Can Intents create
lasting value for NEAR?
The accounts we tracked converted assets into NEAR every day for 90 days. How much value the protocol earns and retains remains open.
Observation window
6 Aug–3 Nov 2026 UTC
Accounts linked to three treasury wallets
Executive summary
The thesis. If growing Intents activity generates lasting protocol revenue and some of that revenue reaches the token, it could support long-term value for NEAR.
We observed a recurring conversion channel. Purchases fell in NEAR terms and were small beside supply growth. Fee records cover only part of the funding path, so the investment case remains conditional.
Fewer tokens in each 30-day window
The cohort acquired less NEAR in each window. Higher daily prices offset much of the drop in token quantity when we valued the final window in dollars.
Acquisitions by 30-day window
The cohort acquired 31.35% less NEAR in the third window than in the second.
Values and valuation method
| Window | NEAR acquired | Est. USD |
|---|---|---|
| 06 Aug–04 Sept | 244,845.70 | $437,374 |
| 05 Sept–04 Oct | 165,809.44 | $557,914 |
| 05 Oct–03 Nov | 113,825.00 | $547,026 |
NEAR quantities are rounded as recorded. The USD estimates use daily start-of-day prices. Using end-of-day prices gives a 5.45% decline from the second window to the third.
Hypothesis assessment at a glance
Research question and canonical hypotheses
The thesis rests on four questions: whether activity grows protocol revenue, how revenue reaches the token, whether burns offset issuance, and how much value the protocol retains.
H1 — Unresolved
Revenue growth
Does increasing Intents activity correspond to sustainable protocol-revenue growth?
Daily economically attributable protocol revenue is not established (open question OQ-01): the preferred revenue specification is NOT_READY — the protocol-revenue metric is gated with zero observations. Gross-fee association alone is insufficient.
What would weaken it. Weakened if comparable net revenue stagnates or falls while activity rises after adjusting for policy, price and costs.
H2 — Partial evidence
Value capture
How much generated revenue translates into measurable buybacks or other capture?
Partial route evidence only: fee-funded conversions are established for a narrow cohort on narrow dates (CL-005, CL-006); the full fee-funded share of observed purchases is unknown because pooled balances prevent physical tracing.
What would weaken it. Weakened if executed fee receipts cannot be reconciled to conversion and retention, or if material outflows reverse the observed accumulation.
H3 — Not supported on current evidence
Supply offset
Under what conditions can buybacks and burns offset issuance?
Total supply grew over the measured window (CL-007) and purchases are not burns: the confirmed purchases are a small fraction of the net supply increase, and gross issuance/burn are known only as conservative lower bounds (CL-008).
What would weaken it. A supply-offset claim fails without actual destruction matching issuance; liquid-supply effects would require a separate verified retention/lock model.
H4 — Unresolved
Retention share
Does protocol retention share rise or fall with adoption?
A compatible revenue numerator and economic-activity denominator are missing: the capture-rate metric is gated and never emits a value from gross fees or cash payouts.
What would weaken it. Weakened if matched accrual retention falls persistently or if costs and incentives exceed retained revenue.
Ecosystem overview
Four economically distinct layers matter here: base-chain execution and gas economics; Intents-related product activity; distribution and affiliate economics; and governance-controlled accounts. A token can be the settlement or treasury asset without every dollar of ecosystem activity accruing to token holders.
The official tokenomics discussion describes discretionary governance tools; it is treated as policy context, not as a guaranteed cashflow claim.
Cross-chain execution and Intents
The relevant observed mechanism is: incoming multi-asset balances are exchanged into wrap.near, followed by native withdrawals arriving as treasury receipts, and subsequent holdings or use (CL-001, CL-003).
The fee configuration documents declared appFees; a successful on-chain receipt and economic entitlement must be proven separately (CL-005). Common pooled balances prevent physical tracing of interchangeable tokens from a particular fee to a particular payout — which is why CL-006 is a conservative accounting model, not token tracing.
