Is NEAR's protocol revenue real, durable, or bought with incentives, and does it offset the 2.4% a year of new NEAR?
- NEAR earns a fee on swaps through NEAR Intents and uses it to buy NEAR: that revenue is real, rebuilt from the chain to within 1%.
- It more than doubled from August to October while an airdrop rewarded confidential swaps, which explain a fifth to most of the rise; it has eased since.
- It buys back between a seventh and a fifth of the new NEAR the network issues, depending on the window, so supply still grows about 2% a year.
- On 1Click, most of what users pay goes to the apps that bring them; NEAR keeps about 6%.
- Nothing bought is burned or locked: it is a policy, not a rule, and the price is set by far larger flows.
Rebuilt from the chain, reconciled across sources, tested against hypotheses written before the data.
Built at NEAR Hackathon 2026. Every figure below can be re-run.
00 / Live reading
For every 100 NEAR issued
NEAR's revenue is real: it reconciles to the chain within 1%. It grew sharply from September, as confidential swaps took off under an airdrop campaign, and has eased since. And at today's market cap it buys back only a fraction of what is issued. Recomputed from live data each time the page loads.
NEAR bought back with protocol revenue, last 30 days
21/100
| Measure | Value | Context |
|---|---|---|
| Revenue captured, 30d | $1.86M | 543k NEAR at $3.42 |
| Take rate | 4.9 bps | of $3.77B Intents volume |
| Market cap / revenue | 205x | break-even below 41× |
| NEAR issued, 30d | 2.6M NEAR | 2.43% a year, on-chain |
What this audit found
Key findings
Verdict for an investor (a thesis on public data, not personal advice): NEAR's fee capture does not support its price today. Market cap is 205x captured revenue; supply stops growing below 41x. Nothing bought back is burned or locked, and 47% of it sits in a fund whose policy one Admin vote can change (a normal DAO set-up, but a policy, not a rule). We turn constructive if, for a quarter, two of these hold: a take of 5 bps or more after the airdrop, revenue covering half of issuance, a burn or lock written in code. Rule or choice →
Follow $100 swapped through NEAR Intents over the 30 days to 2026-11-12: $0.049 reaches NEAR's revenue wallets, while $0.238 of new NEAR is issued over the same days. Revenue covers 21% of issuance; supply stops growing at 100%. (Repository snapshot on GitHub.)
Follow $10 to $1M on 1Click: what a swap costs, and who gets it
Measured on the 1Click swaps of the last 30 days in the band from each size to three times it (the $1M row: $300k–1M). Falls with size to about $3k, flat to $100k, then rises as the solvers' spread widens, while NEAR's slice shrinks. Rebuild: make swap-cost
01 / Simulator
What would it take?
Move the inputs and see how many NEAR the protocol buys back for every 100 issued. It starts from today's live values. The identity is exact: buybacks match issuance when market cap is below about 41 times annual captured revenue. A governance vote on 2 November approved cutting issuance from 2.5% to 1.6% a year over 24 months: tick the box to see the fully phased-in case, where the threshold rises to about 64 times.
02 / Kill criteria
What would prove this wrong
Three criteria were written before the analysis (HYPOTHESES.md). Each invalidates the thesis if it holds for 30 consecutive days. Two are checked live on every visit; one needs swap-level data.
03 / Supply
Buybacks against issuance
NEAR captured by the revenue wallets, as a share of the NEAR issued that month (total supply read from the archival RPC). The red line is break-even: above it, NEAR supply would shrink.
Source: Archival RPC total supply (net issuance) and NEAR inflows to the 3 revenue wallets (NearBlocks). Rebuild: make supply · src/analysis/supply.py
Show the numbers
Checked against NEAR's own numbers
NEAR published a revenue dashboard (revenue.near.org) in November. Its series are compared here with this audit's rebuild, which was made before it existed. Both say the same thing: about one NEAR bought back for every six issued.
Show the numbers
04 / Pricing
NEAR's cut doubled in September
Captured revenue in basis points of volume. Volume did not drive the rise. It moved with confidential swaps, which went from 13% to 41% of volume between August and October. In the first full week of November the take fell back to 3.9 bps while volume held: on the NEAR reaching the wallets, on DefiLlama (3.8) and, by about 30%, on NEAR's own revenue series. On 1Click, NEAR also repriced (a 20 bps Trust Wallet ZEC flow, higher fees on swaps without an integrator).
