Loops HouseEpoch Ledger

NEAR HACKATHON
Links
Team
1 member- LIOwner
Liam Lakevold
Overview
Problem: NEAR pays validators by minting about 2.5% of supply a year. A holder needs to know what returns that dilution, and how far those returns are from covering it.
What we built: Epoch Ledger, a long-term investment thesis for NEAR that is generated from data on every run. It walks epochs from block headers (each header points to the last block of the previous epoch) and reads total_supply at the first and last block of every epoch, which gives minted and burnt NEAR exactly, with no indexer. On top of that ledger it adds NEAR Intents volume, fees and buyback revenue (DefiLlama, labelled third party), price and activity (NearBlocks), and validator concentration and liquid staking yield (RPC view calls).
Result, 90 days to 2026-11-11: issuance is 2.49% a year. Gas burn offsets 0.98% of it. The Intents buyback offsets 17.7%. Holders are diluted a net 2.03% a year. The thesis reduces to one number: Intents needs about $755M a day, 5.6 times today, for burn plus buyback to equal issuance. The third-party buyback series is checked on chain: the NEAR balance of the three revenue accounts rose 0.97 times what the series implies.
Judged on evidence: 15 claims, each a sentence with a test, and python run.py exits 1 if one stops holding. 6 kill criteria state what would retire the thesis. The evidence against is in the thesis itself. A model with four sliders runs in the dashboard.
Reproducible: standard-library Python, no keys, no packages. python run.py --offline rebuilds every output byte for byte from the committed raw data, which has a SHA-256 manifest. python run.py pulls fresh data from mainnet. Hosted dashboard: https://epoch-ledger.pages.dev
How NEAR fits: the method relies on NEAR block header fields (total_supply, epoch ids, chunk balance_burnt), the validators RPC, view calls to linear-protocol.near and meta-pool.near, and the accounts that hold NEAR Intents revenue.