The economic mechanism
The investment question is a chain of distinct steps, each of which can hold or break on its own evidence. Treating them as one number is the mistake this project exists to avoid.
Activity
People use applications on NEAR. Usage is observable on-chain, but raw transaction counts say nothing about economic value by themselves.
Fees
Usage produces fees. Fee amounts are exact on-chain facts and are the first quantity worth measuring precisely.
Revenue
Fees accrue somewhere — to validators, to a protocol treasury, to application contracts. Which part is economically attributable revenue is a definition question, answered per source with explicit boundaries.
Conversion
A treasury may convert holdings or purchase NEAR on the market. Purchases are observable events with their own provenance.
Retention & supply
Purchases can support demand; burns permanently remove supply. These are different effects and are never presented as one movement.
- Treasury purchases do not reduce total supply — burns do.
- Gross fees, attributable revenue and treasury payouts are different metrics; this project never relabels one as another.
Adoption and network activity
Provider volume can describe Intents turnover; it is not independently reconstructed market-wide activity (CL-004).
The measured cohort contains accounts that withdrew to three selected treasury wallets during the 90-day window. Collector accounts are not users, customers or independent integrators; their concentration says nothing about customer concentration.
Network transaction and active-address series stay separate and unavailable until they are defined and sourced — the gated registry emits null with a recorded reason, never zero.
The longer view
Explore the history behind the thesis
Follow Intents volume and gross fees across the available history. Observed NEAR purchases remain a separate 90-day cohort. Gross fees are not net protocol revenue.
Loading the selected history…
Revenue generation
Gross provider fees are not protocol profit (CL-004). A source-qualified fee/volume series can support an exploratory association only, with the mechanical fee–volume construction, provider revisions, incentive policy and endogenous adoption disclosed — the readiness gate admits exactly that named proxy, with limits.
Monthly captured revenue is not allocated across days. Provider sources disagree on overlapping days, and reported fee history has been revised after publication; the disagreement is recorded, not averaged away.
Revenue capture and actual observed findings
For 1–3 November, we matched fee records to on-chain transfers and internal credits. This supports the fee-to-conversion route for part of the activity in that period. How much of the full 90-day cohort’s purchases came from fees is still unknown.
CL-001 and CL-003 are different flows and must not be added as independent token demand. Even the route attribution in CL-006 is a conservative accounting model, not an exact causal estimate of market impact.
Compare NEAR and estimated USD across the three windows
Purchases throughout the study
The selected accounts converted other assets into wrap.near every day of the study. The totals sum successful on-chain events after removing duplicates.
Only accounts linked to the selected treasury wallets are included. This sample cannot measure market-wide demand or protocol revenue.
Data and sources for CL-001
CL-001 — Observed — cohort-scoped
CL-001 recorded quantities, scopes and sources Recorded quantity Scope Source 54,991 successful observed conversion events 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive), accounts that withdrew to the three selected treasury wallets Recorded sourcedocs/INVESTMENT_THESIS.md CL-001; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) 524,480.127692882832410490617319 NEAR acquired 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive), accounts that withdrew to the three selected treasury wallets Recorded sourcedocs/INVESTMENT_THESIS.md CL-001; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) 90 of 90 days contained at least one observed conversion 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-001; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) Definition versions: conversion_count v1 (availability A); observed_near_purchases v1 (availability A)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Fewer NEAR in each window
The accounts bought fewer NEAR in each successive window. The conversion route continued operating, but its token volume declined.
Window totals are rounded. In the last two windows, dollar estimates changed much less than token quantities.
