A yield-generating alternative to ENS

At the time of writing, Bitcoin is more than 50% below its all-time high, spot is around $59,300, and the Fear and Greed Index is at 15, which means the market is in extreme fear. Most of the funding and hype has moved to the hot new thing, which right now mostly means AI.

It forces the remaining crypto teams to stop performing ideology and start working toward product-market fit. I believe that there are two use cases for crypto: DeFi and speculation.

These are the only narratives that have generated revenue, but most speculative projects follow the same transient spike pattern: a parabolic pump followed by a near-vertical dump, whether measured by volume, price, users, or ENS registrations. The ENS registration graph below is a good example. DeFi, by contrast, has followed a more organic adoption curve: slower, less explosive, but structurally more persistent because usage is tied to recurring financial activity rather than one-off speculation.

This is why I see DeFi as the most promising use case for crypto. Speculation will always exist, but it is more of a gambling-like side of crypto, and not one I am particularly interested in.

The core utility of Ethereum is that it provides a shared economic substrate. Applications deployed on it can access existing accounts, capital, assets, protocols, and execution environments. A new product can bootstrap from a network that already exists instead of constructing one from scratch.

A lot of projects use decentralization as a narrative even when the actual product is just another business model with crypto aesthetics. ENS is one of the cleaner examples.

ENS improved Ethereum by turning complex addresses into human-readable names. But the economic model still looks like the old domain world. Users pay recurring registration fees. If they do not renew, the name expires. The relationship between a user and a name is fundamentally rental-based.

That may be acceptable for a namespace business, but it is a strange foundation for durable internet identity. Beyond ENS's network effects, NFT speculation, and historical importance, I think the ENS economic model is inferior for many name ownership use cases.

The registration data has been trending downward for a long time. That raises the uncomfortable question of whether much of the growth was speculation rather than durable demand. ENS captured a real early category, but it did not sustain the growth curve. At this point the project feels almost ancient: important historically, but less useful as a guide for what Ethereum naming should become next.

ETH Bonded Names is a small experiment around a different question: can Ethereum names use refundable, yield-generating bonds instead of recurring rent?

DeFi and Speculation

In the intro I separated crypto into two categories that have actually produced revenue: DeFi and speculation. If ENS unfortunately falls closer to the speculation bucket, then the useful question is not just how to make a better name registry. It is how to design a naming system around DeFi instead.

A DeFi-native naming system should connect naming to capital, yield, liquidity, and durable economic activity. The name should still be useful as identity, but the economic model should look more like a financial position than a subscription or a tradeable meme.

That is the intuition behind ETH Bonded Names. Instead of asking users to pay rent forever, the registry asks them to bond ETH. The capital remains locked while the name is held, can earn yield underneath, and can be withdrawn when the user releases the name. The naming layer becomes an application of DeFi primitives rather than another speculative namespace cycle, with yield shared between the user and the protocol.

Example implementation

The ETH Bonded Names repo is a proof-of-concept implementation of the idea. It is not a full ENS replacement, but it does implement the core economic model and exposes the key data points.

  • Both the user and the protocol benefit from registration. In the example model, yield is split 50/50 between the registrant and the protocol.
  • Because the locked ETH generates revenue for the protocol, there is no expiration date and no renewal path. The name remains assigned to the registrant while the bonded position remains locked.
  • The example implementation uses Lido, but the design does not need to be permanently tied to Lido. Other DeFi protocols, like Aave, could also be used as the ETH-denominated yield layer.
  • Normal names require a small bond. Short names require much larger bonds. That keeps the casual case cheap while making scarce names meaningfully expensive to occupy.
  • The UX is simpler than ENS in a few ways.
    • Registration does not need a reserve flow because mempool frontrunning of usernames is a much less realistic real-world issue now than it was in the early Ethereum naming design space.
    • The registry maps directly to human-readable strings instead of ENS-style label hashes, which makes names easier to inspect.
    • The namespace is intentionally narrow: lowercase letters, numbers, and hyphens; no Unicode; no spaces; no uppercase normalization problem; no double hyphens; no names longer than fifteen characters. It is not a universal naming system. It is a small registry with fewer places for ambiguity to hide.

You can read the example implementation in the eth-bonded-names repo.