[RESEARCH] Economic Value Distribution in Blockchain Ecosystems - June 2026

This research is an empirical study of annual fee revenue, infrastructure costs, and stakeholder compensation in layer-1 and layer-2 protocols based on comprehensive case studies of 25+ chains and L2 solutions, 20 top protocols, and 14 oracles between December 2024 and June 2026.

Crypto Market Cap as of June 20, 2026: $2.28T[^s1_1] — down from the $3.6–4.3T cited at the October 2025 baseline. The thesis below did not move with the price. The dollars shrank; the dependency did not. 🔷 HARD DATA

---

Table of Contents

1. Money Flow Categories
2. Major L1 Networks: Money Allocation Analysis
3. Layer 2 Networks: Fee Distribution
4. Infrastructure Layer: The Hidden Recipients
5. Ecosystem Funding: Foundation and VC Money Flows
6. Comprehensive Money Flow Summary

---

Executive Summary

This analysis examines the distribution of economic value generated when users interact with blockchain networks. For every $1 in transaction fees, value fragments are distributed across multiple recipients. It can be on-chain, among validators, miners, foundations, token holders, oracle networks and MEV searchers. Or it can be off-chain among venture investors, infrastructure operators, oracle networks and other service providers.

Throughout this report we separate organic user fees (what users actually pay for blockspace and services) from three categories of non-fee-funded value flows: consensus/security issuance (a designed security budget, not a temporary handout), external venture capital, and insider supply transfer via token unlocks. Keeping these four buckets distinct is the whole point — collapsing them into one word ("subsidy") is exactly the imprecision this report exists to correct.

A note on the denominator, stated up front because the headline depends on it. The $12.8B figure we anchor to is retained protocol revenue — what protocols, validators and miners keep after paying LPs and suppliers. Gross fees across DeFi are higher, roughly $20.3B/yr; the gap is LP payments, supplier costs and gas-rebates that flow through the system but are not retained by any tokenholder[^s1_4]. We report the ratio against both windows so the reader can see exactly which base produces which number.

Given the extensive amount of data, a GitHub repository has been created as the core reference repository[^s1_2]. This report examines the core dynamics of economic value distribution across the blockchain ecosystem. However, for further, comprehensive details, methodology, and supporting case studies under other points of view or assets, refer to the full repository information.

A key finding of this report is that oracle networks monetize primarily through non-public commercial contracts rather than transparent on-chain fee mechanisms while constituting critical Web3 infrastructure, as referenced in the oracles infrastructure analysis[^s1_3]. This is true of the dominant subscription model (Chainlink); pull-based oracles such as Pyth do charge on-chain update fees and run staking/usage models that are partly visible on-chain, so the opacity claim is a tendency, not a universal[^s1_3].

To create this report, we analyzed end-to-end cash flows across 25+ major networks, 20 leading protocols, dApps per TVL and fees generated, as well as meta-research, using quantitative data and/or expert assessment for cases where public information is limited or unavailable. This report extends the analysis not only to on-chain token unlocks and issuance mechanisms, but also to off-chain financial flows across 14 oracle providers, major infrastructure entities, venture capital and foundation ecosystems, which together represent the largest hidden layer of the blockchain economy. These flows, though largely opaque, ultimately shape what end users pay and receive. This underscores the extent to which blockchain remains an externally supported financial experiment rather than a fully self-sustaining system.

Key Findings

Strip away the narrative and one number refuses to behave. As of June 20, 2026, the blockchain sector retains roughly $12.8B per year in transparent, on-chain revenue[^s1_4] — yet an estimated $50–60B per year of non-fee-funded value flows alongside it: issuance, venture capital, and insider unlocks. The machine looks self-sustaining. It isn't.

How large is the gap? It depends entirely on the denominator, and we refuse to hide that:

• Against gross fees ($20.3B/yr) — the broadest "organic demand" base — non-fee-funded flows are roughly 72% of total value flows[^s1_4][^s1_5].
• Against retained revenue ($12.8B/yr) — the cash protocols actually keep — the share rises to roughly 80–81%[^s1_4][^s1_5].
• A real slice of those "fees" is itself incentive-driven and circular — emissions-farmed DEX volume and perp wash-trading — so the truly exogenous demand sits below the $20.3B gross-fee line. Haircutting gross fees for that circularity (an estimated 25–40%) pulls the gross-fee ratio up from 72% toward the high-70s, closing most of the gap with the retained-revenue number[^s1_4].

Triangulating those, we report a defended range of roughly 75–82%, central near 80%, of measured blockchain value flows originating from sources other than organic user demand[^s1_4][^s1_5]. The October 2025 report put that ratio at 85–90% against an $86–113B base; eight months of price compression — BTC at $63,932, ETH at $1,731, SOL at $71 — deflated the absolute dollars while the structural ratio merely eased into the high-70s to low-80s. The dependency is intact. Only the denomination changed.

Measurement-basis caveat (read before citing the ratio). These flows are not all the same kind of dollar. Fees, revenue and VC are realized cash changing hands. Issuance and token unlocks are marked-to-market notional value of newly-available supply — no cash necessarily moves, and the "value" is endogenous to the same token price that deflates fees. The ratio therefore compares total economic value-at-stake, not like-for-like cash flows. Roughly 60% of the non-fee numerator (issuance + unlocks) is notional; the denominator (fees) is cash. We keep the sum because it is the thesis, but the reader should hold the basis-mix in view[^s1_7][^s1_12].