Data and sources for CL-002
CL-002 — Observed — contrary to an accelerating-demand narrative
CL-002 recorded quantities, scopes and sources Recorded quantity Scope Source 244,845.70 NEAR acquired First 30-day window, 2026-08-06 through 2026-09-04 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-002; research/near-dollar/FINDINGS.md (verified 4 November 2026) 165,809.44 NEAR acquired Second 30-day window, 2026-09-05 through 2026-10-04 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-002; research/near-dollar/FINDINGS.md (verified 4 November 2026) 113,825.00 NEAR acquired Third 30-day window, 2026-10-05 through 2026-11-03 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-002; research/near-dollar/FINDINGS.md (verified 4 November 2026) 31.35 percentDecreaseBasis: third window relative to the preceding window, in NEAR quantities Third window relative to the second window Recorded sourcedocs/INVESTMENT_THESIS.md CL-002; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) Definition versions: observed_near_purchases v1 (availability A)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Daily treasury payouts
Native NEAR payouts from the Intents contract reached the three selected wallets every day. Transfers between those wallets are excluded.
Payouts were matched by amount, recipient and a unique candidate within 300 seconds, rather than a full receipt graph. Do not add them to purchases as separate demand.
Data and sources for CL-003
CL-003 — Observed — matched flows, not provenance
CL-003 recorded quantities, scopes and sources Recorded quantity Scope Source 1,405,659.135989071036891726256814 NEAR in matched native payouts 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive), three selected treasury wallets, internal treasury transfers excluded Recorded sourcedocs/INVESTMENT_THESIS.md CL-003; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) 3,165 matched native payout events 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-003; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) 502,794 NEAR of internal treasury transfers excluded Same 90-day window; transfers between the selected wallets are not counted as new payouts Recorded sourcedocs/INVESTMENT_THESIS.md CL-003; research/near-longitudinal/FINDINGS.md (verified 4 November 2026) Definition versions: treasury_payouts_near v1 (availability A); treasury_payout_count v1 (availability A)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
A steadier dollar value
The estimated dollar value changed little between the last two windows. Higher price marks explain the arithmetic difference from falling NEAR quantities.
Daily price marks approximate value; they are not trade prices. Missing prices remain missing, and purchases are not added to payouts.
Data and sources for CL-004
CL-004 — Estimate — valuation, not execution
CL-004 recorded quantities, scopes and sources Recorded quantity Scope Source 437,374 USD (approximate acquisition value)Basis: daily start-of-day price marks First 30-day window, 2026-08-06 through 2026-09-04 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-004; research/near-dollar/FINDINGS.md (verified 4 November 2026) 557,914 USD (approximate acquisition value)Basis: daily start-of-day price marks Second 30-day window, 2026-09-05 through 2026-10-04 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-004; research/near-dollar/FINDINGS.md (verified 4 November 2026) 547,026 USD (approximate acquisition value)Basis: daily start-of-day price marks Third 30-day window, 2026-10-05 through 2026-11-03 (UTC, both dates inclusive) Recorded sourcedocs/INVESTMENT_THESIS.md CL-004; research/near-dollar/FINDINGS.md (verified 4 November 2026) 1.95 percentDecreaseBasis: start-of-day price marks Third window relative to the second window Recorded sourcedocs/INVESTMENT_THESIS.md CL-004; research/near-dollar/FINDINGS.md (verified 4 November 2026) 5.45 percentDecreaseBasis: end-of-day alternative price marks — a second valuation scenario, not a confidence interval Third window relative to the second window Recorded sourcedocs/INVESTMENT_THESIS.md CL-004; research/near-dollar/FINDINGS.md (verified 4 November 2026) 3.3648 USD quantity-weighted average daily priceBasis: quantity-weighted average of daily start prices, not an execution price Second 30-day window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 4.8059 USD quantity-weighted average daily priceBasis: quantity-weighted average of daily start prices, not an execution price Third 30-day window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 42.83 percentIncreaseBasis: change of the quantity-weighted average daily price between the second and third windows Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 0.68648 ratio of NEAR quantitiesBasis: third window relative to the second window — quantity factor of the recorded decomposition Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 1.42828 ratio of weighted daily pricesBasis: third window relative to the second window — price factor of the recorded decomposition Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 0.98048 ratio of USD valuesBasis: product of the two factors above — the recorded decomposition shows the unit effect, not causation Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 12.54 percentDecreaseBasis: provider-reported Intents volume, fresh dashboard retrieval of 4 November 2026; revisable Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 21.88 percentDecreaseBasis: provider-reported all fees, fresh dashboard retrieval of 4 November 2026; revisable Third window relative to the second window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 10.06 percent of all reported feesBasis: descriptive ratio of observed purchase value to all reported fees — not a contracted rate and not a proven fee share First 30-day window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 7.37 percent of all reported feesBasis: descriptive ratio of observed purchase value to all reported fees — not a contracted rate and not a proven fee share Second 30-day window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) 9.25 percent of all reported feesBasis: descriptive ratio of observed purchase value to all reported fees — not a contracted rate and not a proven fee share Third 30-day window Recorded sourceresearch/near-dollar/FINDINGS.md (verified 4 November 2026) Definition versions: estimated_purchase_usd v1 (availability B — valuation estimate); near_price_usd v1 (availability P); provider_intents_volume_usd v1 (availability P); provider_gross_fees_usd v1 (availability P)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Which fees reached purchases
Some declared fees can be matched from the API record through an on-chain transfer to a unique internal credit. This establishes fee funding for part of the purchases on those dates; unmatched records remain unresolved.
