Loops Housezeat

NEAR HACKATHON
Team
1 member- DOOwner
Dolapo
Overview
The problem: NEAR Intents earns about $21M a year, while NEAR mints about $140M of new tokens a year. Most theses measure that revenue. An investor needs to know whether it can last.
What I built: Zeat, a reproducible research dashboard that tests whether Intents' revenue is defensible. Three causal tests, each against a synthetic control built from competitors (THORChain, Chainflip, Maya, SOCKET, 0x, all-DEX volume), with the rules committed to git before any result existed:
- Pricing power (Feb 23 fee switch): PASS. Volume ran +16% vs control while NEAR's cut of fees rose from 7.8% to 13.8%.
- Trust (Oct 1 exploit): FAIL. Volume trailed the control by 21% in weeks 4-5 (placebo p = 0.05); the page shows why this result is fragile.
- Price premium: FAIL. Intents charged 14.0 vs 18.4 bps for rivals; it wins on price, not premium. Verdict: 1 of 3. Intents has proven it can charge, not yet that its revenue is defensible.
How NEAR fits in: NEAR Intents is the subject, and NEAR is the evidence. Zeat reads 3,346 NEAR blocks directly (no keys) to check the aggregator data against the chain, measure solver concentration from token_diff events (top 3 accounts: 60% of solver activity) and read exact NEAR issuance from block headers. Intents revenue covers 15% of issuance value. Raising NEAR's cut alone can't close that gap; growing volume can, and volume is what trust puts at risk.
Anyone can re-run it: one command (python3 -m zeat build), Python only, under a minute, rebuilds every number byte for byte. A daily job re-tests it on fresh data. Live: https://dydex.github.io/zeat/