Valuo

The long-term case, made falsifiable

4 of 8 conditions for NEAR's long-term case are on track today

The long-term case for NEAR at $4.34B: 4 of 8 measured conditions are on track. It holds if the market keeps an L1 multiple (135x revenue needed against a 2,498x peer median) while Intents grows into it; it fails if volume growth stays near 28% a year against the 133% the price implies.

4 on track4 behind0 against

ConditionRequiredLiveTrendStatus

Intents volume compounds fast enough to grow into the price
Holding every other input at base, the market FDV needs this annual public-volume growth for three years; live is trailing-12-month growth of 30-day volume.
Required133% a year
Live28% a year
Trend+93% 30-day volume vs 90 days earlier
Behind
Holder revenue (buybacks) scales to support the price at the model's multiple
Year-3 annualised holder revenue the market FDV needs at the model's terminal multiple, against today's run-rate.
Required$156.1M a year
Live$21.88M a year
Trend+300% 30-day holder revenue vs 90 days earlier
Behind
The market keeps valuing NEAR like an L1, not like a fee-only app
FDV-to-holder-revenue multiple the market FDV implies, against the peer median for L1s (SOL, ETH, HYPE). This is the only single lever that closes the gap inside measured evidence.
Required135×
Live2,498×
Trendno series
On track
1Click keeps a pricing edge where it competes (cross-VM routes)
Median headroom over the best competitor on destination units, live quote race. Native BTC -2.97 bps and EVM L2 -1.02 bps: 1Click trails there.
Requiredabove 0
Live+0.98 bps
Trendno series
On track
Confidential (private) flow keeps growing
Net 30-day inflow into the Confidential Intents treasury, read on-chain; treasury TVL $132.6M. Swap volume on the private chain itself is not public, so this is flow, not volume.
Requiredabove 0
Live+$15.16M
Trend+392% confidential TVL vs 90 days earlier
On track
Fees retire more NEAR than issuance creates
Daily Intents volume at which buybacks, gas burn and state lock offset issuance: $386.0M/day under the 1.6% glidepath, $608.0M/day at today's 2.5%. Sinks cover 22% of issuance now.
Required$386.0M a day
Live$125.8M a day
Trendno series
Behind
Governance cuts issuance from 2.5% to 1.6%
Protocol max inflation read from EXPERIMENTAL_protocol_config; the cut halves the volume needed for net deflation.
Required1.6% cap
Live2.5%
Trendno series
Behind
Solver competition stays broad enough to keep spreads tight
Solver HHI from sampled execute_intents fills (90% CI 2,898-4,772); above the bear-case threshold means a few solvers set prices.
Requiredat most 3,500
Live3,197
Trendno series
On track

Method and sources

Conditions the long-term bull case needs, graded against this run's live data. Requirements come from market_implied (one lever at a time), supply_physics and bear_case thresholds; nothing is re-estimated here.

Requirements for volume growth, holder revenue and the multiple are the reverse-valuation solves (one input moved at a time until the model's expected-path FDV equals the market FDV), so each says what that condition alone must deliver. The deflation requirement is the breakeven daily volume under the 1.6% glidepath; solver competition uses the bear-case HHI threshold. Trends compare the last 30 days with the 30 days ending 90 days earlier in DefiLlama's daily series, and confidential TVL with its value 90 days earlier in the NEAR-reported history. Fields: bull_case.conditions[].

Investment thesis. What the market is pricing, against a 36-month cash-flow model and 10,000 seeded paths

NEAR's $4.34B market value needs 133% annual volume growth or a 56.3 bps take rate on Intents cash flow alone; only a 135× multiple closes the gap inside the evidence

The cash-flow model values NEAR at $1.98B on its expected path and $644.7M at the Monte Carlo median; the $4.34B market FDV is 2.2× the expected path. Peer L1s trade at a median 2,498× FDV to revenue and NEAR at 186×, so the market prices NEAR on an L1 multiple that carries a monetary premium, not on Intents cash flow at the model's 60× cap.

Market FDV

$4.34B

Fully diluted, at spot

CoinGecko

Model, expected path

$1.98B

Market is 2.2× this

Model, Monte Carlo median

$644.7M

Market is 6.7× this

Levers that close the gap inside evidence

1 of 6

5 reachable at any value; one input moved at a time

What the market is pricing

Each input is moved alone, all others at the snapshot base, until the model's expected-path FDV equals the market FDV. The required value is then set against the comparator this run measured.

Lever	Model base	Required for market FDV	Evidence this run	Required ÷ evidence	Status
Volume growth Expected annual public-volume growth (3-year path)	72% a year	133% a year	28% a year trailing-12-month NEAR Intents 30-day volume growth DefiLlama	4.8×	Outside evidence
Justifying $4.34B needs 133% expected annual public-volume growth for 3 years (base 72%); trailing-12-month Intents volume growth was 28%.
Public take Starting gross public take rate	15.3 bps	56.3 bps	3.00 bps live competitive fee ceiling (median best-competitor friction, cross-VM cells) 1Click	19×	Outside evidence
Justifying $4.34B needs a starting gross public take of 56.3 bps (decaying to the 0.98 bps floor at 3%/month) versus 15.27 bps today; the live competitive fee ceiling is 3.00 bps.
Terminal multiple Terminal FDV / annual holder-revenue multiple	60×	135×	2,498× peer median FDV / annualised revenue ex-NEAR (unclipped) CoinGecko	0.1×	Inside evidence
Justifying $4.34B needs a 135x terminal FDV/holder-revenue multiple; the live peer median is 2,498x (HYPE 81x, SOL 2,498x, ETH 3,076x) and the model caps it at 60x.
Discount rate Annual discount rate	25%	None in range	10% low end of the registry discount-rate range (10%-40%)	n/a	Unreachable
Even a 0% discount rate gives $3.82B, short of $4.34B: the gap is not a discount-rate question.
Confidential share Confidential share of volume at month 36	40%	75%	17% on-chain estimate of the confidential share (Confidential Intents Treasury 24h boundary flow / 24h volume) NearStats	4.5×	Outside evidence
Justifying $4.34B needs the confidential share to reach 75% by month 36; the on-chain estimate today is 17% (band 14%-17%).
Year-3 holder revenue Year-3 annualised holder revenue at the base multiple (volumes scaled jointly)	$68.14M a year	$156.1M a year	$21.88M a year current annualised holder revenue (30-day window x 365.25/30) DefiLlama	7.1×	Outside evidence
Justifying $4.34B at the 60x base multiple needs $156M of annualised holder revenue in year 3 (2.3x the base path's $68M), versus $22M today (93%/yr compounded).

4 of 5 levers with a measured comparator need values outside the evidence: volume growth 133% a year against 28% a year (4.8×), public take 56.3 bps against 3.00 bps (19×), confidential share 75% against 17% (4.5×) and year-3 holder revenue $156.1M a year against $21.88M a year (7.1×).
Terminal multiple closes the gap at 135×, inside the peer median FDV / annualised revenue ex-NEAR (unclipped) of 2,498×.
Discount rate cannot close it: not reachable at any discount rate >= 0% (FDV at 0% is $3.82B).

Method and sources

Basis: deterministic_sotp expected path (mean-matching GBM); one input moved at a time, all others at the snapshot base. Solver: scipy.optimize.brentq on [lo, hi] search bounds, xtol = rtol = 1e-12; each solve is checked by re-running the model at the required value (relative error shown below). Horizon 36 months. Fields: market_implied.solves[], market_implied.current_fdv_usd, market_implied.base_fdv_usd, market_implied.base_fdv_mc_p50_usd.

Status: inside evidence when required_vs_evidence.within_evidence is true, outside when false; unreachable when no value in search_bounds reaches the market FDV; a comparator flagged evidence.assumption is not evidence.

public_volume_growth_yoy: search [-99% a year, 1,000% a year], harder direction higher; evidence source: DefiLlama dexs/near-intents daily volume (intents_forensics.fees.daily_series): last 30 days vs the 30 days ending 335 days earlier; check error 1.4e-13.
tau_pub_bps: search [0.00 bps, 200 bps], harder direction higher; evidence source: execution_bench.competitive_fee_ceiling_bps (median over cross_vm cells of the best competitor's friction (raw where available else usd)); check error 4.8e-14.
terminal_multiple: search [0.0×, 10,000×], harder direction higher; evidence source: comps.median_fdv_to_revenue_ex_near (HYPE 81x, SOL 2,498x, ETH 3,076x); check error 2.2e-16.
discount_rate: search [0%, 100%], harder direction lower; evidence source: assumption registry: discount_rate.
confidential_share_target: search [0%, 99%], harder direction higher; evidence source: intents_forensics.volume.confidential_share_estimate (NEAR mainnet RPC, treasury NEP-245 balances); check error 9.2e-15.
holder_revenue_scale: search [$0.00 a year, $1.00K a year], harder direction higher; evidence source: sotp.calibration: alpha_capture x gross_fees_30d_usd (DefiLlama near-intents holders revenue); check error 0.0e+0.

What the cash-flow model values

The model's median value for NEAR is $644.7M, 85.2% below the current $4.34B FDV.

The median is not the base case. Volume follows a mean-matching random walk, so the expected path is worth $1.98B (Monte Carlo mean $1.88B) and the median path reaches 33% of it. Read the P10 to P90 range below, not a single point.

38 labelled assumptions

Median FDV against current

-85.2%

$644.7M median; $4.34B at $3.31 spot

NearStats

Net revenue run-rate

$21.87Ma year

$72.90M gross fees30.0% captured

Annualised from the last 30 complete days

DefiLlama

Daily volume today

$124.4Ma day

Breakeven is 3.10× to 4.89× this

Depending on the issuance regime

DefiLlama

Sink coverage of issuance

21.6%

19,379 NEAR absorbed of 89,757 NEAR issued a day

NEAR RPC

Valuation distribution

Percentiles across the seeded paths. Figures are in ink; only the last column is coloured, by the sign of the median against the current market.

Measure	P10	P25	P50	P75	P90	Current	Median vs current
Implied FDV Discounted, 36-month horizon	$126.5M	$261.2M	$644.7M	$1.64B	$3.92B	$4.34B	-85.2%
Implied price per NEAR	$0.09	$0.19	$0.47	$1.20	$2.87	$3.31	-85.8%
Implied FDV, ex monetary engine Engines 1 and 2 only	$118.3M	$248.3M	$610.3M	$1.54B	$3.67B	$4.34B	-85.9%

Supply crossover by issuance regime

At today's $124.4M of daily volume, buybacks, gas burn and state lock absorb 21.6% of new issuance. Net supply stops growing only once daily volume reaches $608.0M under the status quo and $386.0M under the SVRN glidepath.

Issuance regime	Issuance per day, first to last month	Breakeven daily volume	Multiple of today's volume	Crossover on the expected path	Chance of crossover within 36 months	Median crossover month	Chance supply is shrinking at month 36
Status quo 2.50% inflation cap, flat	89,757 NEAR to 96,537 NEAR	$608.0M $3.25B at the fee ceiling	4.89×	None within 36 months Cumulative basis: none	66.1%	M14	47.2%
SVRN glidepath 2.50% falling to 1.60%	89,757 NEAR to 60,752 NEAR	$386.0M $2.06B at the fee ceiling	3.10×	None within 36 months Cumulative basis: none	73.3%	M14	55.3%

Crossover is the first month in which buybacks, gas burn and state lock absorb all new issuance. The expected path holds price at spot and volume flat; the chances are shares of the seeded paths.

Method and sources

FDV is the present value of holder revenue plus a discounted terminal value, summed over three engines on each path; percentiles are taken over those per-path totals. Fields: sotp.fdv_percentiles, sotp.price_percentiles, sotp.fdv_ex_monetary_percentiles (Engines 1 and 2; Engine 3 is a demand-flow proxy), sotp.current_fdv_usd, sotp.upside_p50_pct. Seed 20261114.

Expected path versus median: deterministic expected-path FDV $1.98B (sotp.calibration.deterministic_base.fdv_total), Monte Carlo mean $1.88B (sotp.calibration.mc_mean_fdv). Public volume at month 36: mean $524.5M a day against a median of $66.17M a day (sotp.calibration.volume_terminal_mean_usd_day, volume_terminal_median_usd_day); monthly volume volatility 0.339 from 11 monthly log returns. Under the mean-matching GBM the median path sits below the expected path by exp(−σ²t/2).

Run-rates: gross fees and net revenue over aligned daily charts, 30 complete UTC days 2026-10-13..2026-11-11, annualised × 365 / 30 (intents_forensics.fees.annualized_gross_fees_30d_usd, intents_forensics.fees.annualized_net_revenue_30d_usd). Net revenue is NEAR swept to its revenue wallets and directed to buybacks.

Today's daily volume: supply_physics.inputs.daily_volume_usd. supply_physics.inputs.daily_volume_usd (measured): intents_forensics.fees.volume_30d_usd: DefiLlama near-intents daily volume summed over the canonical take-rate window (aligned daily charts, 30 complete UTC days 2026-10-13..2026-11-11).

Sink coverage = (buyback + gas burn + state lock) / measured issuance (supply_physics.sink_coverage_ratio). Breakeven V* = (issuance − gas burn − state lock) × price / (gross take × capture) (supply_physics.breakeven_daily_volume_usd; at the competitive fee ceiling, supply_physics.breakeven_daily_volume_usd_at_ceiling). Issuance by month: supply_physics.trajectories[].daily_issuance_start / _end; expected-path crossover supply_physics.trajectories[].crossover_month and cumulative_crossover_month.

Crossover chances: sotp.crossover_month_distribution.<regime>.p_crossover_within_36m, p50_month, p_deflationary_at_horizon. Model note: Path probabilities reflect the dispersion of the volume and price GBMs (a path crosses when holder revenue / price outruns issuance); compare with deterministic_crossover_month (the expected-volume path) and with supply_physics.breakeven_daily_volume_usd before quoting a crossover.

How the thesis is built

Four measured links, each feeding the next. Every figure below is read from the snapshot; the source link on a figure opens its raw payload record.

Step 1: Execution headroom
1Click delivers more than the best competitor on cross-VM routes and less on native BTC and EVM L2. The best competitor's median friction on cross-VM routes is 3.00 bps, which bounds how far rivals can undercut the take rate.
Headroom, cross-VM
ETH USDC to SOL USDC, 3 cells
+0.98 bps1Click
Headroom, native BTC
ETH USDC to native BTC, 3 cells
-2.97 bps1Click
Headroom, EVM L2
Base USDC to Arbitrum USDC, 3 cells
-1.02 bps1Click
Competitive fee ceiling
Median over 3 cross-VM cells
3.00 bps
Quotes returned
Trailing cells are shown as measured
29 of 36
Step 2: Take rate and stickiness
NEAR Intents keeps 16.06 bps of volume before revenue sharing and 4.82 bps after, on $3.73B of 30-day volume. About 16.7% of that volume settles confidentially, estimated on-chain from Confidential Intents Treasury flows (band 13.5% to 16.7%). Fills spread across the equivalent of 3.1 equal-sized solvers.
Gross take
Fees over volume, 30 days
16.06 bpsDefiLlama
Net take, captured by NEAR
30.0% of gross fees
4.82 bpsDefiLlama
Solver concentration (HHI)
90% CI 2,898 to 4,772; 6 solvers in 114 sampled fills; 3.13 effective
3,197NearBlocks
Deposit-source concentration (HHI)
90% CI 1,944 to 4,625; bridges and direct callers, not solvers; 3.27 effective, 50.0% AI-agent-labelled
3,056NearBlocks
Confidential share of volume
On-chain estimate from treasury flows, last 24 h; band 13.5% to 16.7%
Estimate16.7%NEAR RPC
Step 3: Supply sinks against issuance
Buybacks, gas burn and state lock absorb 19,379 NEAR of the 89,757 NEAR issued each day (21.6%). At spot, net supply still grows by 70,377 NEAR a day.
Gross issuance per day
Measured from the last epoch's supply jump
89,757 NEARNEAR RPC
Buyback sink per day
18,099 NEARDefiLlama
Gas burn per day
938 NEARNEAR RPC
State-lock growth per day
Measured on intents.near and aurora
342 NEARNEAR RPC
Net supply change per day
Positive means supply is still rising
+70,377 NEAR
Step 4: Valuation
Valued path by path, the three engines give a median FDV of $644.7M, -85.2% against the current market. The peer terminal multiple of 2,498.2× is clipped to 60.0×, so the multiple is an assumption, not a measurement.
Terminal multiple used
Peer median FDV over revenue, 3 peers
Clipped60.0×CoinGecko
Discount rate
Assumption25%
Public take-rate floor in the model
Mean-reversion target for the public take
0.98 bps
Median FDV against current
$644.7M median
-85.2%

Method and sources

Execution headroom. Benchmark ran 2026-11-13 06:05:11 UTC. Headroom is 1Click output minus the best competitor's output, in destination units for the same input (basis dest_units), as the median over cells where 1Click and at least one competitor quoted. Fields: execution_bench.headroom_by_archetype_bps, execution_bench.headroom_n_by_archetype, execution_bench.competitive_fee_ceiling_bps (median over cross_vm cells of the best competitor's friction (raw where available else usd)), execution_bench.n_quotes_ok.

Take rate and stickiness. Take = fees / volume on one canonical window (aligned daily charts, 30 complete UTC days 2026-10-13..2026-11-11); capture = holders revenue / fees; volume source defillama:dexs/near-intents daily chart summed over the aligned window. The model's public take mean-reverts from its start toward the floor at λ = 0.03 a month (sotp.calibration.tau_floor_bps: max(0.5 bps, min(live public take, live cross-VM headroom)) from the execution benchmark (headroom_by_archetype_bps.cross_vm = 0.98 bps).). Capture source: measured:defillama near-intents dailyHoldersRevenue / fees.

Solver HHI (intents_forensics.solver_concentration.hhi_solver) is computed over 122 attributed fill participations in 114 execute_intents fills (NEP-413 signer payloads); the cartelisation trigger in the bear case reads this index. Deposit-source HHI (intents_forensics.hhi_external) is over deposit callers, not solvers: 175 receipts over 8 pages, AI-agent-labelled share 50.0%, bridge share 25.0%. Confidential share: intents_forensics.volume.confidential_share_30d, status onchain_estimate (on-chain estimate: Confidential Intents Treasury 24h inflow / (24h volume + inflow) = 16.7% (band 13.5%-16.7%); nearstats split implausible). HHI intervals: intents_forensics.solver_concentration.hhi_solver_ci90, intents_forensics.caller_sample.hhi_external_ci90 (bootstrap over the sampled units; within-sample noise only).

Supply sinks. Issuance is measured from the total-supply jump at the epoch boundary (supply_physics.current_daily_issuance_near); buyback = holders revenue / price; gas burn = within-epoch supply drift (supply_physics.current_daily_sinks_near.gas_burn); state lock = archival storage growth × 10⁻⁵ NEAR per byte. Net change: supply_physics.current_net_float_change_near_day.

Valuation. Terminal multiple sotp.calibration.terminal_multiple_used: raw peer-median FDV / revenue ex-NEAR 2,498.2×, guardrails 8× to 60×. Discount rate sotp.discount_rate. The total FDV P50 (sotp.fdv_percentiles.p50) is a percentile of the per-path engine sum, not the sum of engine P50s. Price basis: fdv / (S_0 + gross issuance - buybacks - gas burn - state lock at H, svrn_glidepath trajectory); supply at horizon 1,367,712,771 NEAR (first month whose net float change (issuance - buyback - gas burn - state lock) <= 0).

Evidence scorecard

17 claims, graded by how each was obtained: 3 verified on-chain, 3 verified via live API, 4 derived, 7 not supported. A claim the data does not support stays on the list, marked.