This sample follows request-creation dates, not execution dates. It covers only declared appFees and cannot be extended to all fees, weekly purchases or the full study.
Data and sources for CL-005
CL-005 — Partial — narrow cohort
CL-005 recorded quantities, scopes and sources Recorded quantity Scope Source 21,361 API creation-cohort records Explorer API SUCCESS affiliate records with request-creation dates 2026-11-01 through 2026-11-03 (UTC) Recorded sourcedocs/INVESTMENT_THESIS.md CL-005; research/near-provenance/FINDINGS.md (verified 4 November 2026) 8,420 matched fee events (API record, logical transfer and unique canonical credit) Same creation cohort; execution dates 2026-11-01 through 2026-11-03 Recorded sourcedocs/INVESTMENT_THESIS.md CL-005; research/near-provenance/FINDINGS.md (verified 4 November 2026) 3,377 matches with exact-integer formula agreement Subset of the matched fee events where amountIn × bps / 10000 is an exact integer Recorded sourcedocs/INVESTMENT_THESIS.md CL-005; research/near-provenance/FINDINGS.md (verified 4 November 2026) Definition versions: none — this claim's values come from recorded analysis artifacts, not a registered metric
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
A minimum fee-funded amount
The inventory model finds a minimum amount of confirmed fees that must have been used in NEAR conversions. After subtracting closing stock, a minimum NEAR-equivalent must also have reached payouts under the model's rules.
These are accounting lower bounds for a short window. They do not trace individual tokens, extend to the full study or estimate market impact.
Data and sources for CL-006
CL-006 — Bound — model minimum
CL-006 recorded quantities, scopes and sources Recorded quantity Scope Source 1,427.13 NEAR equivalent sold out of confirmed feesLower bound (at least)Basis: strict short-window inventory model — a minimum within the stated model, not physical token tracing Confirmed fees executed 2026-11-01 through 2026-11-02 against the reconciled week 2026-10-27 through 2026-11-02 Recorded sourcedocs/INVESTMENT_THESIS.md CL-006; research/near-provenance/FINDINGS.md (verified 4 November 2026) 5.69 percent of the reference week's observed purchasesBasis: share of the week's observed purchases accounted for by the strict inventory minimum Relative to the observed purchases of the reconciled week 2026-10-27 through 2026-11-02 Recorded sourcedocs/INVESTMENT_THESIS.md CL-006; research/near-provenance/FINDINGS.md (verified 4 November 2026) 25,082.81 NEAR observed purchases in the reference week Reconciled week 2026-10-27 through 2026-11-02, the cohort baseline referenced by the share above Recorded sourceresearch/near-provenance/FINDINGS.md (verified 4 November 2026) 1,407.21 NEAR equivalent routed to payout in the modelLower bound (at least)Basis: modeled route minimum after subtracting all closing stock — a minimum within the stated model Same reconciled week and model Recorded sourcedocs/INVESTMENT_THESIS.md CL-006; research/near-provenance/FINDINGS.md (verified 4 November 2026) Definition versions: strict_fee_conversion_equivalent_near v1 (availability B — bound/model)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Token economics
Accounting identity for total supply: the net change equals gross issuance minus actual burns, plus any separately identified protocol adjustments. Ordinary purchases and treasury transfers are not a subtraction. Liquid circulating supply requires a distinct operational definition and verified locks or outflows; it is not measured here. Comparing acquisitions with issuance measures scale, not deflation or net selling pressure.