Core on-chain revenues (transparent, API-verifiable):

• Blockchain base-layer fees (BTC + ETH L1 + SOL): approximately $355M/yr — BTC $79.8M, ETH L1 $135.6M, SOL $139.8M[^s1_6] 🔷 HARD DATA
• Gross protocol fees across DeFi, L2s, DEXs, staking services: approximately $20.3B/yr (DeFiLlama 30-day fees of $1.670B, annualized × 365/30)[^s1_4] 🔷 HARD DATA
• Retained protocol revenue (after LP/supplier payouts): approximately $12.8B/yr (DeFiLlama 30-day revenue of $1.050B, annualized × 365/30)[^s1_4] 🔷 HARD DATA
• The $12.8B retained figure is the denominator we anchor the headline to (down from the approximately $13.7B October 2025 figure — note the new number is lower, not higher, despite a maturing sector).

Non-fee-funded value flows (issuance is hard data; aggregates are estimates):

• Bitcoin mining issuance: approximately $10.5B/yr — 3.125 BTC × 144 blocks × 365 days = 164,250 BTC × $63,932 = the single largest line item in the entire industry, by a wide margin[^s1_7] 🔷 HARD DATA (issuance schedule + live price)
• Ethereum gross staking issuance: approximately $1.9B/yr (approximately 1.1M ETH/yr gross consensus issuance run-rate × $1,731)[^s1_8] — this is gross security-budget issuance; net ETH dilution after the EIP-1559 burn is far lower and burn-variable (see Sustainability Gap). 🔷 HARD DATA on cumulative on-chain staking; the 1.1M ETH/yr forward run-rate is a derived ESTIMATE.
• Solana staking inflation: approximately $1.6B/yr (3.795% disinflationary rate × approximately 580M circulating SOL × $71.48)[^s1_9] — ESTIMATE derived from the inflation schedule and live price.
• Core-3 issuance (BTC + ETH + SOL): approximately $14.0B/yr, of which Bitcoin alone is roughly 75% — the industry's security budget is now overwhelmingly a Bitcoin number[^s1_7][^s1_8][^s1_9].
• Venture-capital deployment: approximately $16B/yr cyclical run-rate — see range note below[^s1_10][^s1_11]. CYCLICAL RUN-RATE, not a forecast.
• Insider supply / value transfer via token unlocks: approximately $18–24B/yr net (wide-error-band ESTIMATE; no verified public aggregate exists)[^s1_12].
• Total non-fee-funded base: approximately $50–60B/yr (ESTIMATE — the spread is driven mainly by the VC cyclical range and the unlock error bars).

A note on the spread, because it is the whole game. BTC, ETH and SOL issuance together total a clean approximately $14.0B/yr of hard-data inflation (plus an estimated $1–2B from smaller L1s such as Tron, Cardano, Avalanche and Cosmos)[^s1_7][^s1_8][^s1_9]. The two soft inputs — VC and token unlocks — are where the uncertainty lives. The institutional-standard Galaxy Digital tally puts 2026 crypto VC on an approximately $16B annualized run-rate (Q1 2026 was $4.0B across roughly 355 deals, down approximately 50% QoQ, with median deal size at an all-time high above $4.5M); full-year 2025 ran near $20B[^s1_10]. Broader trackers that bundle M&A and later-stage rounds reach approximately $27B[^s1_11]. Insider unlocks add an estimated $18–24B/yr of value transferred from new market buyers to vesting insiders — marked at market price, which dwarfs the VC cost basis embedded inside it, so unlocks are not a re-count of VC dollars (see footnote [^s1_12] for the netting). Pair Galaxy's $16B with issuance and net unlocks and the non-fee base lands near $50–55B; use the higher VC tracker and it stretches toward $60B+. We lead with the Galaxy-anchored central case.

The following numbers would make Milton Friedman faint: an industry that mints, vests, and venture-funds its way to the appearance of viability while organic income covers barely a fifth of the bill. Written per dollar of retained user revenue, $1.00 kept by protocols, validators and miners runs alongside roughly $4 of non-fee-funded flows under the $50–55B central case (approximately $5 at the upper $60B+ scenario) — issuance, vesting, and venture capital, none of which a user ever sees on a fee receipt[^s1_5]. BNB's approximately $4B/yr quarterly auto-burn is deliberately excluded from this total: it removes supply rather than funding activity, and counting it would flatter the deflationary case it actually represents[^s1_13].

Notable Exceptions

A limited subset of ecosystems is close to genuinely self-sustaining models, and they cut against the "5–15% organic" pattern that holds for the issuance-funded L1 sample:

• Hyperliquid: Generates genuine, protocol-retained trading-fee revenue rather than relying on issuance — the rare chain whose income line is real[^s1_14].
• Base: Captures sequencer revenue and operates profitably, now among the largest L2s by DeFi TVL (approximately $4.22B)[^s1_15] 🔷 HARD DATA.
• Optimism: Operates a Superchain architecture that captures portions of the fees of other OP Stack L2s, but is not yet at breakeven[^s1_16].
• Stablecoin settlement is a fourth fee-real exception in spirit: chains carrying heavy stablecoin transfer volume (Tron, and increasingly Ethereum and Solana) earn fees tied to genuine payment demand rather than emissions, though much of that value accrues to the issuer rather than the chain[^s1_6].

Yet even these exceptions face material long-term risks: large scheduled team and investor token unlocks continue to test whether fee-funded models can survive their own emission schedules.

The Sustainability Gap

• Bitcoin funds its issuance-based security budget at roughly 130–135:1 versus user fees — approximately $10.5B/yr of block-subsidy issuance against roughly $79.8M/yr in actual user fees, meaning fees cover less than 1% of what Bitcoin spends to stay secure[^s1_6][^s1_7]. This is a designed security budget, not a temporary handout — but as the price compressed, the fee base thinned faster than the subsidy, leaving Bitcoin's long-run fee-only security question more exposed, not less.
• Ethereum has tilted back to net inflationary (approximately +0.9%) as L2s — Base foremost — absorbed mainnet activity. The directional claim is solid; the magnitude is not fixed, because the EIP-1559 burn swings with blob and blockspace demand. At the current record-low burn pace (approximately 16,800 ETH/yr, roughly $29M/yr) against approximately $1.9B/yr of gross staking issuance, net inflation is modestly positive — but a demand spike could compress or briefly reverse it[^s1_8][^s1_17].
• Solana depends on approximately $1.6B/yr of inflation against approximately $140M/yr in network fees — issuance still dwarfs organic fees by roughly an order of magnitude[^s1_6][^s1_9].

