An evidence-graded index of NEAR on-chain markets.
The case for NEAR, tested daily
3 mixed, 1 not supported, 2 not enough data. Six claims, each tested against a threshold written down before the data was read.
Updated 17h ago · Snapshot 2026-11-14 · Thresholds v1 (DRAFT, not locked)
Six claims
Each rail shows where the measured value sits against its two pre-registered thresholds. Open a claim for the evidence, sources and a command to re-run it.
H1. NEAR network usage holds up independently of the NEAR price.
Mixedρ −0.35, outside the ±0.30 band Supported needs; ρ ≥ 0.60 would mean usage tracks price.
Measured ρ −0.35needs |ρ| < 0.30Grade B180 days
Show evidenceHide evidenceMetric and value
Spearman rho of 7d-MA daily active accounts vs 7d-MA NEAR price, 180d. ρ −0.35
Evidence for H1 Measure Value Source Spearman ρ, 7d-MA active accounts vs 7d-MA NEAR price (180d)-0.35 snapshot ↗ Daily active accounts, 30d avg vs prior 30d+2.4% (supports the claim) snapshot ↗ Thresholds
Supported needs |ρ| < 0.3 and 30-day active-account growth ≥ 0%; ρ ≥ 0.6 means usage is tracking price (Not supported).
As written in thresholds/v1.yaml: Supported “abs(rho) < 0.30 AND 30d avg active accounts growth >= 0%”; Not supported “rho >= 0.60”; otherwise Mixed.
What would change this verdict
|ρ| must fall by 0.055 to reach Supported.
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-05-18 to 2026-11-13 (180 days) · sample 180
Sources
- series.network snapshot ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H1H2. NEAR Intents, not the on-chain AMM, is where NEAR’s capital now sits.
Not enough data yetNeeds 30 days of data; has 2 (7%).
Measured 2 of 30 daysneeds 30Grade B2 days
Show evidenceHide evidenceMetric and value
intents_custody_usd / rhea_clean_tvl_usd, daily, last 30d. 2 of 30 days
Evidence for H2 Measure Value Source Assets held by intents.near (latest day)$101.60M snapshot ↗ RHEA TVL (verified pools) (latest day)$38.91M snapshot ↗ Custody ÷ TVL (verified pools) (latest day)2.61× (supports the claim) snapshot ↗ Days in window with both values (from same-day snapshots)2 of 30 snapshot ↗ Thresholds
Supported when custody ÷ TVL (verified pools) ≥ 2× on ≥ 80% of the last 30 days; Not supported when it is < 1× on > 50% of days.
As written in thresholds/v1.yaml: Supported “ratio >= 2.0 on >= 80% of days”; Not supported “ratio < 1.0 on > 50% of days”; otherwise Mixed.
What would change this verdict
2 of 30 daily snapshots so far; needs ≥ 30 (window fills one day per snapshot).
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-11-13 to 2026-11-14 (2 days) · sample 2
Sources
- intents.custody_usd snapshot ↗
- rhea.clean_tvl_usd snapshot ↗
- series.h2_daily snapshot ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H2H3. NEAR’s markets can absorb institutional-size orders beyond NEAR itself.
Mixed2 tokens absorb $100k at ≤ 1% impact; Supported needs 5, 1 or fewer is Not supported.
Measured 2 tokensneeds 5Grade B1 day
Show evidenceHide evidenceMetric and value
count of non-NEAR, non-wNEAR tokens with $100k price impact <= 1.0% (RHEA router + Intents dry-run quotes, best of both). 2 tokens
Evidence for H3 Measure Value Source Tokens quoted at $100k (excluding NEAR/wNEAR)22 snapshot ↗ With a measured $100k quote on at least one route21 snapshot ↗ Best-route impact ≤ 1.0%2 (ZEC, NEAR20) snapshot ↗ $100k quotes that failed (error / no route / paused / not configured)33 snapshot ↗ Thresholds
Supported with ≥ 5 non-NEAR tokens at ≤ 1.0% best-route impact for $100k; Not supported with ≤ 1.
As written in thresholds/v1.yaml: Supported “>= 5 tokens”; Not supported “<= 1 token”; otherwise Mixed.
What would change this verdict
Needs 3 more token(s) to flip to Supported.
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-11-14 to 2026-11-14 (1 day) · sample 21
Sources
- depth snapshot ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H3H4. Trading activity on RHEA is mostly organic.
Not supported41% of volume from verified pools, below the 80% needed and under the 60% floor.
Measured 41% from verified poolsneeds 80%Grade B30 days
Show evidenceHide evidenceMetric and value
volume from non-rejected, non-flagged pools / total reported volume, 30d. 41% from verified pools
Evidence for H4 Measure Value Source Denominator: volume across all 4312 live pools (2103 traded), 30d, rejected and flagged included$229.50M snapshot ↗ Of which from 471 pools we didn’t count$841.2K (0.37%) snapshot ↗ Of which from 4 wash-flagged pools$135.38M (59.0%) (counts against the claim) snapshot ↗ Flagged: USDC/wNEAR (DCL) · turnover 7.47×/day · 52.6% of volume vs 1.38% of TVL (verified pools)$120.63M (counts against the claim) snapshot ↗ Flagged: wNEAR/ZEC (AMM) · turnover 2.01×/day$12.37M (counts against the claim) snapshot ↗ Flagged: QTC/wNEAR (DCL) · turnover 2.61×/day$1.75M (counts against the claim) snapshot ↗ Flagged: FRAX/USDC (STABLE) · turnover 4.38×/day$618.3K (counts against the claim) snapshot ↗ Volume from verified pools (not rejected, not flagged)40.6% (counts against the claim) snapshot ↗ Same share without the wash flags (pools we didn’t count only removed)99.6% snapshot ↗ Concentration: top 3 pools’ share of volume67.8% snapshot ↗ Denominator check: all-pool volume on the last full UTC day ÷ RHEA’s reported rolling 24h volume91% ($9.13M vs $10.05M) snapshot ↗ Thresholds
Supported at ≥ 80% of 30-day volume from pools that are neither rejected nor wash-flagged, out of volume across ALL live pools; Not supported below 60%. Wash flags: mean daily volume > 2× pool TVL, or ≥ 10% of all volume while > 5× its share of TVL (verified pools).