Recorded evidence · CL-007 / CL-001
Purchases stay small beside net supply growth
- Observed purchases: 90-day observed cohort, 2026-08-06 through 2026-11-04 (UTC, end exclusive), accounts that withdrew to the three selected treasury wallets524,480.127692882832410490617319 NEAR
- Net supply increase: Boundary difference over 2026-08-06 through 2026-11-03 (UTC day closes), verified on four exact boundary blocks7,936,181.547057263680300503519861 NEAR
Observed purchases are 6.60872 percent of the net supply increase; the net increase itself is an increase of 0.60915 percent of the opening supply. A scale comparison — not deflation, not a capture rate.
Purchases are never subtracted from total supply. Emission does not equal market selling pressure; recipients may hold or stake their rewards.
Purchases and supply growth
Total supply grew by much more than the observed purchases over the same period. The comparison shows their relative scale, rather than a capture rate or deflation.
Purchases leave total supply unchanged. Newly issued tokens may be held or staked, so issuance alone does not measure selling pressure.
Data and sources for CL-007
CL-007 — Observed — scale comparison
CL-007 recorded quantities, scopes and sources Recorded quantity Scope Source 7,936,181.547057263680300503519861 NEAR exact boundary total-supply increase Boundary difference over 2026-08-06 through 2026-11-03 (UTC day closes), verified on four exact boundary blocks Recorded sourcedocs/INVESTMENT_THESIS.md CL-007; research/near-supply/FINDINGS.md (verified 4 November 2026) 0.60915 percent of the opening supplyIncreaseBasis: rounded; an observed change over 90 days, not a forecast or an annualized rate Same 90-day window Recorded sourcedocs/INVESTMENT_THESIS.md CL-007; research/near-supply/FINDINGS.md (verified 4 November 2026) 6.60872 percent of the net supply increaseBasis: scale comparison of observed purchases with the net supply increase — not deflation and not a capture rate Observed purchases over the same 90-day window Recorded sourcedocs/INVESTMENT_THESIS.md CL-007; research/near-supply/FINDINGS.md (verified 4 November 2026) Definition versions: net_supply_change_near v1 (availability A); purchases_to_net_supply_pct v1 (availability A)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
What the incomplete audit establishes
The unfinished mint-header audit establishes minimum totals for issuance and burns. Unchecked blocks may contain more of both.
These are not exact gross totals. The table labels each lower or upper bound explicitly.