Across the issuance-funded L1 sample analyzed here, user fees represent at best 5–15% of total value flows even for major established networks — with the fee-real exceptions above (Hyperliquid, Base, stablecoin-settlement volume) as the deliberate counterexamples.

The Hidden Economy: MEV and Extraction

Maximal Extractable Value (MEV) is part of the non-fee economy, but it is not monolithic and should not be painted as pure theft:

• Extractive MEV (sandwich attacks, frontrunning) is a genuine tax on users and the clearest "hidden cost" line.
• Efficiency MEV (arbitrage, liquidations) performs a real economic function — it keeps DEX prices aligned with markets and clears bad debt — even as searchers profit from it.
• Protocol-captured MEV (e.g. via order-flow auctions, MEV-Boost redistribution, application-level capture) returns part of the value to validators, protocols or users rather than leaking it entirely.

We keep MEV inside the hidden economy because end users ultimately bear its extractive share, but the framing is "a mix of tax and infrastructure," not "pure extraction"[^s1_18].

Market Opacity Warning

Caution remains warranted, as blockchain markets exhibit persistent structural opacity:

• Exchanges charge $1–5 million listing fees while engaging in wash and proprietary trading.
• Market makers demand 10–15% token loans with options allocations.
• Venture funds frequently coordinate distribution cycles via private communication channels.

The opacity is not academic — it surfaces violently. Between June 4 and June 6, 2026, a leveraged liquidation cascade wiped out over $3 billion in positions as Bitcoin fell from approximately $67,000 to a cycle low near $59,100; longs accounted for roughly 85% of BTC-specific losses, and open interest collapsed 22% in a single day[^s1_19]. Cascades like this are the recurring tell of a market still propped up by leverage and subsidy rather than organic demand. The "Digital Asset Treasuries (DAT)" narrative — corporate vehicles that raised an estimated approximately $29B through 2025 to hold tokens on balance sheets[^s1_20] — appears primarily designed to repackage illiquid tokens for secondary distribution, coinciding with a sustained collapse in retail participation since 2021.

---

[^s1_1]: CoinGecko — Global Charts — Total crypto market cap $2.28T, BTC dominance 56.2%, retrieved via CoinGecko Global API (June 20, 2026). 🔷 HARD DATA

[^s1_2]: Blockchain Payment Flow Analysis — GitHub Repository — Core reference repository for all chain, protocol, and infrastructure case studies (June 2026).

[^s1_3]: Oracles Infrastructure Analysis — Oracle monetization. The opacity claim is specific to Chainlink's dominant subscription/commercial-contract model; pull-based oracles (Pyth) charge on-chain update fees and run staking/usage models that are partly on-chain visible (June 2026).

[^s1_4]: DefiLlama — Fees & Revenue Overview — Total DeFi protocol fees 30-day: $1.670B (annualized × 365/30 = approximately $20.3B/yr; trailing-1y $24.9B); total DeFi protocol revenue 30-day: $1.050B (annualized × 365/30 = approximately $12.8B/yr; trailing-1y $14.08B), retrieved via DeFiLlama fees overview API (June 20, 2026). Both figures are cited. The headline ratio anchors on retained revenue ($12.8B), with the gross-fee figure ($20.3B) reported alongside as the broader denominator; the 30-day run-rate is used (not trailing-1y) so the numerator and denominator are both marked at today's depressed prices. 🔷 HARD DATA

[^s1_5]: DefiLlama — Fees & Revenue Overview — Per-dollar multiple derived from the $50–55B central non-fee base ÷ $12.8B retained revenue ≈ $4.0–4.3 of non-fee flows per $1 of retained revenue (rising to approximately $5 at the $60B+ upper scenario). ESTIMATE; the VC and token-unlock inputs are non-hard and notional. (June 20, 2026)

[^s1_6]: DefiLlama — Chain Fees — Base-layer fees annualized from 30-day actuals × 365/30: BTC L1 $79.8M, ETH L1 $135.6M, SOL $139.8M (total approximately $355M/yr), retrieved via DeFiLlama fees API (June 20, 2026). 🔷 HARD DATA

[^s1_7]: Bitcoin Halving Schedule — Bitcoin Foundation — Post-April 2024 block reward 3.125 BTC; 3.125 × 144 blocks/day × 365 = 164,250 BTC/yr × $63,932 ≈ $10.5B annual gross issuance. Next halving April 2028. Issuance value is marked-to-market notional supply, not a cash flow. 🔷 HARD DATA (issuance schedule) combined with CoinGecko — Bitcoin live price (June 20, 2026).

[^s1_8]: Etherscan — ETH Supply Statistics — Cumulative Eth2 staking rewards 2,940,327 ETH since the Merge; cumulative EIP-1559 burnt 4,630,257 ETH, retrieved via Etherscan ethsupply2 API (June 20, 2026). 🔷 HARD DATA (on-chain cumulative). The forward gross issuance run-rate of approximately 1.1M ETH/yr (× $1,731 ≈ $1.9B/yr) is a DERIVED ESTIMATE — the cumulative average since the Merge is lower (approximately 784k ETH/yr), but the current run-rate is higher as ETH staked has grown toward approximately 39M; plausible range 1.1–1.4M ETH/yr. This is gross security-budget issuance, not net dilution.

[^s1_9]: Solana Compass — Tokenomics & Inflation Schedule — Disinflationary schedule (8% initial, −15%/yr, 1.5% floor); current approximately 3.795% × approximately 580M circulating SOL ≈ 22M SOL/yr × $71.48 ≈ $1.6B/yr (June 2026). ESTIMATE — derived from the inflation schedule and live circulating supply; notional, not a cash flow.

