subsidized state · an investment thesis on near, checked against the chain

block · · every number is a link to its query

if chain abstraction killed gas burn, what does the near token actually capture?

so far, almost nothing. against every 100 near of new supply last year, was burned as gas, piled up in two fee vaults, and was locked under one contract’s state, topped up by a treasury bot.

  1. 01 the lock
  2. 02 the sawtooth
  3. 03 who pays
  4. 04 the gap
  5. 05 verdict

the ground is the real trace: intents.near’s spare balance over days. each red step is about near arriving from intents_refiller.near.

01 — the lock

the lock

every megabyte a contract stores locks near. intents.near held gb of state a year ago and holds gb now. it has near locked under it. that is of everything the contract owns.

+near locked in days

locked, not burned. delete the state and the near comes back.

02 — the sawtooth

the ground it walks on

intents.near keeps almost nothing spare: between and near on top of locked. new state eats the spare. when it falls to about , near arrives. times in days, about every hours.

every riser is one transfer from intents_refiller.near. tap one for its receipt.

03 — who pays

who pays

the refills come from one account, intents_refiller.near. it is funded by a dao treasury that has approved near to it in proposals. in the last days, % of top-ups came this way. users pay transaction fees. they do not post the collateral.

04 — the gap

the gap

in a year the supply grew by near (). state collateral at intents.near locked . gas burned . the lock is of issuance; the burn is .

today · gbbreak-even · gb

break-even means the lock equals issuance. it would still be locked, not burned. reset

05 — verdict

verdict

today the near token captures a deposit, not a toll: near a year locked under state and burned as gas, against issued. two fee vaults gained in the same year. that near is held, not burned.

what to watch

  • the fee vaults fall from and near, and total supply falls by the same amount: a burn, not a transfer.
  • a different payer is identified by receipt while the lock keeps growing. today % of top-ups come from intents_refiller.near.
  • the lock passes 1% of issuance. today it is .
  • gas burn passes 2% of issuance. today it is .

the case for

  • the lock is real and protocol-enforced: near under one contract, readable with one rpc call.
  • it grew at every one of checkpoints across the year, from to .
  • it scales with stored state, which is what agents and accounts produce, not with gas price.
  • fee vaults already hold near that governance could burn.

the case against

  • it is too small: of net issuance.
  • it is locked, not burned. deleting state returns it.
  • it is paid by a treasury: of top-ups in days came from one funded account.
  • the fee vaults are balances, not burns. the buyback vault moves on of keys.
  • this measures one contract. it is not a census of chain state.

what would change my mind

  • state growth near gb a year instead of ; a tenth of that would already matter.
  • another payer, identified by receipt, posts the collateral while the lock keeps growing. refills stopping on their own would not show that.
  • vault balances are burned and total supply falls by the same amount.
  • any of the three tests below fails on a rebuild.

three tests, thresholds fixed before the rebuild script existed · tests/claims.json

also true at block

  • bridge-mng.near holds the on omft.near. it has ( with full access) and no contract.
  • v1.signer needs of its participants to sign.
  • omft.near tokens: across contracts, of it zec (prices from a third-party api).
  • buybacks.multisignature.near holds near, more than days ago. fefundsadmin.sputnik-dao.near holds . both are balances held at this block.

prior work: other entries in this hackathon, including toll road, near toll audit and stakeout, already cover intents fees against issuance. this adds the storage-collateral ledger, the refill trace and who pays for it.