NEAR Against Its Peers
For each dollar of fees paid on-chain, does the market value NEAR more or less than its peers? And which perimeter that includes NEAR Intents can be defended?
Abstract. Counting the Intents revenue its ecosystem captures, NEAR is valued at 0.33 times the peer median, below its typical peer; like for like, on base-layer fees alone, it is valued at 3.07 times, higher than every peer. On the captured-revenue reading, the Intents revenue behind it held up over the last window, but holders do not yet provably capture that revenue. The investment case rests on the treasury burning or locking what it buys back; until then the strict holder reading is BASE. For a holder today, the reading that applies is the like-for-like one: on base-layer fees NEAR is valued at 3.07 times the peer median, higher than every peer. The long-term case turns on a published burn or lock policy for the buyback account (C4) and Intents net revenue staying above $0.87M per 90-day window (C2).
| Reading | NEAR multiple | Peer median | Ratio to the peer median |
|---|---|---|---|
| Counting Intents revenue (BASE+NET) | 441x | 1,326x | 0.33 |
| Like for like, base-layer fees only (BASE) | 4,072x | 1,326x | 3.07 |
Peers are measured on base-layer fees only. The first number (BASE+NET, 441x) also counts the application revenue NEAR captures through Intents, which no peer gets for its own applications; the second (BASE, 4,072x) is the like-for-like comparison. A ratio below one means NEAR is valued below its typical peer. C1 fires above k = 1.00.
Verdicts. C1 valuation holds, C2 durability holds, C3 capture rate holds, C4 holder accrual fires.
The falsification thresholds (k, x, y, z) were frozen after these numbers were visible; the rationale for each one sits next to it in config/panel.toml. Trailing window: 365 days. The full text is THESIS.md; the code and data are on GitHub. Data from DefiLlama. Powered by CoinGecko.
Key findings
Five questions a skeptical investor would ask, each answered from the data and linked to its evidence.
Like for like, how is NEAR valued?
Peers are measured on base-layer fees only, while BASE+NET adds the application revenue NEAR captures through Intents; no peer gets the same credit for its own applications. Like for like, on base-layer fees alone, NEAR trades at 4,072x, 3.07 times the peer median: higher than every peer. See the evidence
Counting the revenue Intents brings, how is NEAR valued?
Counting the Intents revenue its ecosystem captures (BASE+NET), NEAR trades at 441x, 0.33 times the peer median of 1,326x: below k = 1.00, so C1 holds. Holding NEAR's market cap and the peer median fixed, C1 would fire only if trailing-year Intents net revenue fell below $2.39M, a drop of 75%. Counting only Intents revenue booked since the regime break, annualized, the ratio is 0.28 (377x): C1 holds on that reading too. See the evidence
Is the Intents revenue durable?
Intents net revenue rose 199%, from $1.74M over the prior 90-day window to $5.19M over the last one. C2 fires only on a drop larger than 50%: it holds. Over 30-day blocks, Intents net revenue went from $0.50M (18 Mar 2026 to 16 Apr 2026) to a peak of $2.28M (14 Sep 2026 to 13 Oct 2026); the latest block, $1.84M, is below that peak. The share of Intents fees NEAR keeps rose by a relative 41% (+7.97 points), from 19.56% over the prior 90-day window to 27.52% over the last one. C3 fires only on a drop larger than 25%: it holds. See the evidence
Do NEAR holders capture that revenue?
No burn or lock policy is published for the buyback account, which holds about 2,250,000 NEAR; about 12,000 NEAR have already left it. The bought-back NEAR can be spent again, so it is not provably held for holders: C4 fires, and the strict holder reading stays BASE (4,072x, 3.07 times the peer median). C4 has two tests: the policy test fires, while the outflow-share test alone passes (0.53% of the holdings, against z = 1%). Since the regime break (23 Feb 2026), Intents holders revenue equals net revenue every day: all of it funds buybacks. BASE+HOLDERS therefore differs from BASE+NET only by the $1.35M of net revenue booked before the break, which BASE+NET counts and BASE+HOLDERS does not. It is not a separate test of the buyback assumption. See the evidence
How robust is the comparison?
Across the panel itself and 7 variants (adding either candidate chain or dropping any one peer), the NEAR BASE+NET ratio moves from 0.32 to 0.51, and the C1 verdict stays the same in every one. The verdict depends on the perimeter, not on the panel: on BASE, like for like, the ratio is 3.07, above k, so C1 would fire on that reading. Aptos ranks 104 on CoinGecko on 12 Nov 2026, outside the rule's top 30; it stays in the panel, which was frozen with the thresholds. Without it, the NEAR BASE+NET ratio is 0.32 and C1 holds. See the evidence
Method
Daily fees and revenue come from DefiLlama and market caps from CoinGecko, in one snapshot committed with the code. Each chain's multiple is its market cap divided by its trailing-year fees, read like a price-to-sales ratio; NEAR's multiple is then divided by the median of peers chosen by a written rule, so below one NEAR is valued below its typical peer. Most of NEAR's recent revenue comes from NEAR Intents, an application, so NEAR is measured under four perimeters, from base-layer fees only (BASE) to all Intents fees (BASE+GROSS); the thesis defends BASE+NET. Four criteria, C1 to C4, with thresholds frozen in a tagged commit, state what would prove it wrong. Run make judge-demo to rebuild every number offline.
Evidence
Each figure states what it shows in its caption, then the chart or table and its method. The sources follow the evidence.