[^s1_10]: CryptoPotato — Crypto VC Funding Falls 50% After Q4 2025 Surge (Galaxy) — Galaxy Digital Q1 2026: $4.0B across approximately 355 deals, approximately 50% QoQ decline, median deal size at an all-time high above $4.5M → approximately $16B annualized run-rate; FY2025 approximately $20B. Used as the central VC input. ESTIMATE, labeled CYCLICAL RUN-RATE (a Q1 × 4 annualization), explicitly not a forecast (June 2026).

[^s1_11]: Q1 2026 Crypto Fundraising Report — crypto-fundraising.info — Q1 2026: $6.81B across 222 rounds → approximately $27B annualized. Broader tracker that bundles M&A and later-stage rounds; used as the high end of the VC range. ESTIMATE (June 2026).

[^s1_12]: CoinGecko — Incoming Token Unlocks — No verified public annual aggregate exists; CoinGecko, Messari, Tokenomist and CryptoRank track individual events only. The 2026 monthly unlock value averages approximately $2B (→ approximately $24B/yr gross), but that average is inflated by the March 2026 cliff (approximately $6B, of which 69% was a single token — WhiteBIT WBT at $4.18B). Excluding that outlier, gross falls to approximately $19.8B/yr. Net of an estimated VC cost-basis overlap — a coarse, unsourced ESTIMATE with its own wide band ($2.5–6B), since no public decomposition of unlock recipients (VC vs team vs foundation vs ecosystem) exists — the net central lands at approximately $18–24B/yr, ex-WBT closer to approximately $19B. Unlock value is marked at market price and exceeds the VC cost basis embedded within it, so it is not a re-count of VC dollars. ESTIMATE — wide error bars, notional, not hard data and not a cash flow.

[^s1_13]: Chainwire — BNB Chain 35th Quarterly Token Burn — H1 2026 burns annualize to approximately 7M BNB (approximately $4B/yr at the current price). Deflationary supply removal, correctly EXCLUDED from the non-fee base (April 15, 2026).

[^s1_14]: Hyperliquid Payment Flow Analysis — Protocol-retained trading-fee revenue model (June 2026).

[^s1_15]: DefiLlama — Base — Base DeFi TVL approximately $4.22B, retrieved via DeFiLlama API (June 20, 2026). 🔷 HARD DATA

[^s1_16]: Optimism Payment Flow Analysis — Superchain fee-capture architecture, not yet at breakeven (June 2026).

[^s1_17]: ultrasound.money — ETH Issuance & Burn — EIP-1559 burn collapsed to approximately 324 ETH per 7-day period (approximately 16,800 ETH/yr ≈ $29M at $1,731) as L2s absorbed L1 activity; net ETH inflation approximately +0.9%. The burn is demand-variable — a blob/blockspace demand spike can compress or briefly reverse net inflation, so the magnitude is not fixed (mid-June 2026).

[^s1_18]: Flashbots Documentation — MEV Overview — MEV disaggregated into extractive (sandwich/frontrun), efficiency (arbitrage/liquidation) and protocol-captured (order-flow auctions, MEV-Boost redistribution) categories. Annual MEV gross is an ESTIMATE, not on-chain-aggregated; only the extractive share is a pure user tax (June 2026).

[^s1_19]: WazirX Blog — Bitcoin June 2026 Liquidation Cascade — Over $3B liquidated June 4–6, 2026; BTC fell approximately $67,000 → approximately $59,100; longs approximately 85% of BTC losses; open interest down 22% on June 4 (June 2026).

[^s1_20]: The Block — Crypto VCs Share 2026 Funding and Token Sales Outlook — Digital Asset Treasury (DAT) companies raised an estimated approximately $29B through 2025 (per Galaxy research). ESTIMATE — a separate capital channel from VC (2026).

---

Money Flow Categories

Money does not enter a blockchain at the top and trickle down. It enters in several streams at once — one you can see on your receipt, and several that move in the dark. Across the issuance-funded Layer-1 sample that anchors this report, the same conclusion keeps surfacing: the fee a user pays is the smallest number in the room. As of 20 June 2026, transparent on-chain protocol revenue runs at roughly $12.8B/year retained — out of $20.3B/year in gross fees[^s2_cat1] — while the non-fee-funded value flows beneath it run at an estimated $48–60B/year (central approximately $53B)[^s2_cat2]. For every roughly $1 of organic run-rate fees a user pays, on the order of $2.6 of value originating from sources other than organic user demand is already in motion. Here is where each stream goes — sorted into four buckets the rest of the report uses: organic user fees, issuance-funded security budget, external capital (VC), and insider supply transfer (unlocks).

Bucket 1 — Organic User Fees (and where they land)

When users pay transaction fees, the money immediately flows to:

