensus protocol). Applications rarely need the decentralization guarantees provided by the consensus layer but create new experiences through the execution layer (e.g., Seaport/OpenSea, Coinbase Wallet, dYdX).