GradeClaimMeasured valueSource

Verified on-chain
NEP storage staking is a measured on-chain supply sink (1 NEAR per 100 KB)
growth 30,999,489 B/day = 310.0 NEAR/day over 31.1 d (archival); a 0.01 NEAR/KB lock is released if state is deleted, so it is a sink while state grows
intents.near storage_usage 13,647,147,373 bytes = 136,471 NEAR locked; 283,182 NEAR across 8 tracked contracts
High confidence
NEAR RPC
Verified on-chain
Realized issuance is measured from epoch-boundary supply jumps, not assumed from the 2.5% cap
second independent jump 25,422.56 NEAR; within-epoch drift -267.07 NEAR/epoch is the net gas burn
25,544.21 NEAR minted at epoch start 224,308,337 x 1,283.4 epochs/yr = 32,783,583 NEAR/yr (2.4996%), -0.02% vs nominal 32,789,360
High confidence
NEAR RPC
Verified via live API
Cross-VM execution headroom: 1Click out-delivers Relay/LI.FI net of fees on ETH USDC → SOL USDC
$500: +1.24 bps vs relay (rank 1/3); $50,000: +0.98 bps vs relay (rank 1/3); $250,000: +0.98 bps vs relay (rank 1/3); no 1Click capacity failures
median headroom +0.98 bps over 3 comparable cells (dest_units); competitor ceiling 3.00 bps
Medium confidence
Not supported
Native BTC corridor: 1Click delivers more BTC per USDC than LI.FI/Relay on ETH USDC → native BTC
clip-dependent: $500: -18.30 bps vs relay (rank 2/3); $50,000: -2.97 bps vs relay (rank 2/3); $250,000: +8.91 bps vs lifi (rank 1/2); LI.FI's best BTC route is often NEAR Intents itself (tool `near`); venue USD marks of BTC differ by tens of bps and are not used
1Click delivers LESS BTC per USDC than the best competitor (relay): median headroom -2.97 bps over 3 cells (dest_units); the claim is not supported by this run
Medium confidence
Verified via live API
EVM-L2 stablecoin routing is commoditized: Across/Relay match or beat 1Click on Base USDC → Arbitrum USDC (honest negative)
$500: -0.99 bps vs across (rank 3/4); $50,000: -1.02 bps vs across (rank 3/4); $250,000: -1.02 bps vs across (rank 3/4); the moat, where it exists, is corridor-specific (cross-VM, native assets), not EVM-L2 stablecoins
median headroom -1.02 bps over 3 cells (dest_units; 1Click trails)
Medium confidence
Derived
Deposit/caller concentration on intents.near is measurable and AI-agent-labeled flow is material
175 unique receipts over 22 s; a point sample of the live receipt stream, not a market-share estimate. AI-agent is a name-pattern classification (*.ai.near, *.agent.near); escrow.ai.near has no deployed contract, so the label is a heuristic. This HHI is over deposit sources (bridge callbacks, direct callers), not solvers
deposit/caller HHI_external 3,056 [90% CI 1,944-4,625] (N_eff 3.27), ex-AI-agent 2,222; AI-agent share 50.0% (>= 10% materiality line), bridge share 25.0%
Low confidence
NearBlocks
Verified via live API
Solver concentration on intents.near is measured from execute_intents fills, not inferred from deposit callers
462 execute_intents receipts decoded; settlement hub 51e8f94d77b5... (146 counterparties) routes most swaps via transfer legs; see intents_forensics.solver_concentration.method. The CI is within-sample (one window of minutes), so it understates long-run uncertainty
HHI_solver 3,197 [90% CI 2,898-4,772] (N_eff 3.13) over 122 fill participations in 114 fills (37 atomic token_diff fills, 77 hub-settled legs), 6 distinct solvers; CI straddles the 3,500 cartel threshold (inconclusive)
Medium confidence
NearBlocks
Derived
Confidential (dark-pool) execution is a material share of 30d NEAR Intents volume
nearstats' split was rejected by the plausibility guard; the share is estimated from the public treasury boundary instead. Biases: same-hour shield/unshield netting and in-shard turnover push it down, solver inventory pushes it up; 24h window
16.7% (band 13.5%-16.7%) estimated on-chain: Confidential Intents Treasury 24h inflow $29.4M / outflow $27.6M against 24h volume $176.4M; >= 10% materiality line
Low confidence
NEAR RPC
Verified on-chain
Confidential TVL backs the dark-pool engine (Engine 2)
Engine 2 is flagged assumption_heavy and its terminal multiple carries the confidential_multiple_haircut (0.7); the confidential take (20 bps) is an assumption
$132.6M TVL read on-chain from the Confidential Intents treasury in intents.near (NEAR-reported via nearstats: $137.9M)
High confidence
NEAR RPC
Derived
Buyback capture alpha = holders revenue / gross fees is measured, so V x tau x alpha / P is a live sink
DefiLlama revenue methodology: NEAR's net captured revenue, measured as NEAR/wNEAR swept into its revenue wallets (reconciles with revenue.near.org and the on-chain buyback). Split by destination wallet into front-end affiliate revenue and Other. Transfers between those wallets are excluded.
alpha 30.0%; gross take 16.06 bps, net take 4.82 bps; buyback sink 18,099 NEAR/day = 21.6% coverage incl. burn and lock
Medium confidence
DefiLlamaDefiLlama
Not supported
Chain Signatures (v1.signer) is a growing second usage engine beyond Intents
sampled estimate: each window covers minutes; bridge callers serve cross-chain transfers that overlap Intents settlement, so the engine is not fully independent of Intents; storage lock 11.6 NEAR is small. sign() attaches 1 yoctoNEAR (median of sampled requests), so the contract takes no material per-request fee; the requester's cost is gas.
15,597 sign requests/day (7d) vs 12,918/day (30d); ln-trend +39% per 30 d [90% CI -9% to +113%], t = 1.28; top caller omni.bridge.near 42% of 382 sampled sign receipts, *.bridge.near callers 66%; gas 0.0021 NEAR per request (inconclusive: growth not shown at t >= 2)
Low confidence
NearBlocks
Not supported
The volume engine (GBM with measured sigma) is calibrated out of sample: P10-P90 covers ~80% of realized 30-day volume
Tests the volume distribution only; the take rate, holder capture, confidential share, multiple and discount rate are not backtested. 3m and 6m forecasts from monthly origins overlap, so their effective sample is smaller than n.
1m 11/20 = 55% inside P10-P90 (binomial p 0.01); 3m 9/17 = 53% inside P10-P90 (binomial p 0.01); 6m 7/14 = 50% inside P10-P90 (binomial p 0.01); 1m median realized / forecast median 1.37x (under-prediction); 1m coverage 20% for origins before 2026-01-16, 90% from it; 20 monthly origins 2025-03-22 to 2026-10-13
Low confidence
Not supported
A block-bootstrap volume engine (fat tails, autocorrelation) is calibrated out of sample where GBM is not
Inside P10-P90 against an 80% target, exact two-sided binomial p. Low power: at n = 20 the test accepts 14 to 19 hits and rejects a 70%-coverage engine only 39% of the time; 3m and 6m forecasts overlap. Block length (Politis-White) and path count were fixed before coverage was computed; nothing is tuned. The valuation still uses the GBM; this row grades a parallel variant and tests the volume distribution only.
block bootstrap 1m 18/20 (90%, p 0.40), 3m 17/17 (100%, p 0.03), 6m 13/14 (93%, p 0.33); GBM 1m 11/20 (55%, p 0.01), 3m 9/17 (53%, p 0.01), 6m 7/14 (50%, p 0.01); bootstrap at the GBM's expected path (shape only) 1m 12/20 (60%, p 0.04), 3m 6/17 (35%, p <0.01), 6m 0/14 (0%, p <0.01); bootstrap band 3.3x the GBM's log width at 1m; 20 monthly origins 2025-03-22 to 2026-10-13
Low confidence
Not supported
NEAR Intents' realized gross take rate is not trending to zero as volume grows
Graded on the trailing-365d trend only: a decline is counted when the Newey-West 90% interval and the non-overlapping-window Student-t interval both sit below zero; DERIVED needs no such decline and a latest rolling take above the model floor. Gross take includes affiliate fees kept by frontends; NEAR's captured share is the net / holder take. Rolling points overlap, so an interval that includes zero is not proof of a floor. The full-history trend is not used (fee ramp from zero in 2025).
rolling 30d gross take 16.06 bps on 2026-11-11 vs 16.45 bps on 2025-11-12; trend -30.3%/yr, 90% CI -42.3% to -15.8% (excludes zero; Newey-West 60d lag, 365 overlapping points); floor 0.5 bps, 0 of 365 trailing points below it; non-overlapping 30d windows (n=13) -26.6%/yr, Student-t 90% CI -43.1% to -5.3%; trailing 180d trend +22.4%/yr, CI -11.8% to +69.9%
Low confidence
DefiLlamaDefiLlamaDefiLlamaDefiLlama
Not supported
1Click quotes institutional size: at least $1M on every benchmarked corridor
Largest size quoted on an eight-rung ladder of dry quotes, one run. A dry 'no liquidity' is solver quoting policy at that moment, not proven depth, and a dry quote is not a fill.
ETH USDC → SOL USDC: 1Click $500K, best competitor LI.FI $5M; ETH USDC → native BTC: 1Click $1M, best competitor LI.FI $1M (routed through NEAR Intents); Base USDC → Arbitrum USDC: 1Click $500K, best competitor LI.FI $5M (1 of 3 corridors with an attempted 1Click rung at or above $1M)
Low confidence
Not supported
NEAR Intents is gaining share of cross-chain routing fees
Fee share, not volume share; adapter fee definitions differ; the pool is the protocols DefiLlama covers. A share trend is DERIVED from live fee series, never VERIFIED.
20.4% of the six-protocol fee pool (30 days to 2026-11-11), rank 2 of 6 behind Relay at 55.4%; share at window start 35.0% (202-day common window); trend -6.59 pp per 30 days, 90% interval -9.50 to -3.67; ex-Relay 45.7%, trend -0.89 pp per 30 days, 90% interval -1.65 to -0.14 (share is falling)
Low confidence
Derived
The model's terminal multiple cap sits inside the observed peer range of FDV/revenue
Rule: DERIVED when the cap is at or below the 75th percentile of the wide set and n_peers >= 8. Here the cap is inside the observed range and at or below its upper quartile. FDV/revenue on 30 days annualised, one snapshot; revenue definitions differ by chain adapter; peers ranked by revenue and fixed before any multiple was computed (inclusion_rules, excluded).
cap 60x vs 15 chain tokens: 2 of 15 trade at or below it (13th percentile); P25 249x, median 531x [bootstrap 90% 274x to 999x], P75 1,748x, range 3.1x to 6,508x; NEAR L1-only 3,907x (87th percentile); without flagged rows: median 600x, n 10
Medium confidence
DefiLlamaCoinGecko

Verified on-chain outranks verified via live API, which outranks derived, which outranks assumption. Grades and confidence are written by the pipeline, not by hand. Source links resolve against manifest.json (229 records, 26 failed).

Three valuation engines

The engine medians sum to $387.1M, 0.60× the $644.7M median of the path totals. Percentiles do not add: each path is valued as a whole and the median is taken over those totals.

Engine	P10	P50	P90	Expected path
Clearinghouse Measured inputs	$28.68M	$205.3M	$1.70B	$966.3M
Confidential execution Assumption-heavy	$21.13M	$171.4M	$1.70B	$864.1M
Monetary premium Assumption-heavy	$1.17M	$10.37M	$163.4M	$147.2M

The expected path is additive by construction: $1.98B in total. The Monte Carlo mean is $1.88B against a current FDV of $4.34B; the mean sits above the median because the paths are right-skewed, so quote percentiles, not the mean.

Method and sources

Engine 1: cross-chain clearinghouse (public intents holder revenue); Engine 2: confidential execution franchise (dark-pool holder revenue); Engine 3: monetary / supply-sink premium (gas burn + state-lock demand flow).

Fields: sotp.engines[].p10, .p50, .p90; total sotp.fdv_percentiles.p50; expected path sotp.calibration.deterministic_base.engine_<key> and fdv_total (deterministic_sotp() at the median inputs); mean sotp.calibration.mc_mean_fdv.

Caveats

The pipeline recorded 15 warnings and 26 failed fetches in this run. Read them before quoting any figure above.

Pipeline warnings (15)

fee_pool
across: 3 malformed, negative or duplicate chart rows skipped
Execution benchmark
ETH USDC to SOL USDC $500: venue USD oracles disagree with the cell's common price by more than the +1.24 bps headroom (1click +0.0 bps, relay +2.3 bps, lifi -43.1 bps); headroom is measured on destination units, not USD legs
Execution benchmark
ETH USDC to SOL USDC $50,000: venue USD oracles disagree with the cell's common price by more than the +0.98 bps headroom (1click +0.0 bps, relay +2.3 bps, lifi -43.1 bps); headroom is measured on destination units, not USD legs
Execution benchmark
ETH USDC to SOL USDC $250,000: venue USD oracles disagree with the cell's common price by more than the +0.98 bps headroom (1click -0.0 bps, relay +2.3 bps, lifi -43.1 bps); headroom is measured on destination units, not USD legs
Execution benchmark
ETH USDC to native BTC $500: venue USD oracles disagree with the cell's common price by more than the -18.30 bps headroom (1click +0.0 bps, relay -3.1 bps, lifi +104.2 bps); headroom is measured on destination units, not USD legs
Execution benchmark
ETH USDC to native BTC $50,000: venue USD oracles disagree with the cell's common price by more than the -2.97 bps headroom (1click +0.0 bps, relay -3.1 bps, lifi +104.2 bps); headroom is measured on destination units, not USD legs
Execution benchmark
ETH USDC to native BTC $250,000: venue USD oracles disagree with the cell's common price by more than the +8.91 bps headroom (1click +0.0 bps, lifi +104.2 bps); headroom is measured on destination units, not USD legs
Execution benchmark
Base USDC to Arbitrum USDC $500: venue USD oracles disagree with the cell's common price by more than the -0.99 bps headroom (1click +0.0 bps, relay +2.0 bps, lifi -11.4 bps); headroom is measured on destination units, not USD legs
Execution benchmark
Base USDC to Arbitrum USDC $50,000: venue USD oracles disagree with the cell's common price by more than the -1.02 bps headroom (1click -0.0 bps, relay +2.0 bps, lifi -11.4 bps); headroom is measured on destination units, not USD legs
Execution benchmark
Base USDC to Arbitrum USDC $250,000: venue USD oracles disagree with the cell's common price by more than the -1.02 bps headroom (1click +0.0 bps, relay +2.0 bps, lifi -11.4 bps); headroom is measured on destination units, not USD legs
Intents forensics
confidential split: nearstats reports confidential 30d $0.10 (0.0000%), rejected (tvl_without_volume: confidential TVL $132.64M but all-time confidential volume $0.35 of $38.55B; stream_ok: nearstats sourceHealth.confidentialStream.status = ok; split_withheld: check confidential-vs-shield (skip) says 'split withheld — NEAR total ≈ public'). Share ESTIMATED ON-CHAIN from the Confidential Intents Treasury: 16.7% (band 13.5%-16.7%; 24h inflow $29.43M, outflow $27.60M, volume $176.44M), replacing the registry assumption 40%
Intents forensics
take-rate window leaves out 2026-11-12: DefiLlama reports revenue above fees for that day, so its fee figure is not final; the window ends the day before
Intents forensics
nearblocks: 25 duplicate receipt(s) seen on overlapping pages were counted once (id-based cursor vs time-ordered rows); unique receipts sampled = 175
Valuation model
confidential share 16.7% is an ON-CHAIN ESTIMATE from Confidential Intents Treasury boundary flows (band 13.5%-16.7%), not nearstats' split (rejected as implausible); 24h window, biases listed in intents_forensics.volume.confidential_share_estimate
Valuation model
peer median FDV/revenue 2498.2x clipped to 60.0x (guardrails [8, 60])

Failed raw fetches (26, recorded, not substituted)

Across quote, ETH USDC to SOL USDC, $50,000
HTTP 400: ROUTE_NOT_ENABLED: Route is not enabled.
Across quote, ETH USDC to SOL USDC, $500
HTTP 400: ROUTE_NOT_ENABLED: Route is not enabled.
Across quote, ETH USDC to SOL USDC, $250,000
HTTP 400: ROUTE_NOT_ENABLED: Route is not enabled.
Relay quote, ETH USDC to native BTC, $250,000
HTTP 400: INSUFFICIENT_LIQUIDITY: Solver has insufficient liquidity for this swap
across/ladder:across:eth_usdc_sol_usdc:1000
HTTP 400: ROUTE_NOT_ENABLED: Route is not enabled.
relay/ladder:relay:eth_usdc_sol_usdc:2500000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $1,749,518 USD. Requested amount exceeds provisioned liquidity.
across/ladder:across:base_usdc_arb_usdc:1000000
HTTP 400: AMOUNT_TOO_HIGH: Amount is higher than available liquidity. Max amount is 609774.717686 USDC.
relay/ladder:relay:eth_usdc_sol_usdc:5000000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $1,749,518 USD. Requested amount exceeds provisioned liquidity.
relay/ladder:relay:eth_usdc_btc:250000
HTTP 400: INSUFFICIENT_LIQUIDITY: Solver has insufficient liquidity for this swap
across/ladder:across:base_usdc_arb_usdc:2500000
HTTP 400: AMOUNT_TOO_HIGH: Amount is higher than available liquidity. Max amount is 609774.717686 USDC.
relay/ladder:relay:eth_usdc_btc:500000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $465,515 USD
relay/ladder:relay:eth_usdc_btc:1000000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $465,515 USD
across/ladder:across:base_usdc_arb_usdc:5000000
HTTP 400: AMOUNT_TOO_HIGH: Amount is higher than available liquidity. Max amount is 609774.717686 USDC.
relay/ladder:relay:base_usdc_arb_usdc:5000000
HTTP 400: AMOUNT_TOO_HIGH: Quote amount exceeds the maximum for this pair
relay/ladder:relay:eth_usdc_btc:2500000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $465,515 USD
relay/ladder:relay:eth_usdc_btc:5000000
HTTP 400: INSUFFICIENT_LIQUIDITY: Amount is higher than the available liquidity. Max amount is $465,515 USD
oneclick/ladder:1click:base_usdc_arb_usdc:1000000
HTTP 400: No liquidity available
oneclick/ladder:1click:eth_usdc_sol_usdc:1000000
HTTP 400: No liquidity available
oneclick/ladder:1click:eth_usdc_btc:2500000
HTTP 400: No liquidity available
oneclick/ladder:1click:base_usdc_arb_usdc:2500000
HTTP 400: No liquidity available
oneclick/ladder:1click:eth_usdc_sol_usdc:2500000
HTTP 400: No liquidity available
oneclick/ladder:1click:eth_usdc_btc:5000000
HTTP 400: No liquidity available
oneclick/ladder:1click:base_usdc_arb_usdc:5000000
HTTP 400: No liquidity available
oneclick/ladder:1click:eth_usdc_sol_usdc:5000000
HTTP 400: No liquidity available
lifi/ladder:lifi:eth_usdc_btc:2500000
HTTP 404: No available quotes for the requested transfer
lifi/ladder:lifi:eth_usdc_btc:5000000
HTTP 404: No available quotes for the requested transfer

38 inputs are labelled as assumptions, each with a default, a range and a justification; they drive the SOTP tab sliders and are never presented as measured. 203 of 229 raw payloads returned successfully across 10 sources.

Method and sources

Warnings: snapshot.warnings and sotp.calibration.warnings, deduplicated, in pipeline order. Failed fetches: provenance_summary.failed, each named by its manifest record. Assumptions: entries in snapshot.assumptions with source: assumption.

Execution quality

1Click out-delivers the best competitor by +0.98 bps on cross-VM routes and trails on native BTC and EVM L2

Each venue receives an identical dry quote for the same input; deliveries are compared in destination units, so no price oracle enters the result. 9 cells (3 routes at 3 trade sizes), run 2026-11-13 06:05 UTC.

Cross-VM headroom, median

+0.98 bps

3 of 3 sizes scored

Native BTC headroom, median

-2.97 bps

3 of 3 sizes scored

EVM L2 headroom, median

-1.02 bps

3 of 3 sizes scored

Competitive fee ceiling

3.00 bps

Best competitor's all-in cost on cross-VM, median of 3

Quotes returned

29 of 36

All venues in 3.6 s

1Click settlement universe

230assets

On 39 chains

1Click

Live quote race

Re-run any cell of the benchmark now, against the same four endpoints, and read it next to the committed result.

Run the race yourself

Sends the same dry quote request to 1Click, Relay, LI.FI and Across from this server, at once. Nothing is signed or sent on-chain; every raw response is hashed before it is parsed.

RouteTrade size

Venues called: 1Click, Relay, LI.FI, Across. Same asset on both sides, so cost is read on raw units with no price oracle.

No race run in this session. Pick a route and a trade size, then run it.

Committed benchmark

The reproducible run stored in the snapshot. Green cells are sizes where 1Click delivers more than every competitor; red cells are sizes where a competitor delivers more.

1Click headroom by route and trade size

Basis points more (green) or less (red) that 1Click delivers than the best competing quote for the same input.

Route	$500	$50,000	$250,000	Median	What the quotes show
Cross-VMUSDC on Ethereum to USDC on Solana	+1.24 bpsvs Relay	+0.98 bpsvs Relay	+0.98 bpsvs Relay	+0.98 bps3 of 3 scored	1Click delivers the most at every scored size ($500, $50,000, $250,000). Across cannot route this pair.
Native BTCUSDC on Ethereum to BTC on Bitcoin	-18.30 bpsvs Relay	-2.97 bpsvs Relay	+8.91 bpsvs LI.FI	-2.97 bps3 of 3 scored	Mixed: 1Click leads at $250,000 and trails at $500, $50,000. Across cannot route this pair.
EVM L2USDC on Base to USDC on Arbitrum	-0.99 bpsvs Across	-1.02 bpsvs Across	-1.02 bpsvs Across	-1.02 bps3 of 3 scored	Across delivers more than 1Click at every scored size.

Every quote in the benchmark

Amount each venue would deliver for the same input, with its all-in cost and settlement estimate. Each figure links to the raw response it was parsed from.

USDC on Ethereum to USDC on Solana

Trade size	1Click	Relay	LI.FI	Across	1Click headroom
$500	499.889031 USDC 2.22 bps cost, 42 s 1Click	499.827036 USDC 3.46 bps cost, 1 s Best competitor Relay	498.750000 USDC 25.00 bps cost, 18.0 min LI.FI	Unsupported route Route is not enabled. Across (failed)	+1.24 bpsvs Relay, rank 1 of 3
$50,000	49,989.890546 USDC 2.02 bps cost, 42 s 1Click	49,984.977036 USDC 3.00 bps cost, 24 s Best competitor Relay	49,875.000000 USDC 25.00 bps cost, 18.0 min LI.FI	Unsupported route Route is not enabled. Across (failed)	+0.98 bpsvs Relay, rank 1 of 3
$250,000	249,949.492626 USDC 2.02 bps cost, 42 s 1Click	249,924.977036 USDC 3.00 bps cost, 24 s Best competitor Relay	249,375.000000 USDC 25.00 bps cost, 18.0 min LI.FI	Unsupported route Route is not enabled. Across (failed)	+0.98 bpsvs Relay, rank 1 of 3

USDC on Ethereum to BTC on Bitcoin

Across not quoted: Across is EVM-only; no Bitcoin route.