1. Validators/Miners: Network security providers receiving fee revenue. The headline number is brutal — base-layer chain fees across the three largest networks total just approximately $352M/year (BTC approximately $79.8M, ETH L1 approximately $133.8M, SOL approximately $138.0M, annualized from 30-day actuals)[^s2_cat3]. That is what users actually hand to the people securing the chains.
2. Token Burn Mechanisms: Reducing supply to benefit all token holders. On Ethereum, EIP-1559 (now layered with Fusaka-era blob pricing) burns the base fee — but L2s have hoovered up mainnet activity, collapsing the burn to roughly 16,800 ETH/year (approximately $29.1M), a record-low pace that has left the network directionally net-inflationary in mid-2026[^s2_cat4]. Treat this as a snapshot, not a fixed rate: the base-fee burn swings with blob and blob-fee demand, so the net-issuance sign can move with congestion.
3. Protocol Treasuries / Retained Revenue: DAOs and foundations receiving fee shares. Of the approximately $20.3B/year in gross DeFi protocol fees, only approximately $12.8B/year is retained as revenue — the rest is paid through to suppliers, LPs, and stakers[^s2_cat1]. A real share of even that gross-fee figure is incentive-driven and circular — emissions-farmed DEX volume and perpetuals wash-trading inflate "fees" that no exogenous user would pay absent the token reward — so truly-exogenous user demand sits below the $20.3B gross. That haircut is precisely why the headline ratio sits toward the high end of its range (see Bucket 4).
4. Layer 1 Settlement: L2s paying for Ethereum security. The cruel irony of 2026 — the very L2 success that drained ETH's L1 fees means rollups now settle to Ethereum for a fraction of what mainnet once charged.
5. MEV Extractors: Searchers and validators capturing MEV value. On Ethereum, Flashbots' MEV-Boost alone paid validators approximately $241.4M over the trailing year (cumulative approximately $1.66B since launch)[^s2_cat5]; on Solana, Jito tips delivered approximately $164.8M to validators in the same window[^s2_cat6]. (MEV is disaggregated below — not all of it is pure extraction.)

A scoping note on the "smallest number in the room" framing: it is built on the issuance-funded L1 sample (Bitcoin, Ethereum, Solana and similar consensus-subsidized chains). It is not universal. Fee-real exceptions exist — Hyperliquid and Base capture organic fees that are a far larger share of their economics, with little or no issuance subsidy underneath. And a large share of crypto's transfer value — stablecoin settlement — moves enormous notional with minimal protocol fee capture, so "fees are tiny" cuts both ways: tiny relative to value moved, but for the right reason on those rails.

Bucket 2 — Issuance-Funded Security Budget

This is the largest non-fee value flow, and it is a designed security budget, not a temporary external subsidy. New token issuance dilutes existing holders to pay for consensus security:

1. Bitcoin mints 164,250 BTC/year (approximately $10.5B at $63,953) in block subsidy — against roughly $80M in actual user fees[^s2_cat7]. Reframed honestly: fees cover well under 1% of Bitcoin's security spend; the rest is the issuance-funded security budget. Bitcoin is approximately 76% of core-3 issuance — the network security budget is now overwhelmingly a Bitcoin number.
2. Ethereum issues an estimated approximately 1.0–1.1M ETH/year (approximately $1.8B) in gross consensus issuance to stakers[^s2_cat8]. This is a gross security-budget figure, not a net-dilution figure: net ETH issuance after the EIP-1559 burn is far lower and burn-variable (see Bucket 1, item 2).
3. Solana inflates at 3.795% (approximately 22M SOL ≈ $1.57B/year at $71.51) on roughly 580M circulating SOL, declining 15%/year toward a 1.5% floor[^s2_cat9].

Core-3 issuance totals approximately $13.9B/year; adding other issuance-funded L1s (Tron, Cardano, Avalanche, Polkadot, NEAR, Cosmos, Aptos) brings the central estimate to approximately $15.3B/year (band $14.3–15.8B)[^s2_cat2]. Issuance alone exceeds all transparent on-chain retained revenue combined. Foundation grants sit alongside this bucket: Ethereum, Solana, and dozens of L1/L2 foundation treasuries quietly underwrite the developers, audits, and events that fees never cover.

Bucket 3 — External Capital (VC)

Private capital formation flowing to protocols and infrastructure. This is a cyclical run-rate, not a forecast: Galaxy Research counted approximately $4.0B deployed across approximately 355 deals in Q1 2026 (−50% QoQ, −16% on deal count), which annualizes to approximately $16B; full-year 2025 ran near $20B[^s2_cat11]. Broader trackers run hotter — one logged $6.81B across 222 rounds in Q1 2026 (annualizing toward $27B)[^s2_cat10] — but those bundle M&A-style rounds. The defensible run-rate band is approximately $16–20B/year, anchored on Galaxy's institutional-standard count; even the floor dwarfs the approximately $352M users pay in base-layer fees. (Notably, median deal size hit an all-time high above $4.5M in Q1 2026 — capital concentrated into fewer, larger checks.) Airdrops are a related user-acquisition cost, paid in inflation rather than cash — billions in token value distributed annually to bootstrap activity that fees cannot fund.

Bucket 4 — Insider Supply / Value Transfer (Token Unlocks)

Token unlocks are value transferred from new market buyers to insiders (VCs, teams, foundations) via vesting schedules — minted supply hitting the market at market price. 2026 monthly unlock value has averaged approximately $2B (Tokenomist/CryptoRank), implying roughly $24B/year gross[^s2_cat16]. Two adjustments matter:

• Outlier contamination: that monthly average is inflated by the March 2026 cliff, which spiked to approximately $6B — of which 69% was a single token (WhiteBIT WBT, approximately $4.18B). Stripping the WBT spike pulls the run-rate down materially; ex-WBT the central is closer to approximately $19–20B/year gross. We therefore center the net figure at approximately $21.5B/year but flag that ex-outlier it is nearer $19B — the band is wide ($18–24B) and no verified public aggregate exists, so this is an ESTIMATE, not hard data.
• VC overlap (coarse): unlocks are not a re-count of VC dollars. Unlock value is marked at market and dwarfs the VC cost basis embedded in it. We net only a small VC cost-basis overlap, but with no public decomposition of unlock recipients (VC vs team vs foundation vs ecosystem) this netting is itself a coarse estimate (approximately $2.5–6B), widening the band rather than narrowing it. We make no precise overlap adjustment.

Kept in the thesis as a non-organic value flow per the report taxonomy.

