wut.meme / thesis / near

NEAR “Volume is not capture”

Agents settle through NEAR Intents and the fees buy back NEAR, if agent flow does not land on the cheapest fee tier.

1. 1
   Consensus

   Agents will transact on-chain and the settlement layer wins. Its metric is agent payment counts, which x402, Bloomberg and a16z all publish. NEAR rose 164 % in the 30 days to 2026-09-26 on this story, to a $6.46 B market cap.

2. 2
   Mechanism

   Agents settle through NEAR Intents, fees buy back NEAR, so agent volume becomes NEAR demand.

3. 3
   Unmeasured link

   The fee tier agent flow lands on. Agent payments are small stablecoin transfers (x402 averages 30 cents), and Intents charges stablecoin and same-asset routes 1 bp, its lowest tier. Nobody publishes the share of Intents volume on that tier.

4. 4
   Proxy

   30-day Intents revenue captured for NEAR holders, from DefiLlama (it reconciles with revenue.near.org): $1.64 M on 2026-09-26, 3.6 bps of $4.53 B volume. Secondary: captured revenue per dollar of volume, which is the margin question in one number.

5. 5
   Evidence source

   The fee schedule and the on-chain buyback sweep, not the base case. The front end charges 20 bps and the average charged is 14 bps, but 3.6 bps reaches the token. A base case that puts 30 bps on NEAR's side overstates capture about eight times.

6. 6
   Forced move

   If the mechanism holds, captured revenue grows with volume at a stable or rising bps. If agent flow is the growth, bps falls toward 1 while volume rises, and the token sees little of it.

7. 7
   Name

   Volume is not capture.

8. 8
   Kill condition

   30-day captured Intents revenue below $3.35 M (double the publish-day figure) on the last DefiLlama point on or after 2027-03-28.

Positioning and belief, not advice. Live numbers refresh hourly; the baseline and kill condition are fixed at publish.