Loops HouseSettlemark

NEAR HACKATHON
Links
Team
1 member- HEOwner
Henry Sam Marfo
Overview
Settlemark: NEAR is becoming a settlement layer. Its token has not caught up.
Settlemark is reproducible research on where onchain money settles. Note 01 tests one investment thesis on NEAR with public data, and gives the evidence against as much room as the evidence for.
The claim has two halves, tested separately. First, NEAR Intents usage grows independently of the NEAR price. Second, holders capture little of that flow.
Evidence for: weekly Intents volume rose from $0.8M to $859.7M (first versus last 4 weeks) while NEAR fell from $5.30 to $4.38. Intents went from 0.003% to 1.46% of all onchain DEX volume. Weekly changes in Intents volume and NEAR price correlate at minus 0.05 (90% bootstrap interval minus 0.21 to plus 0.12), neither leads the other, and it is not Bitcoin beta.
Evidence against: Intents kept $9.4M of revenue on $32.97B settled in a year, about 2.8 basis points. NEAR's market cap is 477x that revenue. Daily active accounts and stablecoins on NEAR both fell. Verdict: the use is real, but NEAR is a long term buy only if value capture improves. Watch revenue, not price.
Five falsifiers are checked in code on every rebuild. Today 0 of 5 are met.
How it is built: one command (npm run rebuild) refetches nine datasets from DefiLlama, NearBlocks and CoinGecko with retries and caching, validates each against a schema, recomputes every number, redraws six charts in light and dark, and rewrites THESIS.md and the live page. No number is typed by hand: every CSV row stores its source URL and fetch time, and results.json maps each number to its CSV. 37 tests cover the statistics, schemas, HTTP layer and a byte identical rebuild. A BigQuery cross check on the NEAR public dataset is included. MIT licensed. My other entry is Pocket (github.com/henrysammarfo/pocket).
Live: https://near-settlement-thesis.vercel.app Code: https://github.com/henrysammarfo/near-settlement-thesis