Hidden Value Extraction (cross-cuts the buckets)

Additional parties extract value without a direct, line-item user payment:

1. MEV Infrastructure — disaggregated, not pure theft. Gross MEV on Ethereum is estimated at $480–720M/year and on Solana at $207–237M/year — a combined approximately $690M–960M/year[^s2_cat14]. But it splits three ways: extractive MEV (sandwich attacks) is a genuine tax on ordinary users; efficiency MEV (arbitrage, liquidations) keeps prices aligned and bad debt cleared, value the system needs someone to capture; and protocol-captured MEV (e.g. via MEV-Boost / Jito tips, items in Bucket 1) is recycled back to validators rather than skimmed. Only the extractive slice is the "tax on the user who thought $1 was the whole bill."
2. Exchange/Market Maker Partnerships: Off-chain deals — listing fees, market-making rebates, revenue splits — move value that never touches a public ledger.
3. Institutional Services: A growing slice of staking issuance is captured by custody and staking-as-a-service operators who take a cut before rewards reach delegators.
4. Infrastructure Services (oracles, RPC, indexing): Chainlink's on-chain oracle fees run approximately $73.5M/year (annualized from $6.04M/30d)[^s2_cat12] — but this is specific to Chainlink's dominant request-and-pay model; other oracle networks monetize differently (Pyth uses on-chain pull fees, staking/Oracle Integrity Staking, and usage-based models), so "oracles are subsidized" is not universal[^s2_cat12]. The RPC and node-infrastructure market — led by Alchemy's estimated approximately $447M ARR — totals an estimated $600–900M/year (estimate, not hard data)[^s2_cat13]. Indexing: The Graph now collects barely approximately $99K/quarter in query fees while paying out approximately $7.6M/quarter in indexing rewards funded by inflation[^s2_cat15] — the gap is the clearest single example of infrastructure running on issuance, not user payment.

The pattern across the buckets is the same. Organic fees are real but small (approximately $352M base-layer, approximately $12.8B total retained protocol revenue out of $20.3B gross). The issuance-funded security budget (approximately $15.3B), external capital (approximately $16–20B), and insider unlocks (approximately $18–24B) together run roughly $48–60B/year, an order of magnitude larger. And hidden extraction (gross MEV approximately $0.7–1B, only part of it extractive) quietly cross-cuts all of it.

Deriving the headline honestly. A measurement-basis caveat first: fees, retained revenue and VC are realized cash; issuance and unlocks are marked-to-market notional supply (no cash necessarily changes hands, and that "value" is endogenous to the same token price that deflates fees). The ratio below compares total economic value-at-stake, not like-for-like cash — roughly 60% of the non-organic numerator is notional. With that flagged, the non-fee-funded share lands in a defended range, not a single figure:

• vs gross fees ($20.3B run-rate): non-organic / (non-organic + fees) ≈ 72% — the floor, before any fee-circularity haircut.
• Apply the fee-circularity haircut (a real share of gross fees is emissions-farmed / wash-traded), shrinking the truly-organic denominator below $20.3B → pushes to approximately 77–80%.
• vs retained revenue ($12.8B run-rate): ≈ 80–81% — the ceiling.

Triangulating across those, the defended headline is approximately 75–82%, central approximately 80% — and we do not print a bare "80%" without naming retained revenue (and the haircut) as the base that gets it there[^s2_cat2]. The receipt shows $1. The machine moves several.

---

[^s2_cat1]: 🔷 HARD DATA — DefiLlama — Fees & Revenue Overview: total DeFi protocol fees 30-day $1.670B (30-day-annualized approximately $20.3B/yr; trailing-12m $24.9B, inflated by the late-2025 price peak); protocol revenue retained 30-day $1.0495B (30-day-annualized approximately $12.8B/yr; trailing-12m $14.08B). Headline anchored on the 30-day run-rate because the issuance numerator is also marked at today's depressed BTC/ETH/SOL prices. Retrieved via DefiLlama fees/dailyRevenue overview API, June 20, 2026.

[^s2_cat2]: DefiLlama — Fees & Revenue — Non-fee-funded value flows derived: issuance-funded security budget approximately $15.3B (band $14.3–15.8B) + external capital/VC approximately $16–20B run-rate + insider unlocks (net) approximately $18–24B ≈ $48–60B/yr (central approximately $53B). Share-of-total: vs $20.3B gross run-rate fees ≈ 72% (floor); after a fee-circularity haircut ≈ 77–80%; vs $12.8B retained run-rate revenue ≈ 80–81% (ceiling). Defended range approximately 75–82%, central approximately 80%. Measurement-basis caveat: issuance and unlock values are marked-to-market notional supply, not realized cash; VC and fees are cash — the ratio compares total economic value-at-stake, not like-for-like cash. Estimate, not hard data — VC and unlock inputs are expert ranges, not audited aggregates. Down in absolute dollars from the Oct 2025 thesis ($86–113B base) due to BTC/ETH/SOL price compression; structural share held near 80%.

[^s2_cat3]: 🔷 HARD DATA — DefiLlama — Chain Fees: Bitcoin L1 30-day fees $6.56M (approximately $79.8M/yr); Ethereum L1 30-day $11.14M (approximately $133.8M/yr); Solana 30-day $11.49M (approximately $138.0M/yr); combined approximately $352M/yr. Retrieved via DefiLlama fees API, June 20, 2026.

[^s2_cat4]: ultrasound.money — ETH Issuance & Burn — approximately 16,800 ETH/yr burned (approximately $29.1M at $1,731.92), a record-low pace as L2s absorbed mainnet activity; network directionally net-inflationary in mid-2026. Snapshot, not a fixed rate — base-fee burn swings with blob/blob-fee demand. Burn-pace figure search-confirmed, June 2026.

[^s2_cat5]: 🔷 HARD DATA — DefiLlama — Flashbots: MEV-Boost paid validators approximately $241.4M over the trailing year; approximately $1.665B cumulative since launch. Retrieved via DefiLlama API, June 20, 2026.

[^s2_cat6]: 🔷 HARD DATA — DefiLlama — Jito MEV Tips: Jito tips paid Solana validators approximately $164.8M over the trailing year; approximately $1.417B cumulative since launch. Retrieved via DefiLlama API, June 20, 2026.