Data and sources for CL-008
CL-008 — Bounds — incomplete audit
CL-008 recorded quantities, scopes and sources Recorded quantity Scope Source 8,012,834 NEAR gross issuanceLower bound (at least)Basis: rounded down so the lower bound stays mathematically correct; simultaneous burns in the same blocks could hide further issuance Same 90-day window; positive supply changes plus verified simultaneous burn, less a conservative tolerance Recorded sourcedocs/INVESTMENT_THESIS.md CL-008; research/near-supply/FINDINGS.md (verified 4 November 2026) 76,652 NEAR burnsLower bound (at least)Basis: rounded down so the lower bound stays mathematically correct; simultaneous issuance in the same blocks could hide further burns Same 90-day window; negative supply changes plus verified simultaneous burn, less a conservative tolerance Recorded sourcedocs/INVESTMENT_THESIS.md CL-008; research/near-supply/FINDINGS.md (verified 4 November 2026) 6 of 297 positive supply-jump blocks with fully audited headers Archival header audit stopped after rate limiting; the remaining blocks are unverified, not assumed zero Recorded sourcedocs/INVESTMENT_THESIS.md CL-008; research/near-supply/FINDINGS.md (verified 4 November 2026) 6.54551 percent of gross issuanceUpper bound (at most)Basis: purchases divided by the strictly positive issuance lower bound, rounded outward — not an exact share Observed purchases over the same 90-day window Recorded sourcedocs/INVESTMENT_THESIS.md CL-008 (exact bound; the supply findings round it to about 6.55 percent) Definition versions: issuance_lower_bound_near v1 (availability B — lower bound); burns_lower_bound_near v1 (availability B — lower bound); purchases_to_issuance_upper_pct v1 (availability B — upper bound)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Treasury balances can be spent
The available index shows no outgoing receipt-actions from the selected wallets during the audited month. The funds were held as native balances under multisig or DAO control, where governance could decide to use them later.
The index cannot rule out every sale, lock or external delegation. Holding these balances does not establish permanent supply removal. The audit covers 2026-10-05 through 2026-11-03 (UTC).
Data and sources for CL-009
CL-009 — Scoped absence — retention, not removal
CL-009 recorded quantities, scopes and sources Recorded quantity Scope Source 4,140,077.36974892 NEAR combined native balance of the three walletsBasis: accumulated balances at a fixed finalized slice — not weekly revenue and not permanently removed supply Fixed RPC slice 2026-11-04 11:31:03 UTC at finalized block 223010763; storage collateral is included in the raw amounts Recorded sourceresearch/near-treasury/FINDINGS.md (verified 4 November 2026) Definition versions: treasury_balance_near v1 (availability A — saved point)
Dataset versions: Recorded analysis artifacts and manifests in the local research archive; approved application dataset versions staged by NI-028 are delivered read-only through the delivery APIs (the landing metric cards render the approved dataset key, version and coverage), but there is no public dataset catalog.
Econometric evidence
The available data support an exploratory comparison of volume and gross fees. The dataset still lacks comparable daily protocol revenue, and the archive fit is pending.
Model design and release status
Release status (original research release, November 2026): the preregistered readiness gate, the fixed deterministic estimator and the diagnostics battery are implemented and reviewed, and the estimator is verified on synthetic known-behavior fixtures only. The real-archive fit run is pending operator execution; its reproducible recipe is published on the methodology page. No fitted model, coefficient, standard error, confidence interval or causal conclusion from the real archive is published anywhere, and the public model catalog stays empty until a reviewed run is published. Until then this section reports an unmet release gate rather than results.
The preregistered readiness gate has decided what the saved archive may support: the overall verdict for the 90-day window is PROXY_ONLY_READY_WITH_LIMITS, with gross_fee_activity_v1 the only admitted specification — always with limits, never labeled protocol revenue. The preferred true-revenue specification is NOT_READY, and lower-bound supply series are REJECTED as regression outcomes.
The admitted specification is a Δlog fee-on-volume association estimated by ordinary least squares with a preregistered Newey–West HAC covariance (Bartlett kernel, fixed lag count from the config — never data-selected). Its limitations are attached by the specification itself, not by this page: a largely mechanical fee–volume construction, a single provider lineage, a short sample with no long-run claim, no causal identification, and the prohibited labels protocol_revenue, token_valuation, causal_effect, demand_elasticity and sustainable_revenue.
Statistical nonsignificance is a valid finding. When an operator run is verified and published through the release gate, its exact recorded values render verbatim on the model pages with the limits and prohibited labels displayed before any coefficient; nothing here is rounded, refitted or re-selected.
Alternative explanations
- Price changes can reduce token quantity despite stable USD flows — the recorded decomposition in CL-004 shows this mechanically, without implying causation in either direction.
- Fee-rate changes, rebates, recipient allocations, batching and accrual lags, treasury stock movements and source revisions can alter observed relationships.