Falsification scorecard
| Criterion, claim and formula | Statistic | Threshold | Verdict | Distance to firing | If it fires |
|---|---|---|---|---|---|
| C1 Valuation NEAR BASE+NET is no more expensive than the peer median. NEAR BASE+NET multiple / median(peer multiples) > k (trailing 365 days) | 0.33 | k = 1.00 | holds | Intents net revenue would have to fall 75% | Even counting Intents, NEAR is not valued below its peers: the valuation case fails. |
| C2 Durability Intents net revenue does not collapse from one window to the next. Intents net revenue, last 90 days < (1 - x) * prior 90 days (data after regime_break) | +199% | x = 0.50 | holds | Last-window net revenue below $0.87M | Intents revenue is not durable: the captured-revenue reading loses its base. |
| C3 Capture rate NEAR keeps its share of Intents fees. net / gross, last 90 days < (1 - y) * prior 90 days (data after regime_break) | +41% | y = 0.25 | holds | Last-window share below 14.67% | NEAR's cut of Intents fees is being competed away. |
| C4 Holder accrual Bought-back NEAR provably stays out of circulation. outflows from the buyback account > z% of holdings, or no burn/lock policy | 0.53% | z = 1% | fires | Policy test fires; outflow-share test holds | Holders do not provably capture the revenue: the strict holder reading is BASE. |
Method. Each criterion is a tripwire: if it fires, that part of the thesis is wrong. The thresholds were frozen in one tagged commit (thresholds-frozen) after these numbers were visible and before the thesis was written.
Market cap per dollar of yearly fees
NEAR, by perimeterNEAR BASE+GROSS, upper bound onlyPeerPeer median, 1,326x
Method. Multiple = market cap on 12 Nov 2026 / fees summed over the trailing 365 days. Log scale. NEAR appears under its four perimeters; the peer median uses the peers only, never NEAR or the candidate chains. Peers are measured on base-layer fees only: only NEAR BASE is like for like, while BASE+NET also counts NEAR's application revenue from Intents.
Which NEAR fees count
| Perimeter | Fees counted | Multiple | Ratio to peer median | Meaning |
|---|---|---|---|---|
| BASE | $1.15M | 4,072x | 3.07 | Base-layer gas fees only. The strict holder reading. |
| BASE+NET | $10.65M | 441x | 0.33 | Plus the Intents revenue the NEAR ecosystem keeps. Captured revenue: the defended perimeter, assuming the treasury acts for holders. |
| BASE+HOLDERS | $9.30M | 505x | 0.38 | Plus the Intents revenue DefiLlama books as holders revenue: the part that funds NEAR buybacks. |
| BASE+GROSS | $55.04M | 85x | 0.06 | Plus all Intents fees, most of which never reach NEAR. Upper bound only. |
Method. Every perimeter is a trailing 365-day sum ending on 12 Nov 2026, divided into NEAR's market cap on that day.
Does the result depend on the panel?
Method. Ratio = NEAR BASE+NET multiple / median of the peer multiples. '+ X' adds a candidate chain to the peers (Tron 90x, Cardano 7,016x); '- X' drops one peer. C1 fires to the right of k.
NEAR Intents revenue, and the share NEAR keeps
Net revenue kept by NEAR (labelled)Gross Intents fees
| Window | Dates | Net revenue | Gross fees | Net / gross |
|---|---|---|---|---|
| Prior window | 17 May 2026 to 14 Aug 2026 | $1.74M | $8.88M | 19.56% |
| Last window | 15 Aug 2026 to 12 Nov 2026 | $5.19M | $18.86M | 27.52% |
Method. Whole windows ending on 12 Nov 2026, none before 23 Feb 2026: revenue is booked daily only since then, so earlier data is not comparable. C2 compares the last window's net revenue with the prior one's; C3 compares their net / gross shares, the part of Intents fees NEAR keeps.
Do the buybacks reach holders?
- Account
- buybacks.multisignature.near
- Holdings
- about 2,250,000 NEAR (rounded)
- Outflows
- about 12,000 NEAR (cumulative, from the first buyback to 11 Nov 2026)
- Outflow share
- 0.53%
- Burn or lock policy
- none published
- Checked on
- 11 Nov 2026
Method. Outflows are transfers out of the buyback account since the first buyback. A burn or lock policy would make the bought-back NEAR provably unspendable.
Dilution: circulating supply growth
| Chain | Period | Circulating supply change | Annualized |
|---|---|---|---|
| NEAR | 14 Nov 2025 to 12 Nov 2026 | +2.3% | +2.3% |
| Solana | 14 Nov 2025 to 12 Nov 2026 | +6.5% | +6.6% |
| Ethereum | 14 Nov 2025 to 12 Nov 2026 | +0.8% | +0.8% |
| Avalanche | 14 Nov 2025 to 12 Nov 2026 | +3.8% | +3.8% |
| Sui | 14 Nov 2025 to 12 Nov 2026 | +12.2% | +12.3% |
| Aptos | 14 Nov 2025 to 12 Nov 2026 | +20.7% | +20.8% |
Method. Circulating supply = market cap / price, compared at the start and the end of the period, then annualized.
Sources
- Fees, revenue and market caps by chain (DefiLlama and CoinGecko): NEAR: fees API, methodology, market cap; Solana: fees API, methodology, market cap; Ethereum: fees API, methodology, market cap; Avalanche: fees API, methodology, market cap; Sui: fees API, methodology, market cap; Aptos: fees API, methodology, market cap. Used in Multiples · Panel · Dilution.
- NEAR Intents (DefiLlama): revenue, fees, holders revenue, methodology. Used in Perimeters · Intents revenue.
- Buyback account (nearblocks): on-chain account. Used in Buybacks.
- Panel rule, formulas and frozen thresholds (repository): config/panel.toml. Used in Scorecard · Panel.