[^s2_cat7]: 🔷 HARD DATA (price + fees) / derived (issuance) — Bitcoin block subsidy 3.125 BTC/block × 144 blocks/day × 365 = 164,250 BTC/yr; at BTC $63,953 (CoinGecko — Bitcoin, June 20, 2026) ≈ $10.5B. User fees approximately $79.8M/yr (DefiLlama — Bitcoin Fees). The issuance × price product is a derived ESTIMATE; only the price and fee inputs are 🔷. Halving schedule confirmed post-April 2024.

[^s2_cat8]: 🔷 HARD DATA (on-chain cumulative) — Etherscan — ETH Supply: cumulative Eth2 staking rewards approximately 2.94M ETH since the Merge (🔷, Etherscan ethsupply2 API, June 20, 2026). The annual run-rate of approximately 1.0–1.1M ETH/yr gross (≈ $1.8B at ETH $1,731.92, CoinGecko — Ethereum) is a DERIVED ESTIMATE from cumulative issuance and validator count — not 🔷 — and is GROSS consensus issuance, before the EIP-1559 burn.

[^s2_cat9]: Solana Validator Economics Documentation — disinflationary schedule (8% initial, −15%/yr, 1.5% floor); current rate approximately 3.795% on approximately 580M circulating SOL ≈ 22M SOL/yr ≈ $1.57B at SOL $71.51 (CoinGecko — Solana, June 20, 2026). Issuance figure derived (ESTIMATE) from inflation schedule + supply; only price is 🔷.

[^s2_cat10]: Q1 2026 Crypto Fundraising Report — crypto-fundraising.info — $6.81B across 222 rounds in Q1 2026 (annualizes toward $27B). Tracker bundles M&A-style rounds; figure is an estimate, not audited.

[^s2_cat11]: Galaxy Digital crypto VC data, Q1 2026 (via CryptoPotato) — institutional-standard count: approximately $4.0B across approximately 355 deals in Q1 2026 (−50% QoQ, −16% deal count; approximately $16B annualized; FY2025 approximately $20B); median deal size an all-time high above $4.5M. Cyclical run-rate scenario, explicitly NOT a forecast; methodologies diverge materially from broader trackers.

[^s2_cat12]: 🔷 HARD DATA — DefiLlama — Chainlink: on-chain oracle fees $6.04M (30d), approximately $73.5M/yr annualized (🔷, retrieved via DefiLlama API, June 20, 2026). Scope caveat: this reflects Chainlink's dominant request-and-pay model; other oracle networks monetize differently — Pyth uses on-chain pull fees, Oracle Integrity Staking, and usage-based models — so the "subsidized oracle" framing is not universal.

[^s2_cat13]: Latka — Alchemy company profile — Alchemy approximately $447M ARR (Nov 2025, third-party/unaudited). Total RPC + node-infrastructure market estimated at $600–900M/yr (Alchemy + Infura approximately $60–80M + QuickNode + Ankr + others). ESTIMATE, not hard data.

[^s2_cat14]: ESMA — crypto-asset market analysis and Helius — Solana MEV Report — Ethereum gross MEV approximately $480–720M/yr; Solana approximately $207–237M/yr; combined approximately $690M–960M/yr. Splits into extractive (sandwich), efficiency (arbitrage, liquidation), and protocol-captured (MEV-Boost/Jito tips) slices — only the extractive slice is a direct user tax. Research-consensus ESTIMATE (ESMA, EigenPhi, Helius), not a single audited figure. ⏳ HISTORICAL (ESMA July 2025) — most recent comprehensive regulator gross-MEV baseline; cross-checked against live DefiLlama validator-payment data, June 2026.

[^s2_cat15]: The Graph — Network Data / Dune dashboards — Q4 2025 query fees approximately $98,667 (−8.7% QoQ); indexing rewards approximately 81.6M GRT (approximately $7.6M); GRT $0.0195 (June 20, 2026, CoinGecko — The Graph). Indexing infrastructure funded predominantly by token issuance, not query fees. ⏳ HISTORICAL (Q4 2025) — latest published quarterly network data; GRT price confirmed live June 20, 2026.

[^s2_cat16]: Tokenomist — Token Unlock Schedules and CryptoRank — Token Unlocks — 2026 monthly unlock value averaging approximately $2B (approximately $24B/yr gross), marked at market price. The March 2026 cliff spiked to approximately $6B, of which approximately 69% was a single token (WhiteBIT WBT, approximately $4.18B); ex-WBT the run-rate centers nearer $19–20B/yr gross, net approximately $19B. ESTIMATE with wide error bars (net band $18–24B) — no verified public aggregate exists; not 🔷. June 2026.

---

Major L1 Networks: Money Allocation Analysis

All figures refreshed to 20 June 2026. Prices, fees, revenue, and TVL are live API pulls and carry the 🔷 HARD DATA marker; issuance, MEV, and subsidy totals are calculated or expert estimates and are labeled as such. We follow the report's four-bucket taxonomy throughout: (1) organic user fees — and, separately, the retained revenue inside them; (2) issuance-funded security budgets; (3) external capital; (4) insider supply/value transfer via token unlocks. We do not lump these under one word.

The seven networks below settle the overwhelming majority of public-blockchain value. They also share a fact the marketing decks never put on a slide: for most of them, the transparent, user-funded layer is a fraction of the value moving underneath. Strip away issuance, MEV, corporate burns, and insider unlock schedules, and what users actually pay is a small share of total economic flow on each chain. This section follows a single dollar of user fees through each network and asks one question — when a user pays $1, how much total economic activity actually moves, and how much of it is fee-funded?