Trade size	1Click	Relay	LI.FI	Across	1Click headroom
$500	0.00729133 BTC 70.06 bps cost, 8.2 min 1Click	0.00730470 BTC 51.85 bps cost, 5.9 min Best competitor Relay	0.00728642 BTC 76.75 bps cost, 26.0 min LI.FI	Unsupported route Declared unsupported: Across is EVM-only; no Bitcoin route No HTTP exchange	-18.30 bpsvs Relay, rank 2 of 3
$50,000	0.73095450 BTC 45.25 bps cost, 8.2 min 1Click	0.73117189 BTC 42.29 bps cost, 24 s Best competitor Relay	0.72990539 BTC 59.54 bps cost, 23.2 min Via NEAR Intents LI.FI	Unsupported route Declared unsupported: Across is EVM-only; no Bitcoin route No HTTP exchange	-2.97 bpsvs Relay, rank 2 of 3
$250,000	3.65061116 BTC 56.59 bps cost, 8.2 min 1Click	No liquidity Solver has insufficient liquidity for this swap Relay (failed)	3.64736177 BTC 65.44 bps cost, 23.2 min Via NEAR IntentsBest competitor LI.FI	Unsupported route Declared unsupported: Across is EVM-only; no Bitcoin route No HTTP exchange	+8.91 bpsvs LI.FI, rank 1 of 2

USDC on Base to USDC on Arbitrum

Trade size	1Click	Relay	LI.FI	Across	1Click headroom
$500	499.893705 USDC 2.13 bps cost, 34 s 1Click	499.927691 USDC 1.45 bps cost, 1 s Relay	498.750000 USDC 25.00 bps cost, 18.0 min LI.FI	499.943111 USDC 1.14 bps cost, 2 s Best competitor Across	-0.99 bpsvs Across, rank 3 of 4
$50,000	49,989.895220 USDC 2.02 bps cost, 34 s 1Click	49,994.977686 USDC 1.00 bps cost, 1 s Relay	49,875.000000 USDC 25.00 bps cost, 18.0 min LI.FI	49,994.993111 USDC 1.00 bps cost, 1.8 min Best competitor Across	-1.02 bpsvs Across, rank 3 of 4
$250,000	249,949.497300 USDC 2.02 bps cost, 34 s 1Click	249,974.977686 USDC 1.00 bps cost, 1 s Relay	249,375.000000 USDC 25.00 bps cost, 18.0 min LI.FI	249,974.993111 USDC 1.00 bps cost, 15.0 min Best competitor Across	-1.02 bpsvs Across, rank 3 of 4

Size limits

1Click stops quoting at a smaller size than a competitor on 2 of 3 corridors: ETH USDC to SOL USDC, $500K against LI.FI $5M; Base USDC to Arbitrum USDC, $500K against LI.FI $5M.

Non-binding quotes at 8 sizes from $1K to $5M on each benchmarked route, run 2026-11-13 06:05 UTC. "Largest size quoted" is the top size a venue answered in this run. It is not depth.

Route

Trade size$250K

$1K$5M

3 of 3 venues quote $250K on ETH USDC to SOL USDC; 1Click is cheapest at 2.02 bps

All-in cost in basis points of the amount sent; lower is better. Bars share one scale across sizes on this route (0 to 25.0 bps).

1Click
2.02 bps
Cheapest at this size. Settles in about 42 s
Relay
3.00 bps
Settles in about 24 s
LI.FI
25.00 bps
Settles in about 18.0 min
Across
Does not route this corridor
ROUTE_NOT_ENABLED: Route is not enabled.

1Click on this route: largest size quoted $500K, first size refused $1M. Best competitor: LI.FI $5M.

Cost at every size, ETH USDC to SOL USDC

Venue									Largest size quoted
1Click	1.36	2.03	2.02	2.02	2.02	refused	refused	refused	$500K
Relay	3.23	3.02	3.00	3.00	3.00	3.00	refused	refused	$1M
LI.FI	25.00	25.00	25.00	25.00	25.00	25.00	25.00	25.00	$5M
Across	-	-	-	-	-	-	-	-	does not route

Figures are all-in cost in bps; bold marks the cheapest venue at that size. "refused" is the venue's explicit answer that it will not quote the size, "-" a route the venue does not serve, "not run" a request that did not finish inside the time budget and says nothing about liquidity. ^N marks a LI.FI quote routed through NEAR Intents. "gap" marks a venue that refused a size below its largest size quoted. Select a size in the header to inspect it above.

On this run 1Click quoted at least $1M on 1 of 3 corridors and stopped below a competitor on 2. That cuts against the case that the solver network already carries institutional clips on those routes: fee growth there has to come from retail-size flow until solver inventory deepens. On ETH USDC to native BTC the largest competitor quote (LI.FI $1M) is 1Click execution resold; the largest quote on an independent route was $50K (Relay). Repeated runs in which 1Click quotes the same rungs its competitors do would overturn this reading.

Method and limits

Each cell is one dry quote request sent with the same request body the benchmark uses, at a larger or smaller input. 1Click is called with dry: true, so no deposit address is created and nothing is binding. Cost: (amount in - amount out) / amount in x 1e4: on raw units where both legs are the same stablecoin, else on USD legs valued at one common price per cell (1Click token-list oracle, else the median of the venues' implied prices), as in execution_bench.

This run (2026-11-13 06:05:14 UTC): 59 cells quoted, 21 refused, 0 failed, 0 not attempted, 8 on routes a venue does not serve, of 88; 25.4 s against a 75 s budget. Raw responses are stored and hashed under ladder:<venue>:<corridor_id>:<size_usd>.

A dry 'no liquidity' reflects the solvers' quoting policy at that moment. It is not proven hard depth, and a dry quote is not a fill.
One run is one observation. Solver inventory and venue limits move within minutes; re-run before relying on a rung.
Every figure is the largest size quoted on an eight-rung ladder ($1K to $5M), not depth. Sizes between two rungs were not observed.
Cells marked not attempted did not complete inside the time budget. They carry no information about liquidity.

Method and sources

Headroom. (out_1click - out_best_competitor) / out_best_competitor x 1e4 on destination units for the same raw input; no oracle. Basis dest_units: destination units, the same raw input sent to every venue, no oracle. Positive means 1Click delivers more.

Scored cell. 1Click and at least one competitor returned an ok quote. Archetype medians are over scored cells only (execution_bench.headroom_by_archetype_bps, counts in headroom_n_by_archetype). A route with one scored cell is a single observation.

All-in cost. (input - output) / input in bps: on raw units when both legs are the same stablecoin, else on USD legs valued at oneclick_token_list_oracle. The competitive fee ceiling is median over cross_vm cells of the best competitor's friction (raw where available else usd) (execution_bench.competitive_fee_ceiling_bps).

Run. 2026-11-13 06:05:11 UTC; 29 of 36 quotes ok; wall clock 3.6 s against a 45 s budget, 25 s per-request timeout, concurrency 8. 33 quote records in the manifest (quote:<corridor>:<clip>). Token list ok: true.

Collector summary, verbatim. cross_vm (ETH USDC -> SOL USDC): $500: 1Click delivers 1.24 bps more than relay (rank 1/3, destination units); $50,000: 1Click delivers 0.98 bps more than relay (rank 1/3, destination units); $250,000: 1Click delivers 0.98 bps more than relay (rank 1/3, destination units); unsupported: across (ROUTE_NOT_ENABLED). native_btc (ETH USDC -> native BTC): $500: 1Click trails 18.30 bps behind relay (rank 2/3, destination units); $50,000: 1Click trails 2.97 bps behind relay (rank 2/3, destination units); $250,000: 1Click delivers 8.91 bps more than lifi (rank 1/2, destination units); unsupported: across (Across is EVM-only; no Bitcoin route); relay no_liquidity at $250,000. evm_l2 (Base USDC -> Arbitrum USDC): $500: 1Click trails 0.99 bps behind across (rank 3/4, destination units); $50,000: 1Click trails 1.02 bps behind across (rank 3/4, destination units); $250,000: 1Click trails 1.02 bps behind across (rank 3/4, destination units).

1Click chains. abs, adi, aleo, aptos, arb, avax, base, bch, bera, bsc, btc, cardano, dash, doge, eth, fogo, gnosis, hood, hypercore, ltc, miden, monad, movement, near, op, plasma, pol, prl, qtc, scroll, sol, starknet, stellar, sui, ton, tron, xlayer, xrp, zec

Reproduce. make reproduce or cd quant && uv run python -m valuo.reproduce. Snapshot fields: execution_bench.results[], .corridors[], .competitor_friction_by_venue_bps.

Live race. POST /api/live-benchmark with {corridor_id, clip_usd}; routes from src/lib/corridors.ts (eth_usdc_sol_usdc, eth_usdc_btc, base_usdc_arb_usdc); request payloads per docs/LIVE_ENDPOINT_FACTS.md sections 2 and 3. Venues declared unsupported are re-verified live against the Across chain list.

NEAR Intents collects 20.4% of the six-protocol cross-chain fee pool, rank 2 of 6, and its share is falling

NEAR Intents takes 20.4% of the six-protocol cross-chain fee pool over the 30 days to 2026-11-11, rank 2 of 6 behind Relay at 55.4%; its rolling 30-day share moved from 35.0% to 20.4% over 172 days (low 2.5% on 2026-09-13, high 45.3% on 2026-07-04) and the linear trend (-6.59 pp per 30 days, 90% interval -9.50 to -3.67) is falling, with the interval excluding zero.

Share of 30-day fees, latest

20.4%

30 days to 2026-11-11; rank 2 of 6, behind Relay at 55.4%; 45.7% with Relay excluded

30 days earlier

10.3%

+10.1 ppfrom then to now

30 days to 2026-10-12; 59.1% with Relay excluded

90 days earlier

14.9%

+5.4 ppfrom then to now

30 days to 2026-08-13; 55.5% with Relay excluded

Start of the common window

35.0%

-14.6 ppfrom then to now

30 days to 2026-05-23; 48.8% with Relay excluded

Rolling 30-day share, 173 daily points

NEAR Intents' rolling 30-day fee share went from 35.0% to 20.4%, ranging 2.5% to 45.3%

NEAR Intents' fees over the trailing 30 days divided by the same sum across the pool, in percent. The dashed line drops Relay from the denominator, because its adapter counts integrator app fees and swap costs. The marker is the day Relay's 30-day fees peaked at $205.9M.

30 days to 2026-11-11

Relay collects 55.4% of the pool; NEAR Intents 20.4%

Fees as each DefiLlama adapter defines them, summed over the same 30 days. Pool total $29.40M.

Rank	Protocol	30-day fees	Share of pool	90-day fees
1	Relay	$16.30M	55.4%	$261.9M
2	NEAR Intents	$5.99M	20.4%	$18.92M
3	THORChain THORChain DEX	$3.64M	12.4%	$8.24M
4	Chainflip	$2.92M	9.9%	$6.18M
5	deBridge	$454.0K	1.5%	$1.58M
6	Across	$102.2K	0.3%	$291.9K

Linear trend of the share, percentage points per 30 days

Share trend -6.59 pp per 30 days, 90% interval -9.50 to -3.67

Ordinary least squares on time with a Newey-West 90% interval. A share gain is claimed only when the slope is positive and the interval excludes zero.

Series	Slope	90% interval	R²	Points	Reading
Rolling 30-day share, six-protocol pool	-6.59 pp	-9.50 to -3.67	0.60	173	falling, interval excludes zero
Rolling 30-day share, Relay excluded	-0.89 pp	-1.65 to -0.14	0.15	173	falling, interval excludes zero
Rolling 30-day share, excluding THORChain DEX (zero-fee run in the window) share 23.3% now against 41.6% at the start	-6.86 pp	-9.95 to -3.77	0.60	173	falling, interval excludes zero
Daily share, no overlap (cross-check)	-4.72 pp	-7.19 to -2.25	0.38	202	falling, interval excludes zero

Cuts against the bull case: NEAR Intents is losing share of the fees these routers collect, so fee growth has to come from the pool growing, not from share capture. It is overturned by a positive share trend whose 90% interval excludes zero. Relay collects 2.7x NEAR Intents' fees under its own adapter's definition. NEAR Intents' own 30-day fees went from $2.90M to $5.99M over the window; the pool went from $8.30M to $29.40M. With Relay excluded the share is 45.7% against 48.8% at the start of the window (-0.89 pp per 30 days, 90% interval -1.65 to -0.14).

What each adapter counts as fees (6 DefiLlama adapters)

Relay relay: Platform fees, integrator app fees and swap costs paid by users on Relay bridges and swaps, excluding execution costs (network gas). 203 days from 2026-04-24; payload sha256 6b1b2cf686…; adapter source
NEAR Intents near-intents: Gross fees charged on NEAR Intents swaps - the protocol fee plus the affiliate (distribution) fee each frontend collects, summed across all integrators. 551 days from 2025-05-10; payload sha256 562751003b…; adapter source
THORChain DEX thorchain-dex: Slip-based liquidity (swap) fees paid by users on each chain's THORChain pools, gross outbound fees and slashing fees paid into the Reserve, and affiliate fees charged by interfaces/wallets (affiliate fees are attributed to the THORChain native chain). RUNE block-reward emissions and the native tx fee (counted in the THORChain chain fees adapter) are excluded. 2,041 days from 2021-04-11; payload sha256 c0d2cd3a3d…; adapter source
Chainflip chainflip: Chainflip AMM: Includes Swap, Broker, Ingress, Egress and Network Fees for Buy/Burn Mechanism 1,044 days from 2024-01-04; payload sha256 0d34dd6a62…
deBridge debridge: All fees paid by users for swap and bridge tokens via deBridge. 1,321 days from 2023-04-01; payload sha256 8fe5b22ec5…; adapter source
Across across: Total fees paid by users for bridge txs. 1,289 days from 2023-05-01; payload sha256 dfb2612cd8…; adapter source

Method, window and limits

Source. GET https://api.llama.fi/summary/fees/<slug>?dataType=dailyFees for relay, near-intents, thorchain-dex, chainflip, debridge, across; snapshot section fee_pool_share, fetched 2026-11-13 06:05 UTC.

Window. 2026-04-24 to 2026-11-11, 202 days, starting on Relay's first day. Rule: days from the latest first day to the earliest last settled day on which every included protocol has a point; each series' newest point and any day dated today (UTC) are provisional and left out; a trailing day that is zero or missing for any protocol is cut. Excluded: 2026-11-12 (provisional newest point of near-intents, debridge, thorchain-dex); 2026-11-13 (dated today (UTC), partial; provisional newest point of relay, across, chainflip; not yet published for near-intents, debridge, thorchain-dex).

Trend. OLS of rolling 30-day share (percentage points) on time; slope per 30 days; 90% interval = slope +/- 1.645 x Newey-West standard error with 60 Bartlett lags.

Limit 1. Fee share is not volume share: a router charging more per dollar routed has a larger fee share at the same volume. DefiLlama's bridge-volume API is paid, so no volume-side comparison is made here.

Limit 2. Adapter fee definitions differ, so the pool sums unlike quantities; each adapter's own text is quoted next to its protocol (protocols[].methodology_fees). The ex-Relay share is published because Relay's adapter counts integrator app fees and swap costs.

Limit 3. The common window is 202 days (2026-04-24 to 2026-11-11), set by Relay's first day; the trend describes these months only.

Limit 4. Routers outside DefiLlama's fee coverage (LI.FI and other aggregators) are absent, so the pool is these six protocols, not the cross-chain market.

Limit 5. Rolling 30-day points overlap by 29 days; the interval uses a Newey-West error with 60 lags, and the daily non-overlapping share trend is reported as a cross-check.

Limit 6. THORChain is measured by DefiLlama's thorchain-dex adapter (swap, outbound and affiliate fees); the thorchain slug is the chain's native transaction fee and is not a routing fee.

Limit 7. The share path is not a line: it ranged from 2.5% (2026-09-13) to 45.3% (2026-07-04) and the linear trend explains 60% of its variance. Read the slope as the average drift over the window, not as a forecast.

Limit 8. Relay's 30-day fees peaked at $205.95M on 2026-09-22, 13.4x its window median of $15.34M. NEAR Intents' share around that date moves with that surge in the denominator, not with a change in its own fees. Whether the surge is user-paid routing fees or an adapter artefact cannot be settled from this endpoint.

Limit 9. THORChain DEX reports zero fees on 38 days of the window (longest run 37 days, 2026-05-16 to 2026-06-21). The zeros are kept as reported; if they are an adapter gap, the other protocols' shares are overstated for rolling windows that include those days. See the robustness share excluding it.

Collector warnings. across: 3 malformed, negative or duplicate chart rows skipped.

Take-rate history

Gross take compressed over the trailing year

The rolling 30-day rate is 16.1 bps on 2026-11-11 against 16.4 bps a year earlier, while average daily volume went from $85M to $124M; the trailing-365d log-linear trend is -30% per year (90% CI -42% to -16%), which excludes zero, and the trailing-180d trend, +22% per year (90% CI -12% to +70%), does not show the decline continuing.

Gross take, rolling 30d

16.06 bps

16.45 bps on 2025-11-12; window ends 2026-11-11

Trend, trailing 365d

-30.3%per year

90% CI -42.3% to -15.8%, excludes zero

Trend, trailing 180d

+22.4%per year

90% CI -11.8% to +69.9%, includes zero

Elasticity to volume, 365d

-0.04

90% CI -0.13 to +0.04, includes zero; R2 0.005

Rolling 30-day gross take went from 16.4 bps to 16.1 bps in 364 days, inside a 11.8 bps to 24.5 bps range

Rolling 30-day take in bps of volume, 2025-05-10 to 2026-11-11 (UTC); each point pools 30 settled days and is dated by its last day

Y axis: bps of volume. Trailing-year gross range 11.8 bps (2026-06-22) to 24.5 bps (2025-12-16).

Take rate by window

Pooled over each block: sum of fees over sum of volume

Window	Days (UTC)	Gross take	Net take	Holder take	Avg daily volume
last 30d	2026-10-13 to 2026-11-11	16.06 bps	4.82 bps	4.82 bps	$124.4M
30-60d ago	2026-09-13 to 2026-10-12	16.94 bps	4.74 bps	4.74 bps	$158.8M
60-90d ago	2026-08-14 to 2026-09-12	13.32 bps	2.89 bps	2.89 bps	$121.6M
150-180d ago	2026-05-16 to 2026-06-14	12.40 bps	3.49 bps	3.49 bps	$81.32M
330-360d ago	2025-11-17 to 2025-12-16	24.49 bps	1.39 bps	0.00 bps	$88.62M

Stricter check. 13 non-overlapping 30-day windows give -26.6% per year, Student-t 90% CI -43.1% to -5.3% (excludes zero).
Full history. +86.6% per year, 90% CI -2.1% to +255.8%, over 522 rolling points from 2025-06-08. The series starts at 0.03 bps, so this measures fees being switched on, not competition; the full-history elasticity of +0.70 carries the same ramp.
Regime break. Holders revenue first appears on 2026-02-23. Pooled gross take is 18.95 bps over the 103 days before it that fall inside the trailing year and 15.49 bps over the 262 days since; net take 1.51 bps before and 3.36 bps since.
Floor read. Gross take sat below the 0.50 bps model floor on 0 of 365 trailing rolling points (minimum 11.82 bps, 2026-06-22). Net take sat below the 3.00 bps bear-case fee-race trigger on 281 of them, and on 149 of the 233 windows that start after the break.

This cuts against the bull case: gross take fell at 30% per year over the trailing year with an interval that excludes zero (non-overlapping check: -27% per year, 90% CI -43% to -5%, also excludes zero), so the claim that pricing power has held while volume grew is not supported by the fee history; the level is still 32x the 0.5 bps model floor. The trailing-180d trend is +22% per year (90% CI -12% to +70%): the decline is not visible in the last six months. A trailing-365d interval that includes zero again would overturn this read; a 180d interval wholly below zero would show compression still in progress. NEAR's captured (net) take went from 1.7 bps to 4.8 bps over the same span and sat under the 3 bps fee-race trigger on 281 of 365 trailing rolling points.

Method and sources

Method. Inner join of the DefiLlama near-intents daily charts (gross fees, revenue, holders revenue, dexs volume). Rolling 30-day take = sum(fees) / sum(volume) x 1e4 bps over 30 consecutive settled days (net uses revenue, holder uses holders revenue). Trend = OLS of ln(rolling gross take) on time in years, change per year = exp(slope) - 1, 90% CI from a Newey-West standard error with a 60-day Bartlett lag (twice the window, because consecutive rolling points share 29 of 30 days). Elasticity = OLS of ln(daily gross take) on ln(daily volume), Newey-West 30-day lag, days with zero fees or volume dropped. Trailing days dated on or after the run date, and up to 3 trailing days with revenue above fees, are excluded. Snapshot field take_rate_series.

Data. 551 aligned settled days, 2025-05-10 to 2026-11-11; 522 rolling points; unsettled trailing day(s) excluded: 2026-11-12.

Sources. defillama/fees:near-intents:dailyFees sha256 562751003b45; defillama/fees:near-intents:dailyRevenue sha256 1241564a028c; defillama/fees:near-intents:dailyHoldersRevenue sha256 826a018a3a3c; defillama/dexs:near-intents sha256 3a66e2dab201.

Lag sensitivity, trailing 365d. 30-day lag -41.9% to -16.4%; 120-day lag -41.1% to -17.6%; graded interval uses 60 days.

Limit 1. Early history mixes fee regimes: the DefiLlama fee chart ramps from zero in mid-2025 and net revenue is zero until the first sweep, so the full-history trend measures fees being switched on. Read the trailing-365d figures.

