NEAR Thesis Lab
Thesis: NEAR's usage has moved from headcount to throughput. Value settled through NEAR Intents is growing while raw activity counts shrink, and since February 2026 part of that flow is captured for NEAR holders through buybacks. The capture is real but, today, far too small to offset new supply. Every figure below is recomputed in your browser from public data.
1. Value and headcount moved in opposite directions
How this is computed
Daily rows from NearBlocks/v1/charts joined with DefiLlama NEAR Intents daily volume, aggregated to complete ISO weeks (Monday to Sunday) starting with the first full week of Intents data. Growth compares the mean of the first four weeks with the mean of the last four. Code: src/analysis.js (dailyRows, weekly) and src/thesis.js.
2. Is Intents volume a leading indicator?
How this is computed
Week-over-week log changes (so trends do not create fake correlation). Cross-correlation at lags −8…+8 weeks; a bar at lag +L pairs Intents volume in week t with the other series in week t+L. The dashed lines are the ±1.96/√n band for a single lag. Because eight positive lags are inspected, significance comes from a permutation test of the maximum lead correlation (2,000 shuffles, fixed seed 42, so the p-value is reproducible). Code:crossCorrelation, leadPermutationTest.
3. Does the captured revenue offset new supply?
How this is computed
Net new supply is the change in NearBlockstotal_supply (issuance minus burnt fees) over the 365 days to the end date. Captured revenue is DefiLlama's NEAR Intents dailyRevenue (NEAR/wNEAR swept into NEAR's revenue wallets; used for open-market buybacks since 2026-02-23). Coverage = annualised 30-day revenue ÷ (net new NEAR × price). Break-even price = revenue ÷ net new NEAR. Monthly issuance is valued at that month's average price. Code: supplyEconomics, valueCapture.