- Simultaneous volume and fees may be mechanically linked; the admitted readiness proxy is exactly that construction.
- A growing collector's activity can aggregate many sources; concentration of collectors is not concentration of customers.
- General market conditions may jointly influence adoption and revenue.
Risks and counterarguments
- Governance can change deployment and retention of acquired NEAR at any time (CL-009).
- Fee competition can lower retained margins; costs and incentives are unknown.
- Declared appFees may omit other business channels and funding sources (CL-005).
- Source definitions and historical revisions disagree; provider-reported fee history has been revised after publication.
- Treasury accumulation is reversible, and positive supply growth persists (CL-007, CL-008).
- Existing evidence does not identify token holder contractual claims or a justified valuation multiple.
Long-term scenarios
Qualitative assumptions only — not probability-weighted forecasts, price projections or recommendations. Advanced simulation is outside the MVP scope.
Downside
Adoption grows but retained revenue or fee rates fall and acquired assets are spent; no durable capture.
Conditional base
Recurring receipts and conversions persist with transparent retention, but remain modest relative to issuance and costs.
Upside
Independently reconciled revenue scales, costs are covered and governance implements verifiable durable capture.
Final assessment
The conversion channel is observable and recurrent in the selected cohort. Its NEAR quantity did not grow over the measured windows (CL-002).
Some fee-to-conversion attribution is established on narrow dates (CL-005, CL-006), but neither full 90-day fee funding nor sustained protocol-revenue growth is proven.
Positive supply growth and discretionary treasury use constrain stronger value-capture claims (CL-007, CL-008, CL-009).
The investment thesis remains conditional on stronger revenue attribution, scale, costs and retention evidence. No price target or recommendation follows from these observations.
Limitations
- Cohort selection excludes converters without payouts to the selected treasury wallets.
- Index completeness is not independently universal.
- Amount-and-time payout matching differs from a full causal receipt graph (CL-003).
- USD marks approximate execution rather than recording trade prices (CL-004).
- Daily protocol revenue is missing; gross issuance and burn exist only as incomplete lower bounds (CL-008).
- The provenance period is much shorter than the 90-day cohort (CL-005, CL-006).
- The treasury index cannot establish all possible disposal or lock paths (CL-009).
- No real-archive econometric fit exists yet — the estimator is implemented and synthetic-verified, and the real-archive fit run is pending operator execution (see the econometric evidence section).
All recorded evidence is historical: the cohort was verified on 4 November 2026 and the provider register on 5 November 2026, and the archival mint-header audit stopped incomplete (CL-008). Values on this page are never refreshed in place — a provider revision becomes a new dataset version that supersedes the old one, and a fresh retrieval does not make an old observation fresh.
Sources and reproducibility
Primary recorded findings and artifacts are cited per claim in the registry above, with the repository documents named as the source of every value. The metric catalog, source register and reproduction instructions define exact scope.
The official DefiLlama adapter describes its own accounting; it is not proof of independent attribution.
The original research release (November 2026) is documented in the repository at research/reports/ — the release report and its content/evidence manifest, referenced as checkout paths, never as downloads. The manifest pins the SHA-256 of every published release document and the versioned metric definitions behind every published claim.
Future published claims must bind CL ids to dataset and definition versions, code, lock and config hashes, and an approved public evidence manifest.
The recorded analysis artifacts (manifests and summaries) live in the operator's local research archive and are shared only through verified, hash-checked handoffs; a fresh clone of this repository intentionally does not contain them. Approved application dataset versions are delivered read-only through the delivery APIs — the landing metric cards render the approved dataset key, version, status and coverage — but there is no public dataset catalog, so every quantity on this page carries its document-level source and registered definition version instead of implying a live link.
Recorded evidence basis: the saved 90-day cohort 2026-08-06 through 2026-11-04 (UTC, end exclusive) verified on 4 November 2026, plus the provider register reviewed 5 November 2026. No live values are published on this page; archival collection stays stopped, and a missing or unavailable value is never shown as zero.