Limit 2. Fees, revenue and holders revenue follow DefiLlama's near-intents adapter definitions: gross fees include the affiliate fee each frontend keeps; revenue is NEAR swept into the revenue wallets and is booked on the sweep day, so net take is lumpy and can exceed fees on single days. An adapter revision restates the whole series.

Limit 3. Rolling 30-day points overlap (29 of 30 days shared), so about 12 independent observations sit in a year. A Newey-West interval on such points is too narrow at this sample size: in the seeded null simulation in quant/tests/test_take_rate_series.py a trendless series is flagged well above the nominal 10% of the time. The non-overlapping 30-day check (Student-t) stays near its nominal size and is the stricter read: treat a trend as established only when both intervals exclude zero. An interval that includes zero means no decline was detected, not that none exists.

Limit 4. The elasticity regresses fees/volume on volume: noise in daily volume enters both sides and pushes the estimate negative.

Who settles on intents.near

Two populations, measured separately. Deposit sources and direct callers are context; solver concentration, 3,197 HHI across 3.1 effective solvers (highly concentrated), is what the cartelization trigger reads.

Context only; not a solver measure

Who deposits into and calls the contract

External predecessors of intents.near receipts: bridge token contracts, wallets and agents. 12 external receipts from 6 callers.

Caller concentration (HHI)

3,056

90% CI 1,944 to 4,625; Highly concentrated (above 2,500)

Effective callers

3.27

10,000 divided by HHI

AI agent share

50.0%

escrow.ai.near, classified by name only

Bridge and token-contract share

25.0%

HHI excluding agents: 2,222

Account	Share	Receipts	Example tx
escrow.ai.near AI agent (by name)	50.0%	6	HgpVrEcZ…
wrap.near Named account	16.7%	2	jUSWWL5e…
arb-0xaf…t.near Bridge or token contract	8.3%	1	BD4eQxSc…
base-0x8…t.near Bridge or token contract	8.3%	1	7zp76DnR…
qtc.omft.near Bridge or token contract	8.3%	1	w5AwTtUp…
vault.nullpoint.near Named account	8.3%	1	2cffgRrZ…

Read by the solver-cartelization trigger

Which solvers fill signed intents

Fills decoded from execute_intents receipts: 114 fills, 122 solver participations attributed across 6 accounts.

Trigger inconclusive

Solver concentration (HHI)

3,197

90% CI 2,898 to 4,772; Highly concentrated (above 2,500)

Effective solvers

3.13

10,000 divided by HHI

Fills sampled

114

0 participations unresolved

Receipts calling execute_intents

462

623 signed payloads, 0 undecodable

Account	Share	Fills	Example tx
solver-priv-liq.near 55 payloads signed	43.4%	53	HSVFU98i…
6247d9c6…9cfb29 55 payloads signed	25.4%	31	JDy65co9…
solver-multichain-asset.near 35 payloads signed	25.4%	31	J47aB6BN…
brlabs-solver.near 5 payloads signed	4.1%	5	Hh9i4gDL…
5880ad2b…9247dd 1 payloads signed	0.8%	1	EjWobXKM…
s1.anyliq-solver.near 1 payloads signed	0.8%	1	FevqNGCk…

Sample size and coverage

Both indices are a point-in-time receipt sample, not volume-weighted or lifetime measures. Read the counts before the index.

Receipts sampled

175

8 pages of 25 receipts

External receipts

Excluding system and self-receipts

Lifetime receipts

156.9M

Sample covers 1 in 896,565

Block range

#224,323,509 to #224,323,547

38 blocks

Chain time covered

22 s

Ending 2026-11-13 06:05:10 UTC

HHI, all predecessors

4,360

Including system and self-receipts

The caller index rests on 12 external receipts from 6 callers over 22 s of chain time. One burst from one caller moves it materially: read it as a snapshot of traffic, not a market share.
confidential split: nearstats reports confidential 30d $0.10 (0.0000%), rejected (tvl_without_volume: confidential TVL $132.64M but all-time confidential volume $0.35 of $38.55B; stream_ok: nearstats sourceHealth.confidentialStream.status = ok; split_withheld: check confidential-vs-shield (skip) says 'split withheld — NEAR total ≈ public'). Share ESTIMATED ON-CHAIN from the Confidential Intents Treasury: 16.7% (band 13.5%-16.7%; 24h inflow $29.43M, outflow $27.60M, volume $176.44M), replacing the registry assumption 40%
take-rate window leaves out 2026-11-12: DefiLlama reports revenue above fees for that day, so its fee figure is not final; the window ends the day before
nearblocks: 25 duplicate receipt(s) seen on overlapping pages were counted once (id-based cursor vs time-ordered rows); unique receipts sampled = 175

Receipt methods

All 175 sampled receipts, including system and self-receipts.

Method	Receipts	Share
TRANSFER	80	45.7%
execute_intents	62	35.4%
ft_resolve_withdraw	8	4.6%
ft_transfer_call	7	4.0%
near_withdraw	5	2.9%
do_storage_deposit	4	2.3%
ft_on_transfer	3	1.7%
<no action>	2	1.1%
ft_transfer	2	1.1%
storage_deposit	1	0.6%
mt_batch_balance_of	1	0.6%

Method and sources

HHI. Sum of squared percentage shares (0 to 10,000); effective participants = 10,000 / HHI. Bands follow the US DOJ/FTC merger guidelines: below 1,500 unconcentrated, 1,500 to 2,500 moderate, above 2,500 highly concentrated.

Caller index (intents_forensics.hhi_external): deposit-source concentration (external predecessors: bridge token-contract callbacks, direct callers); not solver concentration. System and intents.near self-receipts are excluded. hhi_all includes them.

Classification. system and intents.near self-receipts excluded from external shares; *.omft.near / *.omni.hot.tg = bridge token contracts; *.ai.near / *.agent.near = AI agents BY NAME SUFFIX ONLY (escrow.ai.near is a plain account with no contract; the label is a heuristic, not a verified classification); 64-hex / 0x40-hex = implicit accounts; a predecessor whose receipt is ft_on_transfer is a NEP-141 token contract and is reported as bridge_relayer. execute_intents fills are relayer self-receipts (predecessor intents.near) and never enter the external HHI; solver concentration is in solver_concentration

Solver index (intents_forensics.solver_concentration.hhi_solver): fill participation = (a) the solver side of an atomic execute_intents fill (>= 2 token_diff signers; solver = pure token_diff signer when the counterparty bundles a withdrawal, else a signer whose name matches (*solver*, *-liq*, market maker) or that the structure rule identified elsewhere in the sample, else the named account facing an implicit 1Click deposit address), or (b) a transfer intent from a solver to the settlement hub (see settlement_hub.rule); HHI over participations per solver account; transfers that are not solver-to-hub legs are inventory or user moves, not fills. hhi_solver_ci90 is a bias-corrected bootstrap (bootstrap_method): a block bootstrap by NearBlocks page when the sample is stratified in time (bootstrap_resampling = block_by_page), else over sampled units.

Cartelization trigger (bear_case.scenarios[solver_cartelization]): reads hhi_solver, fires at > 3,500. sample: 122 solver fill participations over 16,269 s; HHI_solver 3,197, bootstrap 90% CI [2,898, 4,772] (CI straddles 3,500: inconclusive, not graded); solver HHI CI straddles the threshold and cross-VM headroom +0.98 bps, not negative; trigger inconclusive (trigger_breached null, not counted as breached or clear); deposit-source HHI over external predecessors (bridge callbacks, direct callers; not solvers) = 3,056.

Sample. 175 receipts over 8 NearBlocks pages (api.nearblocks.io/v1/account/intents.near/txns), block range intents_forensics.sample_block_range, fetched 2026-11-13 06:05:11 UTC. escrow.ai.near is a plain account with no contract: a caller, not a state sink.

Supply physics

Sinks offset 21.6% of the 89,757 NEAR minted per day; the float grows by about 70,377 NEAR a day

Four flows set the change in circulating NEAR. Epoch issuance adds supply; open-market buybacks funded by NEAR Intents fees, gas burn, and storage staking (1 NEAR locked per 100 KB of contract state) remove it. Each flow is measured from mainnet RPC or DefiLlama, then projected over 36 months under the live 2.50% cap and the proposed SVRN glidepath to 1.60%.

Measured issuance

89,757 NEARper day

-0.02%vs nominal 89,772 NEAR

NEAR RPC

Total sinks

19,379 NEARper day

21.6% of issuance; buybacks are 93% of it

DefiLlama

Net float change

+70,377 NEARper day

Issuance minus sinks; the float is growing

Breakeven volume, status quo

$608.0Mper day

4.8×today's $125.8M/day

SVRN glidepath: $386.0M/day, 3.1× today

DefiLlama

Issuance regime lab

Cumulative issuance against cumulative sinks over 36 months at spot price. Switch the regime or set a cap; the model reruns in the browser from the snapshot inputs.

IssuanceBuybacks

Issuance regime

Intents volume path

Custom cap, annual maximum inflation2.50%

0.50%Live cap: 2.50%3.00%

Moving the slider selects the custom regime. Custom caps have no Python reference in the snapshot.

Under Status quo (2.50% max inflation, flat) with volume flat at the 30-day run-rate, sinks cover 20.8% of cumulative issuance over 36 months and the float grows by 80.79M NEAR. Monthly sinks never exceed issuance within the horizon.

Daily issuance

89,757 NEARat M1

96,537 NEARat M36

Cumulative issuance, 36 months

102.0M NEAR

Supply ends at 1.41B NEAR

NEAR RPC

Cumulative sinks, 36 months

21.23M NEAR

Buybacks 19.83M NEAR, burn 1.03M NEAR, lock 374.6K NEAR

Net float change, 36 months

+80.79M NEAR

Positive means the float grows

Crossover month

None in horizon

Cumulative sinks never catch issuance

Breakeven daily volume

$608.0Mper day

4.8×today's volume

At the terminal 2.50% rate

Client model matches the Python output exactly in all 4 snapshot comparisons.

Method, formulas and port check

Supply compounds monthly: S_m+1 = S_m × (1 + i_m × f × D / 365.25), with D = 30.4375 days (supply_physics.days_per_month) and f = 0.9998 (measured over nominal issuance).

Buyback per day is V × τ / 1e4 × α / P with V $124.4M/day, τ 16.06 bps, α 30.0%, P 3.31 USD. Gas burn (938 NEAR/day) and state lock (342 NEAR/day) are held at their measured run-rates with no growth credit.

Net float change is E_m − (B_m + G_m + L_m). The crossover month is the first month with monthly net float change ≤ 0. Breakeven V* is computed at the regime's terminal rate; at the competitive fee ceiling it would be $3.25B/day for this regime.

Port check: each snapshot regime is recomputed from supply_physics.inputs with lib/supplyClient.ts (inflation path, then trajectory, breakeven and staking row) and compared element-wise with supply_physics.trajectories and supply_physics.growth_trajectories. The custom cap runs the same functions and has no stored reference.

Regime	Volume path	Reference scale	Crossover	Result
Status quo (2.50% max inflation, flat)	Flat volume	102.0M NEAR	Match	Exact
SVRN glidepath (2.50% -> 1.60% linear over 24 months, then flat)	Flat volume	77.37M NEAR	Match	Exact
Status quo (2.50% max inflation, flat)	Volume growth	102.0M NEAR	Match	Exact
SVRN glidepath (2.50% -> 1.60% linear over 24 months, then flat)	Volume growth	77.37M NEAR	Match	Exact

Measured issuance

nearcore mints once per epoch. The supply jump at each epoch boundary, annualised by measured epoch time, is the issuance the model uses.

NEAR RPC

Measured daily issuance The model's issuance base: 89,757 NEAR/day
Nominal daily issuance 2.50% cap × total supply: 89,772 NEAR/day
Daily issuance at the SVRN 1.60% target Assumed governance target: 57,454 NEAR/day
Minted per epoch Previous epoch 25,422.6 NEAR: 25,544.2 NEAR
Epoch duration 0.569 s blocks over 43,200 blocks: 6.83 h
Measured annual issuance 1,283.4 epochs a year: 32.78M NEAR
Measured minus nominal -0.02% of nominal: -5.78K NEAR
Implied annual inflation Protocol cap 2.50%: 2.500%
Net burn within the epoch About 938 NEAR/day of gas burned: -267.1 NEAR

Epoch boundaries read
Epoch start block	Time (UTC)	Minted	Source
#224,308,337	2026-11-13 03:40 UTC	25,544.21 NEAR	NEAR RPC
#224,265,137	2026-11-12 20:50 UTC	25,422.56 NEAR	NEAR RPC

Method and sources

minted_per_epoch = total_supply(epoch_start) - total_supply(epoch_start - 1), read from block headers at height 224308337; epoch_duration = ts(epoch_start) - ts(prev_epoch_start); epochs_per_year = 365.25*24 / epoch_duration_hours; implied_annual = minted * epochs_per_year; net_burn = total_supply(epoch_start - 1) - total_supply(prev_epoch_start) (within-epoch drift); nominal_annual = max_inflation_rate * total_supply. nearcore scales the epoch reward by elapsed time and pays nothing to validators below the uptime threshold, so measured can sit below nominal.

Minted = total_supply(epoch start) − total_supply(previous block); epochs per year = 365.25 × 24 / epoch_duration_hours; measured annual = minted_per_epoch × epochs_per_year; nominal annual = max_inflation_rate × total_supply. Fields: near_rpc.epoch_issuance, near_rpc.measured_daily_issuance_near, near_rpc.gross_daily_issuance_near, near_rpc.gross_daily_issuance_svrn_near.

Epoch 4,965, epoch id DQx5CWhRXbBh…, 43,200 blocks per epoch, protocol v88. Gas burn is read from the block before each epoch start: #224,308,337 predecessor#224,265,137 predecessorInflation cap: protocol config.

Breakeven volume

The daily Intents volume at which buybacks, gas burn and storage lock together equal issuance at each regime's terminal rate, at spot price.

Volume

Today's flows, NEAR per day

Gross issuance89,757
Buybacks18,099
Gas burn938
Storage lock growth342

Breakeven daily volume against today's $125.8M/day
Regime	Terminal rate	At live take	× today	At fee ceiling	× today
Status quo	2.50%	$608.0M/day	4.8×	$3.25B/day	25.8×
SVRN glidepath	1.60%	$386.0M/day	3.1×	$2.06B/day	16.4×

A multiple above 1× is how many times today's volume must grow, at a fixed take rate, before the float stops expanding. A lower take rate raises it one for one.

Method and sources

V* = max(0, I − G − L) × P / ((τ / 1e4) × α), with price P $3.31, gross take τ 16.06 bps, capture α 30.0%, and the competitive fee ceiling 3.00 bps (median over cross-VM cells of the best competitor's friction). Fields: supply_physics.breakeven_daily_volume_usd and supply_physics.breakeven_daily_volume_usd_at_ceiling.

Today's volume is intents_forensics.volume.30d / 30 ($3.77B over 30 days). The sinks were computed from the aligned-window model volume supply_physics.inputs.daily_volume_usd = $124.4M/day. Flows: supply_physics.current_daily_sinks_near and supply_physics.current_daily_issuance_near.

Sources:Gross issuanceBuybacksGas burnStorage lock growthModel volumeTake rate

Storage-lock sink

Contract state must be backed by staked NEAR at 1 NEAR per 100 KB. That NEAR cannot circulate while the state exists, so state growth on busy contracts is a supply sink.

Storage price

NEAR locked in tracked state

283,182 NEAR

Across 8 tracked accounts

Measured lock growth

+341.8 NEARper day

0.38% of daily issuance

NEAR archival RPC

intents.near alone

136,471 NEAR

13.65 GB of state

NEAR RPC

Tracked accounts at block 224,323,532
Account	State size	NEAR locked	Growth per day	Window	Source
intents.nearNEAR Intents clearinghouse (Defuse verifier contract)	13.65 GB	136,471 NEAR	+310.0 NEAR	31.1 days	NEAR RPC
auroraAurora EVM engine	14.38 GB	143,832 NEAR	+31.9 NEAR	31.1 days	NEAR RPC
wrap.nearwNEAR wrapper (balance is user deposits; only the storage lock is a sink)	168.97 MB	1,690 NEAR	Not measured	n/a	NEAR RPC
v1.signerChain Signatures MPC coordinator	1.16 MB	12 NEAR	Not measured	n/a	NEAR RPC
escrow.ai.nearNEAR AI escrow (plain account; a caller of intents.near, not a contract sink); no contract code	428 B	0 NEAR	Not measured	n/a	NEAR RPC
treasury.nearProtocol treasury (receives protocol_reward_rate share of issuance)	357.90 KB	4 NEAR	Not measured	n/a	NEAR RPC
meme-cooking.nearmeme.cooking launchpad	5.97 MB	60 NEAR	Not measured	n/a	NEAR RPC
v2.ref-finance.nearRef Finance AMM (v2)	111.34 MB	1,113 NEAR	Not measured	n/a	NEAR RPC

Method, sources and archival readings

NEAR locked = view_account.storage_usage × storage_amount_per_byte (1e+19 yoctoNEAR per byte). Growth per day is the change in storage_usage between archival reads over the window, times the same price. Fields: near_rpc.tracked_sinks, near_rpc.total_state_locked_near, near_rpc.measured_state_lock_growth_near_per_day, near_rpc.measured_state_lock_growth_sinks.

Archival RPC available. Readings behind the growth measurement:

Account	Block	Time (UTC)	Days before snapshot	State size	Change vs prior	Source
intents.near	#219,769,683	2026-10-13 04:46 UTC	31.05	12.67 GB	Oldest point	NEAR archival RPC
intents.near	#223,260,967	2026-11-06 03:39 UTC	7.10	13.39 GB	+720.31 MB	NEAR archival RPC
intents.near	#224,323,532	2026-11-13 06:05 UTC	0.00	13.65 GB	+252.17 MB	NEAR RPC
aurora	#219,769,683	2026-10-13 04:46 UTC	31.05	14.29 GB	Oldest point	NEAR archival RPC
aurora	#223,260,967	2026-11-06 03:39 UTC	7.10	14.37 GB	+84.41 MB	NEAR archival RPC
aurora	#224,323,532	2026-11-13 06:05 UTC	0.00	14.38 GB	+10.93 MB	NEAR RPC

Second usage engine

Chain Signatures handles about 12,918 sign requests a day; the 30-day sample shows no significant trend

Chain Signatures lets a NEAR account sign transactions for other chains through the MPC network. Each request is one sign receipt on v1.signer. The rate below is sampled from NearBlocks receipt pages, so it is an estimate with a standard error, not an exact count.

Sign requests on v1.signer

12 sample windows, 1,200 receipts, 382 sign requests over the trailing 30 days, to 2026-11-13 06:07 UTC.

Cumulative count

Sign requests, last 7 days

15,597per day

SE 1,734, 6 windows

Sign requests, last 30 days

12,918per day

SE 896, day-weighted strata

Trend, per 30 days

+39%

no significant trend; t = 1.28, 90% CI -9% to +113%

Cumulative receipts

26,408,206

All methods (sign, respond, callbacks), NearBlocks index

NearBlocks

Gas per request

0.0021 NEAR

$0.0069 at $3.31; sign + respond + callback

Gas burnt by v1.signer receipts

29.1 NEARper day

3.1% of network gas burn, 0.03% of issuance; $96.17/day

NEAR RPC

Deposit per request

1 yocto

$0.00/day to the signer at the 30d rate

Storage lock

11.6 NEAR

1.16 MB of state, $38.55

NEAR RPC

Top callers of sign(); HHI 2,639 across 17 callers; 66% from *.bridge.near
Caller	Sign receipts	Share
omni.bridge.near	161	42.1%
fast-auth.near	90	23.6%
zcash-connector.bridge.near	49	12.8%
btc-connector.bridge.near	43	11.3%
tx-bench.near	12	3.1%
kernel-zec.meteor-vaults.near	6	1.6%
kernel-near.meteor-vaults.near	4	1.0%
kernel-usdt.meteor-vaults.near	4	1.0%

Sample windows (newest 100 receipts before each anchor)
Anchor	Before	Newest receipt (UTC)	Span	Receipts	Sign	Sign / day	Source
00:00 UTC	2026-11-13	2026-11-12 23:59 UTC	158 s	100	32	17,299	NearBlocks
block height	#224,193,471	2026-11-12 09:07 UTC	138 s	100	30	18,641	NearBlocks
block height	#223,932,836	2026-11-10 15:08 UTC	126 s	100	32	21,680	NearBlocks
00:00 UTC	2026-11-09	2026-11-08 23:59 UTC	261 s	100	34	11,127	NearBlocks
block height	#223,672,201	2026-11-08 21:24 UTC	203 s	100	31	13,051	NearBlocks
block height	#223,411,566	2026-11-07 03:46 UTC	240 s	100	33	11,781	NearBlocks
00:00 UTC	2026-11-05	2026-11-04 23:59 UTC	259 s	100	33	10,906	NearBlocks
00:00 UTC	2026-11-01	2026-10-31 23:59 UTC	195 s	100	31	13,617	NearBlocks
00:00 UTC	2026-10-28	2026-10-27 23:59 UTC	227 s	100	31	11,675	NearBlocks
00:00 UTC	2026-10-24	2026-10-23 23:59 UTC	169 s	100	31	15,701	NearBlocks
00:00 UTC	2026-10-20	2026-10-19 23:59 UTC	225 s	100	33	12,565	NearBlocks
00:00 UTC	2026-10-16	2026-10-15 23:59 UTC	325 s	100	31	8,156	NearBlocks

Method and sources for Chain Signatures

NearBlocks pages of the newest 100 receipts before 4 block heights in the last 7 days (42 h apart) and before 00:00 UTC every 4 days back to 28 days; sign rate per window = (n - 1) / span x sign share; 30d rate = day-weighted 7d and 7-30d means; trend = OLS of ln(sign rate) on time over the 00:00 UTC series. Sampled estimate, not an exact daily count.