All Intents
Captured revenue, bps of volume, 14-day rolling. Shows what NEAR keeps per dollar swapped. Does not show why: a break search on the weekly take dates the jump to the week ending 20 Sep 2026, the airdrop snapshot's week.
Source: Revenue reaching the 3 wallets ÷ NEAR Intents volume (DefiLlama), 14-day rolling. Rebuild: make revenue · src/analysis/revenue_wallets.py
Show the numbers
1Click channel
NEAR's take, bps of volume, weekly (Feb–Jul: sample weeks)
Source: 1Click swaps from the NEAR Intents Explorer API (2.7M collected, partner key): NEAR's fee ÷ volume, weekly. Rebuild: make swaps · src/analysis/swaps.py
Show the numbers
Confidential swaps: where the jump came from, and what pays for it
From NEAR's own Dune and revenue dashboards. Through October, a weekly fit gave NEAR about 11 bps on confidential volume against 1 bp on public volume; the same fit on week-to-week changes gives about 5.5 bps, so the premium is real but its size is uncertain (Dune's fee table says 7.4). Fitted wallet by wallet, the buyback multisig gets about 7.6 bps, close to Dune's figure (REPORT section 10). That growth followed the NEAR@3.33 airdrop; the timing fits, the cause is not proven. The fit then missed its first unseen week (2–8 November): it predicted $0.5–0.9M of revenue and $0.36M arrived. From 3 November the source labels almost all volume confidential, while 1Click swaps still visible in the public API ran above $20M in half a day. So the label is not a price list: what earned NEAR more was most likely its own front-end, where the confidential users were.
Source: NEAR's Dune dashboard, query 5179085 (public vs confidential volume), read as text into data/raw/dune/. Rebuild: make confidential · src/analysis/confidential.py
Source: Dune query 8644298, then revenue.near.org's confidential TVL series. Rebuild: make confidential official · src/analysis/official.py
Show the numbers
05 / Distribution
Who owns the customer
On 1Click, distribution partners keep most of the fee pool. SwapKit takes 75% of all app fees but sends NEAR almost nothing; NEAR's own revenue comes mostly from users who arrive without an integrator. In November NEAR's 1Click take fell from 4.6 to 2.9 bps, mostly because one burst ended (Trust Wallet, $27M of ZEC at 20 bps on 23–28 Oct); revenue from users without an integrator held at about $5k a day. The volume is broad, not wash-traded (2026-10-14 to 2026-11-12, repository snapshot on GitHub): 235k sending addresses, the top 10 at 19%, round trips at 2.9% of volume. Re-sort to see who matters for whom.
Source: 1Click swaps, 11 Oct–10 Nov 2026: app fees and NEAR's fee by referral (integrator). Rebuild: make swaps · src/analysis/swaps.py
Show the numbers
What a 1Click swap costs the user, by size, and who gets it
Basis points of what the user sends that is gone on delivery, 1Click swaps of the last 30 days, by swap size.
Falls with size to about $3k (fixed network costs), flat to $100k, then rises as the solvers' spread widens, while NEAR's slice shrinks. Source: dollar value sent and delivered per 1Click swap (Explorer API), split by the fees in each record; daily sums committed. Rebuild: make swap-cost · src/analysis/swap_cost.py
Show the numbers
What NEAR earns on, by asset
1Click swaps, 11 Oct–10 Nov 2026, labelled by the volatile leg (BTC→USDT counts as BTC). Stablecoin-to-stablecoin swaps are 30% of volume but pay NEAR 1.3 bps.
Source: 1Click swaps, 11 Oct–10 Nov 2026, labelled by the volatile leg. Rebuild: make swaps · src/analysis/swaps.py
Show the numbers
06 / Valuation
Where it goes by end-2027
Shows the range of outcomes if the past repeats in new orders. Does not show new shocks; tested on 2026, it was too optimistic, so read these as upper bounds. 10,000 simulated paths per scenario. Volume and price are drawn together, in 30-day blocks, from history. Pick a scenario: in the base case the median path stays far below break-even; only several-fold volume growth reaches it. The issuance cut approved on 2 November helps only a little in 2027: its 24-month ramp starts no earlier than 2027-02-14.