As written in thresholds/v1.yaml: Supported “>= 80%”; Not supported “< 60%”; otherwise Mixed.
What would change this verdict
Needs +19.4% to reach Mixed.
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-10-15 to 2026-11-13 (30 days) · sample 2,103
Sources
- series.pool_volume_all snapshot ↗
- rhea.volume_24h_usd snapshot ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H4H5. Fresh capital is entering NEAR’s markets right now.
Not enough data yetNeeds 80% of both the 7- and 30-day windows observed; has 0.5% (the lower of the two).
Measured 0.5% observedneeds 80%Grade B29 days
Show evidenceHide evidenceMetric and value
net Intents flow (deposits - withdrawals), 7d and 30d, plus 30d TVL change. 0.5% observed
Evidence for H5 Measure Value Source Observed net Intents flow, 30d, day-weighted (268 windows, 26,800 events, 254 min observed)$50.4K (supports the claim) snapshot ↗ Observed net Intents flow, 7d, day-weighted (84 windows, 8,400 events)−$113.4K (counts against the claim) snapshot ↗ Coverage: hours of the 30d window actually observed (needs ≥ 80%)0.59% (4.2 of 720 h) (counts against the claim) snapshot ↗ Coverage: 7d window (needs ≥ 80%)0.50% (0.8 of 168 h) (counts against the claim) snapshot ↗ RHEA AMM TVL, 30-day change (data-service daily series)-23.4% (counts against the claim) snapshot ↗ Thresholds
Needs ≥ 80% of the 7d and 30d windows observed before any verdict. Then: Supported when observed 30d and 7d net flows are both positive and 30d TVL change is positive; Not supported when both net flows are negative.
As written in thresholds/v1.yaml: Supported “30d net > 0 AND 7d net > 0 AND 30d clean TVL change > 0”; Not supported “30d net < 0 AND 7d net < 0”; otherwise Mixed.
What would change this verdict
Coverage 0.59% of the 30-day window (4.2 of 720 h) and 0.50% of the 7-day window were observed; needs ≥ 80% of each. Nothing is extrapolated..
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-10-15 to 2026-11-14 (29 days) · sample 26,800
Sources
- series.flow_buckets snapshot ↗
- series.rhea_tvl snapshot ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H5H6. NEAR Intents recovered from the 2026-10-01 exploit without lasting damage.
Mixed84% of baseline on day 30 (the lower of custody and transactions), below the 90% needed.
Measured 84% of baselineneeds 90%Grade B38 days
Show evidenceHide evidenceMetric and value
Intents custody and daily intents.near txns vs pre-incident 7d median. 84% of baseline
Evidence for H6 Measure Value Source intents.near custody (constant prices), 7d pre-incident median$117.65M snapshot ↗ Custody on day 30 (2026-10-31) vs baseline84% snapshot ↗ Daily intents.near txns, 7d pre-incident median94,896 snapshot ↗ Txns on day 30 (2026-10-31) vs baseline103% (supports the claim) snapshot ↗ Incident date (reported; 4 sources, exploit activity 2026-09-29 – 2026-10-01)2026-10-01 src ↗ Thresholds
Supported when custody and daily txns are both ≥ 90% of the pre-incident 7d median on day 30; Not supported when either is < 70%.
As written in thresholds/v1.yaml: Supported “both >= 90% of baseline within 30d of incident_date”; Not supported “either < 70% of baseline at day 30”; otherwise Mixed.
What would change this verdict
Lowest ratio needs +6% to flip to Supported.
Grade Bread from an indexer or public API (RHEA indexer, NearBlocks, FastNEAR) and recomputed by PRIX. Window and sample
2026-09-24 to 2026-10-31 (38 days) · sample 16
Incident sources
Incident date 2026-10-01 (as reported). Reported loss $3.8M to $3.8M; exploit activity 2026-09-29T16:14:00Z to 2026-10-01T08:04:00Z.
- NEAR Intents (@near_intents) on X (official, 2026-10-01)
- Alex Shevchenko (NEAR Intents GM), postmortem on X (official, 2026-10-16)
- CoinDesk (press, 2026-10-01)
- The Block (press, 2026-10-01)
Sources
- series.custody_constant snapshot ↗
- series.intents_txns snapshot ↗
- x.com src ↗
Snapshot hash (sha256)
df9a7f1c26b137126191b7304081bef4a72d699560246bb0484139fef7174174Re-run this claim
pnpm verify --claim H6
Key numbers
What PRIX refused to count: 470 pools, $686.17M claimed. See details
Every figure links to its source.
Missing data reads “Not reported”, never an estimate. The verdicts come from one hashed snapshot that anyone can rebuild with three commands.
Re-run it yourself