Each window is one NearBlocks /v1/account/v1.signer/txns?per_page=100 page, anchored either before_block=H (four heights in the last 7 days, 42 h apart, so they fall at different times of day) or before_date=YYYY-MM-DD (the newest receipts before 00:00 UTC, every 4 days back to 28 days). Receipt rate = (n − 1) / span; sign rate = receipt rate × share of receipts whose action is sign. The 30-day rate weights the last-7-day and 7-30-day means by 7/30 and 23/30; the older stratum is sampled at 00:00 UTC only. The trend is an OLS fit of ln(sign rate) on time over the 00:00 UTC series, so every point is the same time of day; growth = exp(30 × slope) − 1, graded growing only when t ≥ 2. NearBlocks' method=sign filter times out on this account, its count endpoint rejects date filters, and deep before_block queries time out, so exact daily counts are not available key-free.

Gas is receipt_outcome.tokens_burnt summed over every receipt in the windows; the share the contract keeps is burnt_gas_reward from the live protocol config (0%). Storage is the near_rpc.tracked_sinks row for v1.signer. Fields: chain_signatures.*.

What agent vaults would lock

If each AI agent holds its own account or sub-vault, its state pays the same storage stake. The size per agent is an assumption, not a measurement.

Storage price

Agent accounts1.00M

100.0K10.00M

Log scale

State per agent20 KB

5 KBDefault: 20 KB50 KB

Range from the assumption registry

NEAR locked 1,000,000 agents × 20 KB: 200,000 NEAR
Days of issuance absorbed At 89,757 NEAR minted per day: 2.23
Share of total supply: 0.0152%
Relative to intents.near today 136,471 NEAR locked: 1.47×

This is a one-time lock, not a recurring sink: it absorbs issuance once, when the state is created.

Method and the pipeline's fixed grid

Locked = agents × kb_per_agent × storage_near_per_kb (0.01 NEAR per KB). KB per agent is supply_physics.inputs.kb_per_agent_vault: A NEAR account with a few access keys and small contract state costs 5-50 KB; at storage_amount_per_byte = 1e19 yocto 20 KB locks 0.2 NEAR per agent vault.

Scale reference: intents.near view_account. Grid from supply_physics.agent_scale_table:

Agents	KB per agent	NEAR locked	Days of issuance	Share of supply
100.0K	5	5,000 NEAR	0.06	0.0004%
100.0K	20 (default)	20,000 NEAR	0.22	0.0015%
100.0K	50	50,000 NEAR	0.56	0.0038%
1.00M	5	50,000 NEAR	0.56	0.0038%
1.00M	20 (default)	200,000 NEAR	2.23	0.0152%
1.00M	50	500,000 NEAR	5.57	0.0381%
10.00M	5	500,000 NEAR	5.57	0.0381%
10.00M	20 (default)	2,000,000 NEAR	22.28	0.1525%
10.00M	50	5,000,000 NEAR	55.71	0.3812%

Validator security budget

Who is paid from issuance, and what a holder keeps after dilution under each regime's terminal rate.

Validators

Active validators

346

Total staked

554.0M NEAR

42.24% of supply

NEAR RPC

Treasury share of issuance

0.00%

Validators receive 100.00%

NEAR RPC

Nominal staking APR, live cap

5.92%

Cutting the cap to 1.60% changes nominal validator APR by -2.13% and a staker's real return with buybacks by -1.23%. An unstaked holder's real return moves from -2.00% to -1.10%.

Yields at each regime's terminal inflation rate
Regime	Terminal inflation	Nominal staking APR	Real staking APR	Buyback accrual	Real APR with buybacks	Non-staker real return	Daily issuance
Status quo	2.50%	5.92%	+3.42%	+0.50%	+3.92%	-2.00%	89,772 NEAR
SVRN glidepath	1.60%	3.79%	+2.19%	+0.50%	+2.69%	-1.10%	57,454 NEAR

Method and sources

Nominal APR = inflation × (1 − treasury) / staked_ratio; real = nominal − inflation; buyback accrual = buyback NEAR/day × 365.25 / total supply; non-staker real return = −inflation + buyback accrual. Buybacks are open-market purchases funded by NEAR Intents holders revenue, not burns, and accrue to every holder.

Treasury share is protocol_reward_rate, paid to treasury.near. Fields: supply_physics.staking_yield_table, near_rpc.active_validator_count, near_rpc.total_staked_near, near_rpc.staked_supply_ratio, near_rpc.nominal_staking_apr_current.

Sources: Validator setTotal supplyBuyback revenue

Model inputs

Every scalar the trajectories, breakeven volumes and staking rows were computed from: 13 measured, 1 derived, 3 assumptions. Hover an input for its justification and snapshot field.

Input	Value	Range tested	Source
Total supply	1,311,574,393 NEAR	Point value	NEAR RPC
Maximum inflation rate (live cap)	2.50%	Point value	NEAR RPC
Treasury share of issuance	0.00%	Point value	NEAR RPC
Staked share of supply	42.24%	Point value	NEAR RPC
Measured annual issuance	32.78M NEAR/yr	Point value	NEAR RPC
Measured over nominal issuance	99.98%	Point value	NEAR RPC
NEAR price	$3.31	Point value	NearStats
Intents daily volume	$124.4M/day	Point value	DefiLlama
Gross take rate	16.06 bps	Point value	DefiLlama
Share of fees used for buybacks	30.00%	Point value	DefiLlama
Gas burn	938.4 NEAR/day	Point value	NEAR RPC
Storage-lock growth	341.8 NEAR/day	Point value	NEAR archival RPC
SVRN target inflation	1.60%	0.50% to 3.00%	Assumption
SVRN glidepath length	24 mo	6 mo to 48 mo	Assumption
State per agent vault	20 KB	5 KB to 50 KB	Assumption
Volume growth per year	71.69%	12.61% to 161.76%	Derived
Competitive fee ceiling	3.00 bps	Point value	Derived

Method and sources for the headline figures

Measured issuance is supply_physics.current_daily_issuance_near, the epoch-boundary supply jump annualised by measured epoch time and divided by 365.25. Nominal is near_rpc.gross_daily_issuance_near = max_inflation_rate × total_supply / 365.25.

Total sinks is supply_physics.current_daily_sinks_near.total = buyback + gas burn + state-lock growth. Coverage is supply_physics.sink_coverage_ratio = sinks / issuance; net float change is supply_physics.current_net_float_change_near_day = issuance − sinks.

Breakeven volume is supply_physics.breakeven_daily_volume_usd per regime. Today's volume is intents_forensics.volume.30d / 30 ($3.77B trailing 30 days).

Sum-of-the-parts valuation, 10,000 seeded paths over 36 months

The median Monte Carlo path values NEAR at $644.7M, 85.2% below its current $4.34B FDV

The P10 to P90 range is $126.5M to $3.92B; even the 90th-percentile path sits 10% below today's market value. The median falls below the expected path ($1.98B) because volume shocks are lognormal: by month 36 mean daily volume is 7.9× the median. Engines 2 and 3 rest on assumptions rather than measured fee series, so their contribution is the least reliable part of the total.

Median FDV vs current

-85.2%

P50 $644.7M against $4.34B today

NEAR RPC

Expected-path FDV

$1.98B

Deterministic mean path; Monte Carlo mean $1.88B

Terminal multipleClipped

60.0×

Raw peer median 2,498.2× FDV / revenue, capped

CoinGecko

Discount rateAssumption

25%

Annual, applied to 36 months of cash flow and the terminal value

Method and sources

FDV per path = E1 + E2 + E3, each the present value of 36 months of modelled flow plus a terminal value at the horizon, discounted at 25%. Seed 20,261,114, 10,000 paths; GBM shocks on public volume, confidential volume and price; take rates follow a deterministic path because no measured dispersion exists.

Current FDV $4.34B = spot $3.31 × on-chain total supply 1.31B NEAR at the snapshot block. Median vs expected: P50 $644.7M is 32.6% of the expected path. Public volume at M36: mean $524.5M per day vs median $66.17M per day (monthly σ 0.339, n = 11); the median path sits below the mean by exp(−σ²t / 2).

Fields: sotp.fdv_percentiles, sotp.upside_p50_pct, sotp.current_fdv_usd, sotp.calibration.deterministic_base.fdv_total, sotp.calibration.mc_mean_fdv, sotp.calibration.volume_terminal_mean_usd_day, sotp.calibration.volume_terminal_median_usd_day, sotp.calibration.terminal_multiple_raw / _used / _clipped, sotp.inputs.discount_rate.

What you have to believe

Expected annual public-volume growth, terminal FDV / holder-revenue multiple and public take compression speed (per month) carry 84% of the swing-squared variance proxy. Expected annual public-volume growth alone moves FDV from $687.4M to $5.85B across its range. 2 of 17 single-input ranges reach the $4.34B market FDV.

Input at its low endInput at its high endModel baseMarket FDV

Model base $1.98BMarket FDV $4.34B

Expected annual public-volume growth

13% a year to 162% a year, base 72% a year · measured

$687.4M to $5.85B

40.9% of variance

Terminal FDV / holder-revenue multiple

8.0× to 150×, base 60× · derived

$348.9M to $4.80B

30.4% of variance

Public take compression speed (per month)

0% a month to 15% a month, base 3.0% a month · assumption

$600.3M to $3.52B

13.1% of variance

Confidential share at month 36

5.0% to 60%, base 40% · assumption

$1.22B to $2.86B

4.1% of variance

Annual discount rate

10% to 40%, base 25% · assumption

$1.42B to $2.88B

3.3% of variance

Starting 30-day Intents volume (public + confidential)

0.71× base to 1.40× base, base 1.00× base · measured

$1.45B to $2.72B

2.5% of variance

Holder capture (revenue / gross fees)

13% to 32%, base 30% · measured

$929.4M to $2.08B

2.0% of variance

NEAR price drift (gas-burn and state-lock USD value)

-50% to 100%, base 0% · assumption

$1.86B to $2.72B

1.1% of variance

Confidential multiple haircut

30% to 100%, base 70% · assumption

$1.51B to $2.33B

1.0% of variance

Confidential gross take (public take re-reconciled to live fees)

10.0 bps to 30.0 bps, base 20.0 bps · assumption

$1.67B to $2.29B

0.6% of variance

Confidential compression divisor

1.0× to 10×, base 3.0× · assumption

$1.60B to $2.17B

0.5% of variance

Public take hard floor

0.00 bps to 5.00 bps, base 0.50 bps · assumption

$1.98B to $2.39B

0.3% of variance

Blended gross take (public take re-reconciled)

12.4 bps to 17.0 bps, base 16.1 bps · measured

$1.72B to $2.04B

0.2% of variance

NEAR spot price (USD value of gas burn and state lock)

2.46 to 4.45, base 3.31 · measured

$1.94B to $2.03B

0.0% of variance

L1 gas burn (NEAR/day)

704 to 1,173, base 938 · measured

$1.94B to $2.01B

0.0% of variance

Cross-VM headroom (sets the public take floor)

0.74 bps to 1.23 bps, base 0.98 bps · measured

$1.95B to $2.00B

0.0% of variance

State-lock growth (NEAR/day)

256 to 427, base 342 · measured

$1.98B to $1.98B

0.0% of variance

Input range, source

$0$2.0B$4.0B$6.0B

FDV range, share

Method and sources

Basis: deterministic_sotp expected path; one input at its low / high, all others at the snapshot base. Variance proxy: swing^2 / sum(swing^2): first-order share for independent inputs, an approximation. Fields: sensitivity.rows[] (low, high, fdv_low, fdv_high, swing_usd, variance_share), sensitivity.base_fdv_usd, sensitivity.current_fdv_usd, sensitivity.top3_share.

public_volume_growth_yoy (measured): measured 90% interval [0.12614, 1.61761].
terminal_multiple (derived): registry guardrails: terminal_multiple_floor 8x .. terminal_multiple_cap high 150x (live peer median 2,498x, clipped).
fee_compression_lambda_monthly (assumption): registry range [0, 0.15].
confidential_share_target (assumption): registry range [0.05, 0.6].
discount_rate (assumption): registry range [0.1, 0.4].
starting_volume (measured): x exp(+/- measured monthly volume sigma 0.339) (measured: monthly log-return sigma of 30-day volume aggregates).
alpha_capture (measured): min..max over 6 trailing 30-day windows of DefiLlama near-intents revenue / fees.
near_price_drift_annual (assumption): registry range [-0.5, 1].
confidential_multiple_haircut (assumption): registry range [0.3, 1].
confidential_take_bps (assumption): registry range [10, 30].
confidential_compression_divisor (assumption): registry range [1, 10].
public_take_floor_bps (assumption): registry range [0, 5].
gross_take_bps (measured): min..max over 6 trailing 30-day windows of DefiLlama near-intents fees / volume.
price_usd (measured): x exp(+/- measured monthly price sigma 0.296).
gas_burn_near_day (measured): +/- 25% of the live value (sensitivity_unmeasured_band; no measured dispersion).
headroom_cross_vm_bps (measured): +/- 25% of the live value (sensitivity_unmeasured_band; no measured dispersion).
state_lock_near_day (measured): +/- 25% of the live value (sensitivity_unmeasured_band; no measured dispersion).

Not shown: svrn_target_inflation (issuance path only: moves supply at horizon and implied price, not FDV); svrn_glidepath_months (issuance path only: moves supply at horizon and implied price, not FDV).

Scenario model

Change the inputs to recompute the three-engine valuation on its expected path, in your browser. The defaults are the values the snapshot run used. The Monte Carlo percentiles below come from the seeded 10,000-path run and do not move.

Public volume growth, per year+72%

-20%+162%

Measured +72% a year (trend of daily volume, 90% CI +13% to +162%); +4.61% a month on the expected path.

Public take rate at start, gross15.27 bps

1.00 bpsLive competitive ceiling: 3.00 bps20.00 bps

Decays toward a floor of 0.98 bps at 3.0% a month. The ceiling is the best competitor's measured friction on cross-VM routes.

Matches measured fees at 15.27 bps

Confidential share of volume by month 3640.0%

5.0%Live 30-day share: 16.7%60.0%

Assumption; the share moves linearly from the measured 16.7%. Confidential TVL is verified.

Confidential take rate at start, gross20.00 bps

10.00 bps30.00 bps

Assumption, default 20.00 bps: no live confidential fee series exists. Decays at 1.0% a month.

Share of fees spent on buybacks30.0%

0.0%Measured, 30 days: 30.0%50.0%

Holder revenue as a share of gross fees. At the measured share the net take rate is 4.82 bps.

Issuance schedule

FDV under these inputs

$1.98B

Equal tothe snapshot run

Implied price at month 36

$1.446

-56.3%vs spot $3.31

Vs current FDV

-54.4%

Current FDV $4.34B

Contribution by engine

Engine 1, public clearinghouse
PV $50.89M, terminal $1.79B
$966.3M49%
Engine 2, confidential flowAssumption
PV $39.58M, terminal $1.61B, 0.70 multiple haircut
$864.1M44%
Engine 3, monetary premiumAssumption
Gas burn PV $5.76M, state-lock demand $2.10M; not a cash flow
$147.2M7%

Supply crossover

CrossesM31

SVRN glidepath; snapshot run: M31

Sinks as a share of issuance

57.5%

Net float +32.86M NEAR over 36 months

Holder revenue, 36 months

$135.4M

Snapshot run $135.4M

Month-0 fees vs measured

+0.0%

$5.99M vs $5.99M over 30 days

Net change in float per month, NEARIssuance less buybacks, gas burn and state lock; SVRN glidepath

Take rate paths, bpsPublic 15.27 bps to 5.75 bps; confidential 20.00 bps to 14.23 bps

Supply at month 36: 1.37B NEAR used for the price (SVRN glidepath); the model's internal float is 1.34B NEAR. Cumulative issuance 77.37M NEAR, buybacks 40.92M NEAR, net +32,861,115 NEAR.

Method and sources

The panel runs quantClient.deterministicSotp, a line-for-line port of deterministic_sotp in quant/valuo/models/sotp_valuation.py; untouched, it reproduces sotp.calibration.deterministic_base, the expected path under the mean-matching GBM.

Growth: monthly g = (1 + g_yoy)^1/12 − 1 (sotp.inputs.public_volume_growth_yoy); confidential growth is set so the confidential share reaches the target by M36 (odds-ratio construction, sotp.inputs.confidential_share_target).

Public take: τ_t = floor + (τ₀ − floor)(1 − λ)^t, floor = max(0.50 bps, min(τ₀, cross-VM headroom 0.98 bps)), λ = 3.0% per month. The ceiling marker is execution_bench.competitive_fee_ceiling_bps: the median over cross-VM cells of the best competitor's friction. The measured start rate solves (gross fees − V_conf × τ_conf) / V_pub over the 30-day window. Holder revenue = V × τ × α.

Confidential take: sotp.inputs.confidential_take_bps, compressing at λ / 3. Buyback share α = DefiLlama near-intents dailyHoldersRevenue / dailyFees (measured:defillama near-intents dailyHoldersRevenue / fees).

Issuance: sotp.calibration.regime_inflation_paths_annual[regime] × measured scale 0.9998 (per-epoch minting / nominal). Implied price divides FDV by the supply at the horizon for the chosen schedule; current FDV $4.34B is spot × on-chain total supply.

Port check (verifyPort): passed; base-path max relative error 2.8e-16; crossover month matches; heatmap 25 cells, max relative error 2.0e-16.

On-chain, NEAR SocialDB

Publish your forecast

Sign one social.near write from your own NEAR account: the scenario settings above, the model price they imply, your price target and the date it resolves. The forecast board below reads every published forecast back from the chain and scores it against snapshot prices.

Not connected

Your NEAR price target, USDModel at month 36: n/a; snapshot spot $3.310Resolves on (UTC)

The scenario model is not available on this snapshot.

What gets written

Key: <your-account>/valuo/forecast/<YYYYMMDDTHHMMSSZ>
Snapshot: 7b49e1947e0bd6eb…
Target, horizon: n/a on n/a
Volume growth: n/a
Public take: n/a
Confidential: n/a
Buyback share: n/a
Issuance: n/a
Deposit: 0.01 NEAR, plus 0.02 NEAR the first time an account writes to SocialDB (storage registration). It stays in your SocialDB storage balance.

Method and sources

One transaction from your account to social.near, method set, args {data: {<account>: {valuo: {forecast: {<key>: <JSON string>}}}}}, gas 30 Tgas. SocialDB accepts writes only under the signer's own account and records the block height of every value, which the board shows. The key layout mirrors the snapshot anchors in quant/valuo/anchor.py (<account>/valuo/snapshot/<key>).

Deposit: storage_balance_of is read first. An account without a SocialDB storage balance attaches storage_balance_bounds.min (0.02 NEAR on mainnet) plus 0.01 NEAR; a registered account attaches 0.01 NEAR (or 1.5 × the value's bytes × 10¹⁹ yocto, if larger).

The settings stored are the scenario panel's slider values; the model price is the panel's deterministic implied price at month 36. Signing in creates no access key for this site: every write is approved in your wallet.

Read from social.near

Forecast board

Every forecast published under */valuo/forecast/* on NEAR SocialDB, newest write first. A forecast is scored once the latest snapshot's price date reaches its horizon; until then the board shows how far the target is from the latest price.

Reading social.near…

Method and sources

A wildcard-account get on social.near (["*/valuo/forecast/*"]) exceeds the view-call gas limit on mainnet, so /api/forecasts lists accounts through the public SocialDB indexer (api.near.social/get, same key syntax), then re-reads each account from social.near over NEAR RPC with with_block_height. Where that read succeeds, the board shows what the chain returned ("On-chain") and drops indexer entries the chain does not hold; where it fails, the indexer's entries are shown as "Indexer only".

Values are user-written. The API keeps only entries whose key is a UTC timestamp, whose schema is valuo-forecast/1, and whose numbers parse inside fixed bounds; everything else is dropped and counted. Text is rendered as text, never as HTML.

Scoring uses only the snapshot: realized price = intents_forensics.price.current_usd at price.fetched_at. Resolved when that date is on or after the horizon; error = target / realized − 1. The snapshot has no daily price history, so a forecast whose horizon passed before the latest snapshot is scored on the latest snapshot price, and the gap in days is shown on hover.

How the distribution widens over 36 months

Percentiles across 10,000 seeded paths, taken month by month; no single path traces the median line. Dispersion comes from measured volatility: price σ 1.03 a year, volume σ 0.34 a month.

Running implied FDV

P10 to P90P25 to P75MedianCurrent FDV

The FDV an investor would pay at each month's revenue run-rate, at the 60.0× terminal multiple.

Cumulative holder revenue

P10 to P90P25 to P75Median

Fees directed to buybacks, undiscounted. Month 0 reproduces the measured DefiLlama holder revenue.

Method and sources

Running FDV_t = M × 12 × (public holder revenue_t + h × confidential holder revenue_t + L1 gas burn_t), with M = 60.0× and h the confidential multiple haircut. Cumulative holder revenue = Σ (V_pub × τ_pub + V_conf × τ_conf) × days × α.