Source: Monte Carlo, 10,000 paths, volume and price drawn jointly in 30-day blocks from history. Rebuild: make montecarlo · src/analysis/montecarlo.py
Show the numbers
The path to yes: the volume that would stop supply growing
Daily NEAR Intents volume at which buybacks would match issuance, month by month to December 2027, at today's NEAR price, with the issuance cut's ramp from February 2027. At today's take (4.9 bps) it needs $516.5M a day by December 2027, 4.1x today's volume: growth of about 11% a month for 13 months. Shows how far the yes case is. Does not show a forecast: price, take and volume will all move.
Lines: actual volume (monthly average), needed at today's take, needed at 8 bps. Rebuild: make deeper
What break-even takes, at today's price
Share of issuance offset for a daily Intents volume and take rate. Outlined cells reach break-even. Hover a cell to trace its row and column.
07 / Context
Peers, rivals, holders
NEAR Intents led its category on volume from May to October (THORChain is ahead in November so far), but NEAR is priced far above tokens whose holders share in revenue. And for a holder who does not stake, issuance still dilutes about 2% a year.
Market cap / annualised holders revenue
Tokens whose holders receive protocol revenue (DefiLlama, CoinGecko). Log scale.
Source: DefiLlama holders revenue (30 days ×365/30) and CoinGecko market caps. Rebuild: make comps · src/analysis/comps.py
Show the numbers
NEAR Intents' share of cross-chain swap volume
Against THORChain, Chainflip and Maya, monthly (DefiLlama)
Source: DefiLlama cross-chain swap volume: NEAR Intents, THORChain, Chainflip, Maya. Rebuild: make comps · src/analysis/comps.py
Show the numbers
What moved NEAR: seven events against a placebo
Median day, two weeks after against two weeks before each event, ranked against the same split at every ordinary date of the last year. Bold: outside the usual range (p ≤ 0.10).
The airdrop snapshot moved revenue and the price; the supply cuts did not lift the price; the exploit moved neither. Sources: DefiLlama, the wallet rebuild, CoinGecko. Rebuild: make events · src/analysis/events.py
Will rivals compete the take away? The quote race
Source: dry quotes from the public APIs of 1Click, Relay, LI.FI and Across for the same swap, asked within the same minute; every answer is kept in data/raw/quotes/. Positive = 1Click pays the user more, fees included. Quotes are not fills, and origin-chain gas is left out. Rebuild: make quotes · src/analysis/quotes.py
The holder's view
Daily prices since Nov 2024; supply, stake and buybacks at today's pace
08 / Scorecard
Hypotheses, as registered
Each was written down before the test. Reasoning, dead ends and corrections: RESEARCH_LOG.md. The full thesis: REPORT.md.
09 / Checks
Every finding is a test
Each headline finding has an explicit pass condition, re-checked by make verify on every build; the documents are rendered from the same numbers, so they cannot drift. Below: the findings, public claims about NEAR set against the data, and six predictions frozen on 11 Nov 2026 (sha256 committed, Bitcoin timestamp) that resolve from 15 Nov.
Findings
Dated predictions
Frozen 2026-11-11; status as of the data date. "So far" is a reading on the data to date, not a resolution
Public claims, checked
What NEAR, analysts and the press say about revenue, buybacks, issuance and usage, tested on this audit's data at each claim's date
Anchored on NEAR
sha256 of the predictions, the raw-snapshot manifest and the dashboard bundle, written to SocialDB on NEAR mainnet under near-revenue-audit.near/audit and on testnet under near-revenue-audit.testnet/audit. Check with make anchor-check (a view call, no key)
Verify it in your browser
No key, no wallet, no install, about 10 MB: your browser reads the hashes anchored on NEAR mainnet and testnet, hashes the raw files it was computed from, rebuilds NEAR's revenue from the raw receipts and compares it with this audit and with NEAR's own dashboard. Nothing here trusts this page's server.