Mean-matching (Itô) GBM: the deterministic path is the expected path and the median path sits below it. Fields: sotp.fan_chart.fdv_p10 … fdv_p90, sotp.fan_chart.rev_p10 … rev_p90, sotp.calibration.sigma_price_annual, sotp.calibration.sigma_volume_monthly, sotp.calibration.gbm_convention.

Percentiles across paths

FDV and implied price at each percentile, against today's $4.34B FDV and $3.31 spot price.

Percentile	FDV	vs current FDV	Price at M36	vs spot	FDV excl. Engine 3
P10	$126.5M	-97.1%	$0.092	-97.2%	$118.3M
P25	$261.2M	-94.0%	$0.191	-94.2%	$248.3M
P50	$644.7M	-85.2%	$0.471	-85.8%	$610.3M
P75	$1.64B	-62.2%	$1.200	-63.8%	$1.54B
P90	$3.92B	-9.6%	$2.869	-13.3%	$3.67B

When does net supply stop growing?

Issuance schedule	Crosses within 36 months	Median month	P10 to P90 month	Deflationary at M36	Median net float change
Status quo	66.1%	M14	M5 to M30	47.2%	+40.24M NEAR
SVRN glidepath	73.3%	M14	M5 to M29	55.3%	+15.59M NEAR

Share of paths on which monthly issuance falls below buybacks, gas burn and state lock combined. Expected path: M31.

Method and sources

Implied price = FDV / supply at M36 under the SVRN glidepath (1.37B NEAR). FDV excluding Engine 3 sums Engines 1 and 2 per path. The Monte Carlo mean FDV ($1.88B) exceeds the P50 by lognormal convexity and is not a base case.

Crossover: first month whose net float change (issuance - buyback - gas burn - state lock) <= 0. Path probabilities reflect the dispersion of the volume and price GBMs (a path crosses when holder revenue / price outruns issuance); compare with deterministic_crossover_month (the expected-volume path) and with supply_physics.breakeven_daily_volume_usd before quoting a crossover. Paths crossed: Status quo 6,606, SVRN glidepath 7,328. Median net float change is the median over paths of the 36-month sum; positive means net dilution.

Fields: sotp.fdv_percentiles, sotp.price_percentiles, sotp.fdv_ex_monetary_percentiles, sotp.crossover_month_distribution, sotp.calibration.supply_at_horizon_used.

Value by engine

On the expected path, the 2 assumption-heavy engines carry 51% of the $1.98B total.

Expected path, engines add to the total

Engine	P10	P50	P90	Expected path	Median-path share
Engine 1, public clearinghouse	$28.68M	$205.3M	$1.70B	$966.3M	5.2%
Engine 2, confidential flowAssumption	$21.13M	$171.4M	$1.70B	$864.1M	94.7%
Engine 3, monetary premiumAssumption	$1.17M	$10.37M	$163.4M	$147.2M	0.1%

Engine percentiles are taken engine by engine and do not add: the three P50s sum to $387.1M against a total P50 of $644.7M. The last column decomposes the single path closest to the total median.

Method and sources

Engine 1: PV of H months of public holder revenue (V_pub x tau_pub x alpha) at the discount rate, plus year-3 annualized revenue x peer-median FDV/revenue multiple, discounted; tau_pub reverts to the benchmark floor.

Engine 2: Same construction on confidential volume at the assumed confidential take; terminal multiple x confidential_multiple_haircut because nearstats flags confidential TVL as unverified. Starting confidential share 16.7% is an on-chain estimate from Confidential Intents Treasury boundary flows (intents_forensics.volume.confidential_share_estimate), not nearstats' rejected split.

Engine 3: PV of L1 gas burn (grown with public volume) at the peer multiple plus the discounted market value of newly state-locked NEAR (sum state_lock_t x P_t): a demand-flow proxy, not a cash flow.

Engine 2 is marked as an assumption because the confidential TVL is unverified at the source and no confidential fee series exists; Engine 3 values a demand flow (newly state-locked NEAR × price) that is not a cash flow. Engine 1 inherits the clipped peer multiple.

Fields: sotp.engines, sotp.calibration.deterministic_base, sotp.calibration.engine_shares_on_median_total_path (path 2), sotp.fdv_percentiles.p50.

Sensitivity to volume and take rate

FDV with daily volume and net take held flat for 36 months. 17 of 25 cells turn net supply deflationary under the SVRN glidepath.

Live operating point (ink border): $124.4M a day at 4.82 bps netNearest cell $100.0M/day × 6 bps: $725.2M
Net take: Gross 16.06 bps × buyback share 30.0%
Against today's $4.34B FDV: At least one cell in the grid reaches the current market value.

Method and sources

Each cell runs the deterministic model with volume and net take constant for the horizon: buyback share folded into the net take, zero growth, zero compression, zero price drift, SVRN glidepath issuance. Confidential share 16.7% of each column. Basis: held constant for the horizon; alpha folded into the net take; regime svrn_glidepath. The snapshot key is named fdv_p50, but the cells are deterministic, not Monte Carlo medians.

Browser recomputation: 25 of 25 cells reproduced, max relative error 2.0e-16. Fields: sotp.sensitivity_heatmap (x, y, fdv_p50, net_deflation, confidential_share_used), operating point from sotp.calibration.v_pub_30d_usd + v_conf_30d_usd and net_take_start_bps.

Relative value

NEAR's multiples against the live peer set, ranked lowest first. This places NEAR among other L1s; it does not value the Intents cash flow.

Only 3 peers. Peer 'revenue' follows each chain's own DefiLlama methodology, so revenue multiples compare different cash-flow definitions; with n = 3 peers the median is a reference point, not a distribution.

Chain	FDV	Annualised fees	Annualised revenue	FDV / fees	FDV / revenue
HYPE Hyperliquid (parent: perps + spot + HLP + outcomes) Revenue definition	$67.43B	$1.09B DefiLlama	$835.6M	62×	81×
SOL Solana L1 Revenue definition	$49.97B	$201.4M DefiLlama	$20.00M	248×	2,498×
ETH Ethereum L1 Revenue definition	$233.2B	$242.3M DefiLlama	$75.82M	962×	3,076×
NEAR (L1 gas + Intents clearinghouse) Revenue definition	$4.37B	$72.26M	$23.46M	61× rank 1 of 4	186× rank 2 of 4
NEAR L1 gas Component of the NEAR rowRevenue definition	same	$1.60M	$1.12M	2,735×	3,907×
NEAR Intents Component of the NEAR rowRevenue definition	same	$70.66M	$22.34M	62×	196×

NEAR trades at 61x annualised fees versus a peer median of 248x (rank 1 of 4, lowest first; below every peer); on revenue it is 186x versus 2,498x (inside the peer range).
0 of 3 peers trade at or below the model's 60x terminal FDV/revenue multiple; NEAR itself trades at 186x.
At the peer FDV/fees distribution, NEAR's $72M of annualised fees maps to $11.20B-$43.73B (median $17.93B, n = 3) against a $4.34B market FDV; a relative-value cross-check, not a fair value.
NEAR issues $109M a year to validators and delegators (2.5% of FDV) against $72M of annualised fees and $23M of holder revenue (gas burn + Intents buyback): revenue covers 22% of the security budget.
At 2.5% inflation stakers earn 5.92% nominal and 3.92% real including buybacks; non-stakers lose 2.00% a year to dilution net of buybacks.

Peer-multiple cross-check, not a fair value

NEAR's annualised cash flow times the peer multiple distribution, set against the $4.34B market FDV. With 3 peers the quartiles are reference points, and peer multiples carry each chain's monetary premium.

Multiple applied	NEAR cash flow, annualised	Peer P25	Peer median	Peer P75	Peers	Median ÷ market FDV
Peer FDV / fees	$72.26M	$11.20B	$17.93B	$43.73B	3	4.1×
Peer FDV / revenue	$23.46M	$30.25B	$58.61B	$65.39B	3	13.5×

Staking and the security budget

Security budget Issuance to validators and delegators, at spot: $108.5M a year
Security budget as share of FDV: 2.5%
Fees ÷ security budget $72.26M of annualised fees: 67%
Holder revenue ÷ security budget $23.46M of gas burn plus Intents buyback: 22%
Net dilution Security budget less holder revenue: $85.05M a year
Staked share of supply Nominal staking yield 5.92%: 42.2%

Issuance regime	Inflation	Staker yield, nominal	Staker yield, real incl. buyback	Non-staker, real
Status quo	2.5%	5.92%	3.92%	-2.00%
SVRN glidepath	1.6%	3.79%	2.69%	-1.10%

Method and sources

Basis: live comps (DefiLlama fees/revenue, CoinGecko FDV) recomputed; rank 1 = lowest multiple; percentile = share of peers strictly below NEAR. Fields: relative_value.peers[], relative_value.near, relative_value.near_components[], relative_value.coverage. Run-rate multiples annualise the last 30 days; trailing-year multiples are in fdv_to_fees_1y / fdv_to_revenue_1y (NEAR 79× on fees, rank 2 of 4); 424× on revenue, rank 2 of 4.

Caveat as written by the pipeline: Peer 'revenue' follows each chain's own DefiLlama methodology (peers[].revenue_methodology), so revenue multiples compare different cash-flow definitions; with n = 3 peers the median is a reference point, not a distribution.

Model terminal multiple 60×; peers at or below it: 0 of 3 (relative_value.model_terminal_multiple).

Cross-check: relative-value cross-check, not a fair value: NEAR cash flow x the live peer multiple distribution. Fees leg: comps.near_combined.annualized_fees_usd x peer FDV / annualised fees (30-day run-rate). Revenue leg: comps.near_combined.annualized_revenue_usd x peer FDV / annualised revenue (30-day run-rate). (relative_value.implied_fdv_band)

Security budget: security budget = supply_physics.current_daily_issuance_near x 365.25 x NEAR price; fees / revenue = comps.near_combined 30-day run-rate. Measured inflation 2.50%, NEAR at $3.31 (relative_value.staking).

The 60x terminal multiple cap sits at the 13th percentile of 15 peers (2 of 15 trade at or below it) on FDV to annualised 30-day revenue

The wide median is 531x (bootstrap 90% interval 274x to 999x), and NEAR on L1 gas revenue alone trades at 3,907x, the 87th percentile.

A wider set than the three peers the model uses, ranked by revenue before any multiple was computed. It does not feed the valuation.

DefiLlamaCoinGecko

Peer chain tokenFlagged peer, kept in the distributionInterquartile rangeBootstrap 90% interval of the medianModel cap and floorNEAR

FDV divided by annualised 30-day revenue, log scale. Hover a dot for the chain.

Lowest multiple first. NEAR and the model's cap and floor are placed where they fall.
Chain	30-day revenue	FDV	FDV / revenue
Quai Rank 12 by revenueFlagged	$200.4K	$7.65M	3.1×
Model terminal multiple floor Assumption, from the run's registry	n/a	n/a	8×
POL Polygon Rank 5 by revenue	$1.62M	$911.3M	46×
Model terminal multiple cap Assumption, from the run's registry	n/a	n/a	60×
TRX Tron Rank 1 by revenue	$24.62M	$29.54B	99×
NEAR, L1 + Intents The basis the model values NEAR on. Not counted in the set	$1.93M	$4.37B	186×
ARB Arbitrum Rank 9 by revenueFlagged	$524.5K	$1.43B	225×
STRK Starknet Rank 17 by revenueFlagged	$109.4K	$364.8M	274×
INJ Injective Rank 14 by revenue	$144.7K	$719.0M	408×
APT Aptos Rank 15 by revenue	$123.0K	$642.0M	429×
MON Monad Rank 11 by revenue	$253.6K	$1.64B	531×
ICP Rank 13 by revenue	$173.6K	$1.41B	669×
AVAX Avalanche Rank 10 by revenue	$342.8K	$3.81B	914×
FIL Filecoin Rank 16 by revenue	$115.6K	$1.40B	999×
SOL Solana Rank 4 by revenue	$1.64M	$49.97B	2,498×
ETH Ethereum Rank 2 by revenue	$6.23M	$233.2B	3,076×
NEAR, L1 gas revenue only Rank 19 of 69 by revenue. Not counted in the set	$91.98K	$4.37B	3,907×
BNB BSC Rank 6 by revenueFlagged	$1.18M	$82.35B	5,735×
HYPE Hyperliquid L1 Rank 8 by revenueFlagged	$851.5K	$67.43B	6,508×

Peers in the set Top 15 by 30-day revenue with their own token, of 69 chains with revenue: 15
Median FDV / revenue Bootstrap 90% interval 274× to 999× (10,000 resamples, seed 20261114): 531×
Interquartile range 25th to 75th percentile: 249× to 1,748×
Lowest to highest: 3.1× to 6,508×
Model cap, 60× 2 of 15 peers trade at or below it: 13th percentile
Model floor, 8× 1 of 15 peers trade at or below it: 7th percentile
NEAR on L1 gas revenue only, 3,907× 13 of 15 peers trade at or below it: 87th percentile
NEAR on L1 + Intents revenue, 186× The multiple the model uses; no peer row has an equivalent: 20th percentile
Median without flagged rows 10 peers; cap at the 10th percentile: 600×
Three-peer median the model uses Solana, Ethereum, Hyperliquid: 2,498×

The cap is at or below the upper quartile of the set (1,748x), so the model's terminal value does not rest on a multiple richer than what the market pays for chain revenue today. On L1 gas revenue alone NEAR (3,907x) is priced above the wide median (531x), so NEAR is not cheap on chain revenue; the long-term case has to be carried by Intents revenue, not by the L1 multiple. This reading is overturned if a re-run puts the cap above the 75th percentile of the set (peer multiples compressing toward 60x) or if the set shrinks below 8 peers.

Limits of this comparison

FDV to revenue on 30 days annualised is noisy: one busy or quiet month moves a chain's multiple by a large factor.
Revenue definitions differ by chain adapter (burned base fees, sequencer profit, treasury share, buyback and burn); the adapter's own text is carried per row.
Token FDV for an L2 or an exchange-linked token does not accrue chain revenue the way an L1 gas token does; those rows are flagged and a sensitivity without them is shown.
Selection: ranking by revenue picks the chains with the most revenue, which pulls their multiples down relative to the long tail of low-revenue chains.
NEAR's multiple here uses L1 gas revenue only. The model values NEAR on L1 plus Intents revenue (comps.near_combined), which no peer row here has an equivalent of.
One snapshot. Nothing here is a time series, and the cap's position will move with token prices and monthly fee volume.

Inclusion rules, excluded chains and sources

Basis: FDV (CoinGecko fully_diluted_valuation) / annualised chain revenue (DefiLlama overview, dataType=dailyRevenue, total30d x 365/30). annualized_revenue_usd = revenue_30d_usd x 365 / 30 (30-day run-rate, the same convention as comps). Percentile: share of peers with FDV/revenue at or below the value.

Universe: every DefiLlama fees-overview row with category 'Chain' and 30-day revenue (dataType=dailyRevenue) above zero.
Ranking: by 30-day revenue, highest first. No chain is added or removed by hand after seeing its multiple.
Peers: the first 15 chains in that ranking that have their own token with a CoinGecko fully diluted valuation.
Excluded, each listed with its reason: chains with no native token; chains whose token has no FDV on CoinGecko; chains not in the chain-to-token mapping table (configuration in wide_peers.py); a second row for a token already counted.
Flagged, not dropped: L2 tokens, exchange-linked tokens, rows whose adapter books something other than chain gas revenue and rows whose DefiLlama adapter is not typed 'chain'. They stay in the distribution; a sensitivity without them is reported next to it.
NEAR is the subject. It is positioned against the set on the same basis (its own chain row and CoinGecko FDV) and is not counted in it.
Multiple: FDV / (30-day revenue x 365 / 30). Percentile of a value: share of peers with a multiple at or below it.

Excluded inside the cut (2): Base, rank 3 with $2.57M of 30-day revenue (no native token: Coinbase-operated L2, gas is paid in ETH, so there is no token FDV to divide by revenue); Robinhood Chain, rank 7 with $1.08M of 30-day revenue (no native token: Robinhood-operated L2, gas is paid in ETH, so there is no token FDV to divide by revenue). 51 lower-revenue chains fall below the cut and are not listed.

Flagged rows (5): BSC: exchange-linked token: BNB's FDV also prices Binance exchange utility, not only BNB Chain's fee burn; Hyperliquid L1: chain row covers HyperEVM gas only; HYPE's FDV is priced mostly on the HyperCore exchange, whose revenue DefiLlama books under separate Hyperliquid protocols (the model's three-peer set uses that parent); Arbitrum: L2 governance token: sequencer revenue accrues to the DAO treasury, not to ARB by protocol rule; Quai: revenue is merged-mining block rewards earned on other chains and spent on QUAI buyback and burn (the adapter's own Fees text), not gas paid on Quai; Starknet: L2 token: sequencer revenue accrues to the operator and foundation, not to STRK by protocol rule. Without them: 10 peers, median 600×, quartiles 413× to 977×, 1 at or below the cap.

Cap 60× and floor 8×: assumptions registry, preset 'base' (wide_peers.terminal_multiple_cap, wide_peers.terminal_multiple_floor). L1-only: NEAR's own chain row in the same DefiLlama overview (gas revenue) over NEAR's FDV from the same CoinGecko call. This is not the model's combined L1 + Intents multiple, which is comps.near_combined.fdv_to_revenue.

Three-peer contrast: the three-peer median the model uses (Solana, Ethereum, Hyperliquid), for contrast, 2,498×. Recomputed here from the same DefiLlama overview and CoinGecko call as the wide set (Hyperliquid = sum of the rows whose parentProtocol is parent#hyperliquid). The valuation itself reads comps.median_fdv_to_revenue_ex_near, fetched seconds apart, and clips it to the floor and cap.

Fields: wide_peers.peers[], wide_peers.distribution, wide_peers.near, wide_peers.ex_flagged, wide_peers.excluded[]. Bootstrap: percentile bootstrap of the median, resampling peers with replacement; 5th and 95th percentiles.

Calibration

Every parameter behind the paths, marked by how it was obtained.

Parameter	Value	Basis	Note	Source
Volatility
Price volatility, annual	1.026	Measured	364 daily closes over 1y; 0.296 a month	NearStats
Volume volatility, monthly	0.339	Measured	11 monthly returns; confidential leg × 1.00	DefiLlama
Volume engine	GBM (unchanged)	Assumption	Lognormal paths from the sigma above. A block bootstrap of daily changes is reported next to it and is not used: at the same expected path its P10 to P90 for 30-day volume in 12 months is 0.09x to 3.73x today's, against 0.19x to 3.88x for the GBM; swapped into the Monte Carlo it gives FDV P50 $293.1M against $644.7M. Neither engine passes the out-of-sample coverage test.	Backtest, History tab
Volume and fees
Volume, 30 days, public / confidential	$3.11B / $622.5M	Measured	Confidential share 16.7%	NearStats
Public take rate at start	15.27 bps	Derived	Blended gross 16.06 bps = fees $5.99M / volume	DefiLlama
Confidential take rate at start	20.00 bps	Assumption	No live confidential fee series	Model input
Buyback share of fees	30.0%	Measured	Net take 4.82 bps at start	DefiLlama
Take-rate floor	0.98 bps	Derived	Cross-VM headroom; competitive ceiling 3.00 bps	1Click
Growth and decay
Take-rate decay, public / confidential	3.00% / 1.00% a month	Assumption		Model input
Public volume growth	+4.61% a month	Measured	+72% a year, measured trend used as the expected path	Model input
Confidential volume growth	+8.16% a month	Measured and assumed	Share 16.7% measured to 40.0% assumed by M36	NearStats
Valuation
Terminal multiple, raw / used	2,498.2× / 60.0×	Derived	Peer median clipped to the cap	CoinGecko
NEAR price	$3.31	Measured	Current FDV $4.34B	NearStats
Discount rate	25%	Assumption		Model input
Monte Carlo mean FDV	$1.88B	Derived	2.92× the P50; convexity, not a base case	Model output
Supply
Gas burn at start	938 NEAR a day	Measured	On-chain supply drift; DefiLlama cross-check 926 NEAR a dayDefiLlama	NEAR RPC
State lock growth at start	342 NEAR a day	Measured	Storage growth × 1 NEAR per 100 KB	NEAR archival RPC
Supply at M36, glidepath / status quo	1.37B NEAR / 1.39B NEAR	Derived	Issuance scale 0.9998 measured / nominal	NEAR RPC

Model warnings

sotp: confidential share 16.7% is an ON-CHAIN ESTIMATE from Confidential Intents Treasury boundary flows (band 13.5%-16.7%), not nearstats' split (rejected as implausible); 24h window, biases listed in intents_forensics.volume.confidential_share_estimate
sotp: peer median FDV/revenue 2498.2x clipped to 60.0x (guardrails [8, 60])

Method and sources

Take basis: tau = gross fees / volume (bps); alpha = holders revenue / fees; holder revenue = V x tau x alpha. Public take start: reconciled: (gross_fees_30d - V_conf_30d x tau_conf / 1e4) / V_pub_30d. Buyback share: measured:defillama near-intents dailyHoldersRevenue / fees. Floor: derived, max(0.50 bps, min(public take, cross-VM headroom 0.98 bps)).

Volatility: nearstats:price-history, log-return σ × √365 for price; measured: monthly log-return sigma of 30-day volume aggregates for volume. Price drift is zero unless the assumption is changed. Mean FDV: mean over paths; exceeds the P50 by lognormal convexity (mean-matching GBM: the deterministic path is the EXPECTED volume path, the P50 path sits below it, see gbm_convention) and is not a base case.

Gas burn: −net burn per epoch × epochs per year / 365.25 from the epoch-start block vs its predecessor (supply_physics.current_daily_sinks_near.gas_burn); cross-check supply_physics.cross_checks.gas_burn_defillama_near_day. State lock: archival storage_usage growth across tracked sinks × near_rpc.storage_near_per_kb (0.01 NEAR per KB).

Crossover: first month whose net float change (issuance - buyback - gas burn - state lock) <= 0. Price: fdv / (S_0 + gross issuance - buybacks - gas burn - state lock at H, svrn_glidepath trajectory). Current FDV: spot price x on-chain total_supply at the snapshot block. Days per month 30.4375; discount rate 25.00%; horizon 36 months; seed 20,261,114; 10,000 paths.

Fields: sotp.calibration (sigma_price_annual, sigma_volume_monthly, tau_pub_start_bps, tau_conf_start_bps, alpha_capture, tau_floor_bps, lambda_pub_monthly, growth_monthly_pub, growth_monthly_conf, terminal_multiple_raw, terminal_multiple_used, v_pub_30d_usd, v_conf_30d_usd, gas_burn_near_day, state_lock_near_day, supply_at_horizon_by_regime, mc_mean_fdv, warnings), sotp.inputs.

Bear case and falsification

1 of 3 measured early-warning triggers is breached; all four stresses together cut the deterministic value by 89%.

Every value in this section is deterministic at median inputs, not a Monte Carlo percentile. Each scenario changes a few inputs of the same valuation and re-prices it, so any difference comes from the scenario alone. Breached now: Governance stalemate: 2.5% issuance forever. Inconclusive, measured on too short a sample to grade: Solver cartelization: HHI_solver > 3,500, 1Click loses the spread (sample span 4.5 h against 1 h required).

Base value, deterministic

$1.98B

Implied price $1.446

Combined worst value

$211.5M

-89.3%vs base

Implied price $0.152 (-89.5%)

Triggers breached now

1 of 3

1 inconclusive (sample too short)

Spot FDV, for reference

$4.34B

-54.4%base vs spot

NearStats

Scenario changes are measured from the deterministic base, not from spot. For contrast, the seeded Monte Carlo median is $644.7M; the deterministic base sits +206.8% against it.

4 stress scenarios and the combined worst case

Each row re-prices the same deterministic valuation with the inputs in the second column changed.

Scenario	What changes	Value (FDV)	Change vs base	Implied price	Supply crossover
Base case, unperturbed inputs The snapshot's measured inputs and default assumptions. Every change below is measured against this row.	Nothing; measured inputs and default assumptions	$1.98B	Base	$1.446	Q11 (M31)
Aggregator bypass: public take races to 1 pip Third-party frontends route around near.com and the public gross take collapses from 15.3 bps to 0.01 bps (the bare intents.near contract fee); confidential execution reprices from 20 bps to 5 bps. Capture ratio alpha is held at 30%, so the hit is price, not allocation. Deterministic (median-input) FDV −81.7% to $361.0M; implied price $0.26. Evidence against it: live cross-VM headroom is +1.0 bps, so a bypassing frontend must first give up that execution edge. Trigger: DefiLlama 30d net take (gross x alpha) below 3.0 bps (now 4.8 bps).	Public take rate 0.01 bps Confidential take rate 5.00 bps Take-rate floor 0.01 bps	$361.0M $213.8M ex-monetary	-81.7%	$0.264 -81.7%	none within horizon
Governance stalemate: 2.5% issuance forever Validators never enact the SVRN cut: max_inflation_rate stays at 2.50% for the full horizon instead of gliding to 1.60%. Protocol cash flows do not change, so FDV moves +0.0%; the damage is dilution: 24,648,760 more NEAR outstanding at the horizon and a per-token value of $1.42 (−1.8% vs base). Net-float crossover: none within horizon (base: month 31). This scenario is live today: the on-chain cap reads 2.50% and no cut has been enacted.	Issuance target 2.50%	$1.98B $1.83B ex-monetary	0.0%	$1.420 -1.8%	none within horizon
Post-incentive volume decay (NEAR@3.33 unwind) Mercenary liquidity from the NEAR@3.33 incentive program unwinds: public daily volume $103.6M → $41.5M (−60%), confidential $20.7M → $6.2M (−70%) and confidential TVL x0.3 (nearstats flags that TVL unverified). Take rates are unchanged, so revenue falls with volume and the deterministic (median-input) FDV −59.9% to $793.0M; implied price $0.58. Current 30d volume $3.77B is 78% of the 30d record $4.84B; the trigger fires at −40%.	Public daily volume $41.45M Confidential daily volume $6.22M Confidential TVL $39.79M Struck-through input is listed but not applied (no engine input)	$793.0M $645.8M ex-monetary	-59.9%	$0.580 -59.9%	none within horizon
Solver cartelization: HHI_solver > 3,500, 1Click loses the spread Two market makers take most fills (HHI_solver > 3,500) and spreads widen until Relay/LI.FI deliver 2 bps better net execution on cross-VM (headroom −2 bps). Keeping flow costs the same 2 bps of take (15.3 bps → 13.3 bps) while compression runs at 0.15/month. Deterministic (median-input) FDV −40.6% to $1.17B; implied price $0.86. Live: HHI over solver fill participations 3,197 (3.1 effective solvers across 122 fill participations), bootstrap 90% CI [2,898, 4,772] straddles 3,500, so the HHI condition is inconclusive; deposit-source HHI over external predecessors 3,056 (not a solver measure); cross-VM headroom +1.0 bps; trigger at HHI_solver > 3,500 or headroom < 0.	Deposit-source HHI 3,500 Cross-VM headroom -2.0 bps Public take rate 13.27 bps Fee compression rate 0.15 per month Struck-through inputs are listed but not applied (no engine input)	$1.17B $1.03B ex-monetary	-40.6%	$0.859 -40.6%	none within horizon
All 4 scenarios at once (combined worst) Every perturbation applied together; where two scenarios set the same input, the value worse for the thesis wins.	Public take rate 0.01 bps Confidential take rate 5.00 bps Take-rate floor 0.01 bps Issuance target 2.50% Public daily volume $41.45M Confidential daily volume $6.22M Confidential TVL $39.79M Deposit-source HHI 3,500 Cross-VM headroom -2.0 bps Fee compression rate 0.15 per month Struck-through inputs are listed but not applied (no engine input)	$211.5M	-89.3%	$0.152 -89.5%	none within horizon

How far each scenario moves value and price

Zero is the deterministic base. Solid bars are the change in value; light bars the change in implied price, which differs only when a scenario changes NEAR supply at the horizon.

Early-warning triggers

Each scenario is paired with a live metric and a threshold. The thesis is treated as falsified if any trigger holds for 30 consecutive days.

1 of 5 breached, 1 inconclusive

Scenario	What we watch	Now	Fires when	Status	Monitor
Aggregator bypass: public take races to 1 pip	Net take rate, 30 days	4.82 bpsDefiLlama	< 3.00 bps	Not breached	api.llama.fi/summary/fees/near-intents
Governance stalemate: 2.5% issuance forever	Protocol inflation cap Breached while the on-chain cap exceeds the SVRN target: the cut has not been enacted.	2.50%NEAR RPC	> 1.60%	Breached	free.rpc.fastnear.com
Post-incentive volume decay (NEAR@3.33 unwind)	30-day volume versus its 30-day record	-22.0%NearStats	<= -40.0%	Not breached	nearstats.org/api/near/summary
Solver cartelization: HHI_solver > 3,500, 1Click loses the spread	Solver concentration (HHI) Sample: 122 solver fill participations over 16,269 s; HHI_solver 3,197, bootstrap 90% CI [2,898, 4,772] (CI straddles 3,500: inconclusive, not graded); solver HHI CI straddles the threshold and cross-VM headroom +0.98 bps, not negative; trigger inconclusive (trigger_breached null, not counted as breached or clear); deposit-source HHI over external predecessors (bridge callbacks, direct callers; not solvers) = 3,056.	3,19790% CI 2,898 to 4,772NearBlocks	> 3,500	InconclusiveSample span 4.5 h against 1 h required	api.nearblocks.io/v1/account/intents.near/…
Secondary trigger, same scenario	Cross-VM execution headroom	+0.98 bps1Click	< 0.00 bps	Not breached	Source shown under Now

Daily canonical snapshot · the 30-day rule

No trigger has held for 30 consecutive days; the longest current streak is 2.7 days

A streak counts elapsed days between the first and the latest daily record of an unbroken breach. Records more than 36 h apart break it (a missed daily run does not count as breached), and so does a day the trigger was not measured.

Trigger	Streak	Days to falsification	On-chain record
Aggregator bypass: public take races to 1 pip Net take rate, 30 days < 3.00 bps	clear	n/a	none
Governance stalemate: 2.5% issuance forever Protocol inflation cap > 1.60%	breached 2.7 of 30 consecutive days since 2026-11-10 13:04 UTC, 4 daily records; restarted after a gap in the daily records	27.3	none
Post-incentive volume decay (NEAR@3.33 unwind) 30-day volume versus its 30-day record <= -40.0%	clear	n/a	none
Solver cartelization: HHI_solver > 3,500, 1Click loses the spread Solver concentration (HHI) > 3,500	not measured in the latest daily record	n/a	none

On the day a streak reaches 30 days, the daily job (ops/vps/valuo-run.sh snapshot) writes valuo/falsification/<trigger> to NEAR mainnet SocialDB under 0161aff7…81f0, signed with the same key that anchors each snapshot; no one has to act. Live read at 2026-11-14 05:06 UTC: no record under valuo/falsification/* yet.

Conditions under which the thesis is wrong

Written by the quant engine from the same numbers as the tables; identical to the paragraph in the IC memo.

Aggregator bypass: public take races to 1 pip. Public gross take at 0.01 bps and confidential at 5 bps puts the deterministic FDV at $361.0M (−81.7%); watch DefiLlama near-intents 30d net take weekly, trigger < 3.0 bps, now 4.8 bps (not breached).

Governance stalemate: 2.5% issuance forever. No SVRN cut (cap 2.50% for 36 months) leaves the deterministic FDV at $1.98B (+0.0%) but dilutes per-token value to $1.42 (−1.8%) with crossover none within horizon; watch max_inflation_rate in EXPERIMENTAL_protocol_config, trigger > 1.60%, now 2.50% (BREACHED).

Post-incentive volume decay (NEAR@3.33 unwind). Public volume −60% and confidential −70% puts the deterministic FDV at $793.0M (−59.9%); watch nearstats 30d volume against its 30d record weekly, trigger <= −40%, now −22.0% (not breached).

Solver cartelization: HHI_solver > 3,500, 1Click loses the spread. HHI_solver > 3,500 with cross-VM headroom at −2 bps puts the deterministic FDV at $1.17B (−40.6%); watch the execute_intents solver HHI (now 3,197) and the live quote-race headroom (now +1.0 bps) weekly (inconclusive).

All four at once. Deterministic FDV $211.5M (−89.3% vs base $1.98B); 1 of 3 measured triggers currently breached (1 not graded this run, inconclusive or not measured: solver_cartelization), and the thesis is falsified if any trigger holds for 30 consecutive days.

Live values the triggers were graded against

Carried unchanged from the collectors. A missing value is a trigger that could not be evaluated this run, not a pass.

Metric	Value	What it measures	Source
Net take rate, 30 days	4.82 bps	Protocol revenue divided by swap volume over the last 30 complete days.	DefiLlama
Solver concentration (HHI)	3,197	Concentration of fills across solver accounts; the cartelization trigger reads this number.	NearBlocks
Effective number of solvers	3.1	10,000 divided by solver HHI: how many equal-sized solvers the concentration corresponds to.	NearBlocks
Solver fills attributed	122	Fill participations matched to a solver account; the trigger is graded only above a minimum sample.	NearBlocks
Fills sampled	114	execute_intents fills inspected for solver attribution.	NearBlocks
Deposit-source concentration (HHI)	3,056	Concentration of the external accounts that call intents.near (bridge callbacks, direct callers). Not a solver measure; context only, it cannot fire the cartelization trigger.	NearBlocks
Deposit-source HHI, AI agents excluded	2,222	The same caller concentration with AI-agent accounts removed.	NearBlocks
External receipts sampled	12	Receipts in the NearBlocks sample the caller concentration is computed on.	NearBlocks
Sample time span	16,268.9 s	Wall-clock time covered by the sampled receipts.	NearBlocks
Cross-VM execution headroom	+0.98 bps	Median amount by which 1Click out-delivers the best competitor on ETH USDC to SOL USDC clips.	1Click
Protocol inflation cap	2.50%	Maximum annual issuance currently set on-chain.	NEAR RPC
Intents volume, 30 days	$3.77B	NEAR Intents swap volume over the last 30 days.	NearStats
Record 30-day volume	$4.84B	Highest 30-day volume NEAR Intents has recorded.	NearStats
Confidential TVL	$132.6M	Value held in confidential (shielded) execution, as reported by NearStats.	NearStats
Confidential TVL verified at source	Yes	False when NearStats marks its confidential TVL observation as unverified.	NearStats

Stress parameter register

How hard each scenario pushes. Every severity is an assumption with a declared range and a one-line justification; none is presented as measured.

Parameter	Value	Plausible range	Basis	Why this value
Fee race: public take rate	0.01 bps	0.00 bps to 1.00 bps	Assumption	1 pip is the bare intents.near contract fee once third-party frontends route around near.com and its 25 bps / 10-20 bps partner fees (QUANT_PIPELINE_SPEC Model C.1, MODEL_DESIGN 4.3.1).
Fee race: confidential take rate	5.00 bps	0.00 bps to 10.00 bps	Assumption	Confidential execution repriced to 5 bps, a quarter of the 20 bps default, once public routing is free (MODEL_DESIGN 4.3.1).
Fee race: trigger on net take	3.00 bps	1.00 bps to 5.00 bps	Assumption	Early warning: DefiLlama near-intents 30d net take below 3 bps (live ~5.6 bps class) means compression is already past the mid-point to zero (MODEL_DESIGN 4.3.1).
Volume decay: public volume multiplier	×0.40	×0.10 to ×1.00	Assumption	Public volume -60%: QUANT_PIPELINE_SPEC Model C.3 sizes the NEAR@3.33 mercenary-liquidity unwind at -60% of 30d volume.
Volume decay: confidential volume multiplier	×0.30	×0.10 to ×1.00	Assumption	Confidential volume -70%: the shard's TVL is flagged unverified by nearstats, so its volume is assumed more incentive-dependent than public flow (MODEL_DESIGN 4.3.3).
Volume decay: confidential TVL multiplier	×0.30	×0.10 to ×1.00	Assumption	Confidential TVL -70% (QUANT_PIPELINE_SPEC Model C.3).
Volume decay: trigger on volume versus record	-40%	-80% to -10%	Assumption	Early warning: 30d volume more than 40% below the 30d record (nearstats records.max30dUsd) (MODEL_DESIGN 4.3.3).
Cartelization: concentration level	3,500	2,500 to 10,000	Assumption	HHI 3,500 is two actors at ~40% each plus a tail; the US horizontal-merger guideline calls > 2,500 highly concentrated (QUANT_PIPELINE_SPEC Model C.4).
Cartelization: cross-VM headroom	-2.00 bps	-10.00 bps to 0.00 bps	Assumption	Cartel spreads widen until Relay/LI.FI deliver 2 bps better net execution on cross-VM; 1Click must give the same 2 bps of take back to keep flow (MODEL_DESIGN 4.3.4).
Take-rate floor	0.50 bps	0.00 bps to 2.00 bps	Assumption	Floor on the cartel scenario's public take give-back: max(0.5 bps, live public take - 2 bps) (MODEL_DESIGN 4.2 competitive floor).
Cartelization: fee compression rate	0.15 per month	0.00 per month to 0.15 per month	Assumption	Compression speed set to the registry's high bound for fee_compression_lambda_monthly (0.15/month): losing the spread means order flow leaves at the fastest plausible rate.

Method and sources

Value basis. base_fdv_p50 / sotp_fdv_p50 / combined_worst_fdv keep their field names for compatibility but are DETERMINISTIC (median-input) FDVs from deterministic_sotp, not Monte Carlo percentiles; mc_* fields carry the seeded MC medians when computed

Change definition. drawdown_pct = fdv_scenario / fdv_base - 1 (signed fraction) on the deterministic (median-input) FDV; price_drawdown_pct likewise on implied price; mc_drawdown_pct on the Monte Carlo median when computed

Snapshot fields: base bear_case.base_fdv_p50, base_price_p50, base_crossover_month; combined combined_worst_fdv, combined_drawdown_pct, combined_worst_price, combined_price_drawdown_pct, combined_perturbations (unapplied: combined_perturbations_unapplied); per scenario bear_case.scenarios[].sotp_fdv_p50, drawdown_pct, price_p50, price_drawdown_pct, crossover_month, fdv_ex_monetary, mc_fdv_p50; trigger counts bear_case.triggers_breached, triggers_evaluated, triggers_unevaluated; prose bear_case.falsification_summary; register bear_case.stress_parameters. Spot FDV is sotp.current_fdv_usd; the Monte Carlo median is sotp.fdv_percentiles.p50.

Scenario keys and the field each trigger reads:

fee_race_to_zero: Net take rate, 30 days (net_take_bps_30d), read as sum(dailyRevenue chart) / sum(summary/dexs/near-intents chart) x 1e4 over the same 30 complete UTC days.
governance_stalemate: Protocol inflation cap (max_inflation_rate), read as EXPERIMENTAL_protocol_config.max_inflation_rate (num/den).
post_incentive_decay: 30-day volume versus its 30-day record (volume_30d_vs_record_30d), read as intents.totals.volume30dUsd / intents.totalsBreakdown.records.max30dUsd - 1.
solver_cartelization: Solver concentration (HHI) (hhi_solver), read as HHI over solver fill participations decoded from execute_intents args.signed[].payload.message (NEP-413 signer_id, token_diff legs); secondary: headroom_by_archetype_bps.cross_vm (destination units); secondary headroom_cross_vm_bps.

Live metrics, read from bear_case.live_metrics:

net_take_bps: DefiLlama near-intents 30d dailyRevenue / 30d DEX volume × 1e4
hhi_solver: HHI over fill participations per solver, decoded from execute_intents args.signed[].payload.message (NEP-413 signer_id, token_diff legs); intents_forensics.solver_concentration.hhi_solver
n_eff_solver: 10,000 / hhi_solver
solver_fills_attributed: count of solver-attributed fill legs
solver_fills_sampled: count of execute_intents calls decoded
hhi_external: HHI over external predecessor_account_id receipt shares on intents.near, system and self excluded; fills are relayer self-receipts
hhi_external_ex_ai_agent: deposit/caller HHI excluding *.ai.near and *.agent.near callers
external_receipts_sampled: count of external receipts in the NearBlocks page
sample_span_s: wall-clock span of the sampled receipts (seconds)
headroom_cross_vm_bps: median (1Click out − best competitor out) / best, ETH USDC → SOL USDC clips with ≥ 1 competitor quote
max_inflation_rate: EXPERIMENTAL_protocol_config.max_inflation_rate (num / den)
volume_30d_usd: nearstats intents.totals.volume30dUsd
record_30d_usd: nearstats intents.totalsBreakdown.records.max30dUsd
confidential_tvl_usd: nearstats intents.protocolMetrics.confidentialTvlUsd
confidential_tvl_verified: nearstats confidentialTvlObservation.sourceStatus != 'unverified'

Register keys: bear_fee_race_tau_pub_bps (fee race: public take rate), bear_fee_race_tau_conf_bps (fee race: confidential take rate), bear_fee_race_trigger_net_take_bps (fee race: trigger on net take), bear_decay_v_pub_multiplier (volume decay: public volume multiplier), bear_decay_v_conf_multiplier (volume decay: confidential volume multiplier), bear_decay_tvl_multiplier (volume decay: confidential tvl multiplier), bear_decay_trigger_volume_drawdown (volume decay: trigger on volume versus record), bear_cartel_hhi_external (cartelization: concentration level), bear_cartel_headroom_bps (cartelization: cross-vm headroom), bear_tau_floor_bps (take-rate floor), bear_cartel_lambda_monthly (cartelization: fee compression rate).

Perturbation inputs shown struck through in the scenario table are listed by the model but have no slot in the deterministic engine, so they do not move the value.

History

1Click out-delivered the best competitor on cross-VM routes by a median 0.98 bps across 91 races

6 canonical snapshots recorded from Nov 1, 2026 to Nov 13, 2026 (11.9 days) and 92 live quote races over 5.3 days. Cross-VM headroom ranged from +0.98 bps to +2.42 bps. The 90% bootstrap interval of its median is +0.98 bps to +0.98 bps.

Cross-VM headroom, latest race

+0.98 bps

unchangedsince first of 91 races

median +0.98 bps

Native BTC headroom, latest race

-4.84 bps

+0.72 bpssince first of 91 races

median -6.86 bps

EVM L2 headroom, latest race

-1.02 bps

unchangedsince first of 91 races

median -1.02 bps

Competitive fee ceiling, latest race

3.00 bps

unchangedsince first of 92 races

best competitor's cross-VM friction

Solver HHI, latest snapshot

3,197

+1,588since first of 6 runs

90% CI 2,898 to 4,772

Sink coverage, latest snapshot

21.6%

+2.9%since first of 6 runs

sinks / issuance

SOTP P50 FDV, latest snapshot

$644.7M

+$386.2Msince first of 6 runs

10,000-path Monte Carlo median

Market FDV, latest snapshot

$4.34B

-$1.41Bsince first of 6 runs

P50 is -85% vs market

Hourly quote race · win/loss by route

1Click won 45% of 819 contested races across 3 routes over 5.3 days (90% CI 42% to 48%)

It won every decided race on ETH USDC to SOL USDC and lost most races on ETH USDC to native BTC (won 27%, beaten mostly by Relay) and Base USDC to Arbitrum USDC (won 8%, beaten mostly by Across). A race is one route and clip in one hourly run; the three clips of a route are pooled.

Route	Races	1Click won	90% CI	Typical margin	Winner when 1Click loses	Span
ETH USDC to SOL USDC Cross-VM · clips $500, $50K, $250K	273 (+3 excl.)	100% (273–0)	99% to 100%	wins by 0.98 bps	never lost	over 5.3 days
ETH USDC to native BTC Native BTC · clips $500, $50K, $250K	273 (+3 excl.)	27% (73–200)	23% to 31%	wins by 10.12 bpsloses by 12.58 bps	Relay 188 of 200 (LI.FI 12)	over 5.3 days
Base USDC to Arbitrum USDC EVM L2 · clips $500, $50K, $250K	273 (+3 excl.)	8% (22–250, 1 tied)	6% to 11%	wins by 0.24 bpsloses by 1.02 bps	Across 229 of 250 (Relay 21)	over 5.3 days

Hourly quote race

Headroom per route archetype

Median cell headroom of 1Click over the best competing quote, by archetype, in basis points. Above zero, 1Click delivers more output.

Hourly quote race

Competitive fee ceiling

Median friction of the best competitor on cross-VM routes: the fee a router can charge before a rival quote beats it.

Hourly quote race · per route and clip

Where 1Click wins, and how consistently

Each row pools every recorded race for one route and clip size. Headroom is 1Click's output against the best competing quote in the same race; positive means 1Click delivers more.

Route	Clip	Races	1Click quoted	1Click ranked first	Best competitor	Median headroom	90% CI of median	Range	Competitors not quoting
ETH USDC to SOL USDC Cross-VM	$500	92	99% (91/92)	99% (91/92)	Relay 92/92	+1.41 bps	+1.30 bps to +1.54 bps	+1.2 bps to +2.6 bps	Across: unsupported route 86/92, error 6/92
ETH USDC to SOL USDC Cross-VM	$50K	92	99% (91/92)	99% (91/92)	Relay 92/92	+0.98 bps	+0.98 bps to +0.98 bps	+1.0 bps to +2.4 bps	Across: unsupported route 86/92, error 6/92
ETH USDC to SOL USDC Cross-VM	$250K	92	99% (91/92)	99% (91/92)	Relay 92/92	+0.98 bps	+0.98 bps to +0.98 bps	+1.0 bps to +2.0 bps	Across: unsupported route 86/92, error 6/92
ETH USDC to native BTC Native BTC	$500	92	99% (91/92)	2% (2/92)	Relay 89/92, LI.FI 3/92	-18.47 bps	-20.72 bps to -17.02 bps	-52.6 bps to +1.1 bps	Across: unsupported route
ETH USDC to native BTC Native BTC	$50K	92	99% (91/92)	12% (11/92)	Relay 80/92, LI.FI 12/92	-6.73 bps	-8.41 bps to -5.56 bps	-60.2 bps to +23.7 bps	Across: unsupported route; Relay: no liquidity 4/92
ETH USDC to native BTC Native BTC	$250K	92	99% (91/92)	65% (60/92)	Relay 49/92, LI.FI 42/92	+7.87 bps	+2.38 bps to +9.98 bps	-22.9 bps to +22.1 bps	Across: unsupported route; LI.FI: skipped 1/92; Relay: no liquidity 41/92
Base USDC to Arbitrum USDC EVM L2	$500	92	99% (91/92)	24% (22/92)	Across 85/92, Relay 6/92	-0.68 bps	-0.96 bps to -0.38 bps	-1.0 bps to +0.9 bps	Across: error 6/92, skipped 1/92; LI.FI: skipped 1/92; Relay: skipped 1/92
Base USDC to Arbitrum USDC EVM L2	$50K	92	99% (91/92)	0% (0/92)	Across 85/92, Relay 6/92	-1.02 bps	-1.02 bps to -1.02 bps	-1.0 bps to -0.0 bps	Across: error 6/92, skipped 1/92; LI.FI: skipped 1/92; Relay: skipped 1/92
Base USDC to Arbitrum USDC EVM L2	$250K	92	99% (91/92)	0% (0/92)	Across 82/92, Relay 9/92	-1.02 bps	-1.02 bps to -1.02 bps	-1.0 bps to -0.0 bps	Across: error 6/92, skipped 1/92, no liquidity 3/92; LI.FI: skipped 1/92; Relay: skipped 1/92

Daily canonical snapshot

Solver concentration

HHI over solver fill participations on intents.near, against the 3,500 cartelization trigger. The band is the 90% bootstrap interval when recorded.

Daily canonical snapshot

Sink coverage

Buybacks, gas burn and storage lock as a share of daily issuance. At 100% the float stops growing.

Daily canonical snapshot

Model value against the market

SOTP Monte Carlo P50 FDV with its P10 to P90 range, and the market FDV recorded in the same run.

Daily canonical snapshot · bear-case triggers

1 of 4 falsification triggers breached in the latest run

Rule as recorded in the summary: thesis falsified if a trigger holds for 30 consecutive days. A streak runs from the first daily record of the current unbroken breach to the latest one, so a breach seen once reads 0.0 days. Daily records more than 36 h apart break it, as does a day the trigger was not measured.

Trigger	Latest value	Status now	Breached since	Consecutive days breached	Runs breached
Aggregator bypass: public take races to 1 pip Net take rate, 30 days < 3.00 bps	4.82 bps	Not breached	not in a breach run	clear	0 / 6
Governance stalemate: 2.5% issuance forever Protocol inflation cap > 1.60%	2.50%	Breached	2026-11-10 13:04 UTC	breached 2.7 of 304 daily records; restarted after a data gap	6 / 6
Post-incentive volume decay (NEAR@3.33 unwind) 30-day volume versus its 30-day record <= -40.0%	-22.0%	Not breached	not in a breach run	clear	0 / 6
Solver cartelization: HHI_solver > 3,500, 1Click loses the spread Solver concentration (HHI) > 3,500	3,197	Not graded	not in a breach run	not measured	4 / 6

On the day a streak reaches 30 days the daily job writes valuo/falsification/<trigger> to NEAR mainnet SocialDB under 0161aff7…81f0 (make falsification-list ANCHOR=<account> reads it back without a key).

Backtest: volume engine, out of sample

The volume model's P10 to P90 band held 55% at 1 month, 53% at 3 months and 50% at 6 months of realized 30-day volume, against 80% if calibrated

20 monthly origins from 2025-03-22 to 2026-10-13, each calibrated only on DefiLlama NEAR Intents daily volume up to that day (2024-12-10 to 2026-11-12 in all). One month ahead the model under-predicted volume: the median realized 30-day volume was 1.37x the forecast median. One-month coverage was 20% for origins before 2026-01-16 and 90% from it. Only the volume distribution is tested; take rate, capture, confidential share and multiples are not.

P10 to P90 coverage, 1 month ahead

55%

11 of 20 inside; 80% rejected (p 0.01)

P10 to P90 coverage, 3 months ahead

53%

9 of 17 inside; 80% rejected (p 0.01)

P10 to P90 coverage, 6 months ahead

50%

7 of 14 inside; 80% rejected (p 0.01)

Realized / forecast median, 1 month

1.37x

3 months 1.84x; 6 months 3.70x

1 month ahead

Forecast band vs realized, 1 month

30-day NEAR Intents volume at each target date, log scale. Shaded: the P10 to P90 band drawn at the origin; dashed: its median.

3 months ahead

Forecast band vs realized, 3 months

30-day NEAR Intents volume at each target date, log scale. Shaded: the P10 to P90 band drawn at the origin; dashed: its median.

6 months ahead

Forecast band vs realized, 6 months

30-day NEAR Intents volume at each target date, log scale. Shaded: the P10 to P90 band drawn at the origin; dashed: its median.

Coverage, bias and PIT counts per horizon. A calibrated model puts about 80% inside the band and a flat PIT histogram.
Sample	Horizon	n	P10 to P90 coverage	Binomial p vs 80%	Below P10 / above P90	Realized / median	PIT histogram (5 bins)
All origins	1 month	20	55% [37% to 72%]	0.01	0 / 9	1.37x	1 4 2 2 11
All origins	3 months	17	53% [34% to 71%]	0.01	0 / 8	1.84x	0 2 2 4 9
All origins	6 months	14	50% [30% to 70%]	0.01	0 / 7	3.70x	1 0 3 1 9
Origins before 2026-01-16	1 month	10	20% [7% to 46%]	0.00	0 / 8	1.66x	0 2 0 0 8
Origins before 2026-01-16	3 months	10	20% [7% to 46%]	0.00	0 / 8	4.49x	0 1 1 0 8
Origins before 2026-01-16	6 months	10	30% [13% to 56%]	0.00	0 / 7	7.28x	1 0 1 1 7
Origins from 2026-01-16	1 month	10	90% [65% to 98%]	0.70	0 / 1	1.09x	1 2 2 2 3
Origins from 2026-01-16	3 months	7	100% [72% to 100%]	0.36	0 / 0	1.48x	0 1 1 4 1
Origins from 2026-01-16	6 months	4	100% [60% to 100%]	1.00	0 / 0	1.92x	0 0 2 0 2
Variant: trailing measured drift	1 month	20	60% [42% to 76%]	0.04	6 / 2	0.81x	8 4 0 3 5
Variant: trailing measured drift	3 months	17	47% [29% to 66%]	0.00	7 / 2	0.59x	9 4 0 1 3
Variant: trailing measured drift	6 months	14	21% [9% to 43%]	0.00	10 / 1	0.23x	12 0 0 0 2

Second engine, same out-of-sample test

Coverage against 80%: block bootstrap rejected at 1 of 3 horizons, GBM at 3 of 3

Stationary block bootstrap of realized daily log-changes (10,000 paths, seed 20261114), scored on the same origins and realized values as the GBM. It is reported next to the base case; the valuation does not use it.

The block bootstrap does not repair the volume engine: P10-P90 coverage (1m 18/20 (90%, p 0.40), 3m 17/17 (100%, p 0.03), 6m 13/14 (93%, p 0.33)) is rejected against 80% for over-coverage at 3m; the pass at 1m and 6m rests on about 20 origins, which is weak evidence; its band is 3.3x to 7.6x the GBM's log width, so it covers by being wide, not by being sharp; the GBM is rejected at 3 of 3 horizons (1m 11/20 (55%, p 0.01), 3m 9/17 (53%, p 0.01), 6m 7/14 (50%, p 0.01)), so the two engines bracket realized dispersion and neither is calibrated.

P10 to P90 coverage per horizon against the 80% target, exact two-sided binomial test. Both engines are scored on the same forecasts.
Horizon	GBM: inside band	GBM: p vs 80%	GBM: verdict	Bootstrap: inside band	Bootstrap: p vs 80%	Bootstrap: verdict	Bootstrap: realized / median	Band width, bootstrap / GBM
1 month	11 of 20 (55%)	0.01	rejected, band too narrow	18 of 20 (90%)	0.40	not rejected	0.61x	3.3x
3 months	9 of 17 (53%)	0.01	rejected, band too narrow	17 of 17 (100%)	0.03	rejected, band too wide	0.41x	5.7x
6 months	7 of 14 (50%)	0.01	rejected, band too narrow	13 of 14 (93%)	0.33	not rejected	0.06x	7.6x

Bootstrap shape at the GBM's expected path: 1 month 12 of 20 (60%), p 0.04; 3 months 6 of 17 (35%), p <0.01; 6 months 0 of 14 (0%), p <0.01.

Sensitivity, cube-root block length: 1 month 20 of 20 (100%), p 0.02; 3 months 17 of 17 (100%), p 0.03; 6 months 14 of 14 (100%), p 0.09.

Power: with 20 one-month forecasts the test accepts 14 to 19 inside. An engine whose true coverage is 60% is rejected 75% of the time and one at 70% 39% of the time, so a pass on this sample is weak evidence.

What the engine choice does to the valuation inputs, from the 2026-11-12 origin. Headline figures use the GBM rows.
Measure	Volume engine	P10	P50	P90
30-day volume in 12 months, multiple of today's	GBM (used by the valuation)	0.19x	0.86x	3.88x
30-day volume in 12 months, multiple of today's	Block bootstrap, same expected path	0.09x	0.57x	3.73x
30-day volume in 12 months, multiple of today's	Block bootstrap, resampled drift	0.33x	2.06x	13.45x
Model FDV, 36-month run	GBM (used by the valuation)	$126.5M	$644.7M	$3.92B
Model FDV, 36-month run	Block bootstrap, same expected path	$46.59M	$293.1M	$2.78B

The valuation uses the GBM. At the same expected path the bootstrap puts 30-day volume 12 months out at 0.09x / 0.57x / 3.73x of current (P10 / P50 / P90) against the GBM's 0.19x / 0.86x / 3.88x, a band 1.2x as wide in log terms; through the same Monte Carlo FDV P10 / P50 / P90 becomes 0.37x / 0.45x / 0.71x of the base. The backtest rejects the bootstrap for over-coverage, so read this as a stress bound on volume dispersion (with the mean path held fixed, a wider distribution lowers the median), not as a better central estimate.

Block bootstrap: rule, limits and what was not tuned

Stationary block bootstrap (Politis-Romano: geometric block lengths, circular wrap) of realized daily log-changes of volume over the trailing 365 days up to the origin. Mean block length: Politis-White (2004) automatic selector with the Patton-Politis-White (2009) correction on that lookback, capped at min(3 sqrt(n), n/3); rule fixed from the series' autocorrelation before any coverage was computed. 10,000 paths (the valuation Monte Carlo's count), seed 20261114 mixed with the origin timestamp. Each path continues the daily series from the origin day; the forecast is the sum of the last 30 days of the path at the horizon, scored against the same realized 30-day sums as the GBM.

Block length rule politis_white, fixed before any coverage was computed; mean block 7.2 to 58.0 days across origins (median 11.5). Lookback 365 days, the same window the GBM sigma uses. Grading: exact two-sided binomial test of P10-P90 coverage vs 80%, p >= 0.10

Nothing in the bootstrap is tuned to reach 80% coverage. With about 20 origins, choosing block length, lookback or path count by coverage would be overfitting; the block rule and path count were fixed first and the n^(1/3) block rule is shown only as a sensitivity.

Low power: the exact binomial test on about 20 one-month origins accepts a wide range of true coverage, and 3- and 6-month forecasts overlap, so a pass is weak evidence of calibration, not proof.

A bootstrap can only replay changes that occurred in the lookback. It widens the band where the past year was volatile and cannot anticipate a regime the lookback did not contain.

Daily volume carries day-to-day noise that reverses. Blocks shorter than that reversal overstate horizon variance, so part of any coverage gain can come from a wider band, not a better-shaped one; the band-width ratio is reported for that reason.

The path starts at the origin day's volume, one noisy observation, and the registry-drift form holds the mean path fixed, so a wider distribution mechanically lowers its median. Both are stated properties of the design, not fitted choices.

No second design was tried after these results. Resampling changes of the 30-day sum instead of daily volume would remove the day-to-day noise, but choosing between engine designs on about 20 origins is the tuning this test rules out; it is left as a stated next test.

The valuation still runs on the GBM. The bootstrap is a parallel variant; no headline figure uses it.

Backtest method: what was and was not tested

Origins every 30 days, anchored at the last complete day of defillama:dexs/near-intents totalDataChart (raw/defillama__dexs_near-intents.json) (SHA-256 3a66e2dab20191e6..., the same payload recorded in artifacts/manifest.json). At each origin the live calibration is rerun on the days up to the origin only: starting volume is the trailing 30-day sum; the monthly volume sigma comes from intents_forensics.volume_volatility over the trailing 365 days through sotp_valuation._sigmas; the drift is the run's labeled public_volume_growth_yoy (72% a year) with the Ito correction of simulate_paths. The 30-day volume h months ahead is lognormal under that GBM, so the P10, P50 and P90 are closed-form (a unit test checks them against simulate_paths with seed 20261114). The realized value is the 30-day sum ending round(h x 30.44) days after the origin.

Grading: exact two-sided binomial test of P10-P90 coverage vs 80%, p >= 0.10. PIT is the forecast CDF at the realized value; a calibrated model gives a flat histogram, a pile-up in the top bin means realized volume ran above the forecast. The measured-drift variant replaces the registry growth (chosen with the whole history in view) with the trailing mean monthly log return at each origin, so it has no look-ahead.

Not tested: fee take rate (public and confidential) and its compression toward the floor; holder capture ratio alpha; confidential share of volume and its drift to the target share; terminal FDV/revenue multiple and discount rate; NEAR price and supply paths.

Forecasts at h = 3 and 6 months from monthly origins overlap, so their errors are serially correlated: the effective sample is smaller than n and the binomial interval is too narrow for those horizons.

public_volume_growth_yoy is a registry assumption chosen with the full history in view; the 'measured drift' variant removes that look-ahead by using only the trailing mean monthly log return.

The series is total NEAR Intents volume; the live model splits it into public and confidential legs with the same sigma, so the split itself is not tested.

Method, sources and how to add points

Everything on this tab is read from artifacts/history/summary.json (schema valuo.history.summary/1, rebuilt 2026-11-14 04:23 UTC), which quant/valuo/history.py derives from two append-only logs: snapshots.jsonl, one record per canonical pipeline run (de-duplicated by snapshot SHA-256), and bench.jsonl, one record per live quote race. Each race gzips its raw venue responses under artifacts/history/raw/<utc>/ with the SHA-256 of the uncompressed bytes; make history-verify re-hashes them.

The daily series keeps the latest canonical run per UTC day: 6 days shown from 11 recorded runs; 5 same-day re-runs are kept in snapshots.jsonl but not plotted.

Schedule, as committed: the canonical snapshot runs daily at 06:05 UTC (cron "5 6 * * *") (ops/vps/valuo.crontab); the quote race runs hourly at minute 23 (cron "23 * * * *") (ops/vps/valuo.crontab). Both jobs run on a VPS (ops/vps/setup.sh installs them) and commit their output back to the repository, so the series grows by one point per day and one race per hour.

Statistics: 90% percentile-bootstrap CI of the median (2000 resamples, seed 20261114); null when n < 5. Medians, ranges and changes are over non-null observations in time order; a run that did not measure a metric leaves a gap rather than a zero. Race headroom per archetype is the median over cells with a 1Click quote and at least one competitor quote; snapshot series are the values of the canonical latest_snapshot.json of each day. Trigger streaks use the trigger's own breach flag from bear_case.scenarios in each run; a streak is continuous only while consecutive daily records are at most 36 h apart, and a record where the trigger was not measured ends it.

Add points locally: make bench-history runs one live race (all corridors and clips, zero API keys) and appends it; make history appends the current canonical snapshot; make history-summary rebuilds summary.json. Reload this page to see them.

Ask the analyst · NEAR AI

Questions answered only from this snapshot, with every claim cited

Each question retrieves numbered facts from latest_snapshot.json and IC_MEMO.md, sends only those to a model on NEAR AI Cloud, and maps every citation back to its JSON path and SHA-256-hashed raw payload. If the snapshot does not contain the answer, the analyst is instructed to say so.

NEAR AI Cloud

Valuo NEAR Investment Committee terminal

4 of 8 conditions for NEAR's long-term case are on track today

NEAR's $4.34B market value needs 133% annual volume growth or a 56.3 bps take rate on Intents cash flow alone; only a 135× multiple closes the gap inside the evidence

What the market is pricing

What the cash-flow model values

Valuation distribution

Supply crossover by issuance regime

How the thesis is built

Step 1: Execution headroom

Step 2: Take rate and stickiness

Step 3: Supply sinks against issuance

Step 4: Valuation

Evidence scorecard

Three valuation engines

Caveats

Pipeline warnings (15)

Failed raw fetches (26, recorded, not substituted)

1Click out-delivers the best competitor by +0.98 bps on cross-VM routes and trails on native BTC and EVM L2

Live quote race

Committed benchmark

USDC on Ethereum to USDC on Solana

USDC on Ethereum to BTC on Bitcoin

USDC on Base to USDC on Arbitrum

3 of 3 venues quote $250K on ETH USDC to SOL USDC; 1Click is cheapest at 2.02 bps

Cost at every size, ETH USDC to SOL USDC

NEAR Intents collects 20.4% of the six-protocol cross-chain fee pool, rank 2 of 6, and its share is falling

Gross take compressed over the trailing year

Who settles on intents.near

Sinks offset 21.6% of the 89,757 NEAR minted per day; the float grows by about 70,377 NEAR a day

Today's flows, NEAR per day

Chain Signatures handles about 12,918 sign requests a day; the 30-day sample shows no significant trend

The median Monte Carlo path values NEAR at $644.7M, 85.2% below its current $4.34B FDV

What you have to believe

Contribution by engine

Running implied FDV

Cumulative holder revenue

When does net supply stop growing?

Relative value

Peer-multiple cross-check, not a fair value

Staking and the security budget

The 60x terminal multiple cap sits at the 13th percentile of 15 peers (2 of 15 trade at or below it) on FDV to annualised 30-day revenue

1 of 3 measured early-warning triggers is breached; all four stresses together cut the deterministic value by 89%.

Conditions under which the thesis is wrong

1Click out-delivered the best competitor on cross-VM routes by a median 0.98 bps across 91 races

The volume model's P10 to P90 band held 55% at 1 month, 53% at 3 months and 50% at 6 months of realized 30-day volume, against 80% if calibrated

Questions answered only from this snapshot, with every claim cited