A measurement note that governs every multiplier below: gross fees and retained revenue are different numbers, and we say which one anchors each ratio. Gross fees are what users pay for blockspace; retained revenue (DeFiLlama "Revenue") is what validators or the protocol actually keep after burns and LP/supplier payouts. Where a multiplier divides by fees, we use gross fees as the denominator and flag where retained revenue is materially smaller.

---

Ethereum: The Ultrasound Money Thesis, in Abeyance

Ethereum's deflation story is on hold. ETH trades at $1,763.58 as of 20 June 2026[^s3_eth1], down approximately 64% from its $4,946 August 2025 all-time high[^s3_eth1] — among the largest price moves in this refresh. With activity at multi-year lows, the network sits at +0.88% net annual inflation: EIP-1559 burns roughly 40,000 ETH a year at current throughput while staking issues approximately 1.10M ETH[^s3_eth9][^s3_eth10]. (This rate is highly sensitive to blob demand. At today's activity trough, burn runs near 40,000 ETH/yr, implying approximately +0.88% inflation; a return to 2024-level throughput could push burn above 300,000 ETH/yr and compress inflation toward +0.67%. The directional claim — issuance exceeds burn — is robust, but the precise rate swings with demand.) The "ultrasound money" thesis requires mainnet demand to push burn above issuance. It is not close today.

The Fusaka upgrade (5 December 2025) introduced EIP-7918's minimum blob-fee floor[^s3_eth11], but at current L2 volumes the burn impact is marginal. Gas has roughly doubled off its trough — the safe price is now approximately 0.19 Gwei, making a simple transfer cost approximately $0.006[^s3_eth3]. The chain collected $302.7M in user fees over the trailing twelve months[^s3_eth4] 🔷 HARD DATA. Of that gross-fee figure, the priority-fee tip retained by validators is roughly 15% (approximately $45M/yr); the remaining 85% is burned permanently and never reaches a validator's balance sheet. DeFiLlama's "Revenue" line — validator-retained fee income — is $115.3M over the trailing year[^s3_eth4b], which includes MEV-related tips beyond the base priority fee. The $302.7M gross-fee figure is the denominator used for the multiplier below — not retained revenue, which is roughly 2.6x smaller. Against that gross-fee base, the network paid out roughly $1.95B in staking issuance[^s3_eth9] at the live ETH price. DeFi TVL on Ethereum stands at $39.0B[^s3_eth5], with L2s now handling approximately 95% of throughput[^s3_eth16].

When a user pays $1.00 in Ethereum gas fees:

Direct Fee Recipients

• $0.85 — burned via the EIP-1559 base fee (permanent supply reduction; accrues proportionally to all ETH holders)[^s3_eth17]
• $0.15 — validators, as the priority-fee tip[^s3_eth17]

Issuance-Funded Security Budget

• +$6.43 — staking issuance ($1.95B annual gross consensus issuance ÷ $302.7M annual gross fees)[^s3_eth9][^s3_eth4]. This is a designed security budget paid in newly minted ETH, diluting holders who do not stake. Note this is gross issuance; net dilution after the EIP-1559 burn is lower and burn-variable.

Hidden Extraction (disaggregated)

• +$1.82 — MEV flowing to searchers and bots (estimate: approximately $550M/yr ÷ $302.7M fees)[^s3_eth13]. This is not uniform harm: research suggests roughly 40–50% is efficiency MEV (arbitrage and liquidations that incidentally support price discovery and keep lending protocols solvent), while extractive MEV (sandwich attacks, frontrunning) is a direct cost to traders and LPs. A portion of MEV is also protocol-captured — MEV-Boost relay payments that flow back to validators and are partly already counted in the priority-tip line above.[^s3_eth14]

Total Ecosystem Value Flow: approximately $9.24 per $1 of gross user fees — approximately 89% from sources other than organic fees.

The arithmetic deserves a footnote of its own. An earlier draft ran the multiple off a $116M annual-fee base — which is actually DeFiLlama's revenue line, not gross fees — implying a misleadingly high multiplier. The correct gross-fee figure is $302.7M[^s3_eth4], which places Ethereum's true multiple near 9x and its non-fee share in the high-80s. Still overwhelmingly subsidy-shaped; just not a caricature.

---

Bitcoin: An Issuance-Funded Security Budget, Two Orders of Magnitude Over Fees

Bitcoin runs the report's purest issuance-funded security budget. It mints an estimated $10.5B a year in fresh BTC to pay for hashrate, against roughly $58–69M in annual user fees at current throughput[^s3_btc1][^s3_btc2]. That is a security budget where fees cover well under 1% of the spend — a subsidy-to-fee ratio of roughly 150–180:1 at today's depressed fee levels, and it widens, not narrows, as price retreats. BTC trades at $63,951, down approximately 49% from its $126,080 October 2025 ATH[^s3_btc1].

This is a designed mechanism, not a temporary external subsidy: the protocol pays miners in new issuance by construction, and that issuance halves on schedule. The mechanics are simple. Each block pays 3.125 BTC in subsidy and a small fraction of a BTC in fees: 144 blocks/day × 3.125 BTC × $63,951 = approximately $28.8M/day in new issuance, against roughly $159K/day in fees[^s3_btc2]. Fees are approximately 0.55% of miner revenue over the trailing day[^s3_btc2][^s3_btc7]. Hashrate peaked near 1.05 ZH/s in January 2026 before retreating to 937 EH/s by June — the first first-quarter hashrate decline since 2020, as miners pivot rigs to AI compute (Cipher's 15-year, approximately $5.5B AWS deal is the headline)[^s3_btc3][^s3_btc8][^s3_btc13]. With fleet-average production cost estimated near $90,000/BTC against a $63,951 spot, large swaths of the network are mining at a loss[^s3_btc8].

When a user pays $1.00 in Bitcoin transaction fees:

Direct Fee Recipients

• $1.00 — to the block-winning miner (Bitcoin has no burn, no protocol treasury, no developer cut from fees)[^s3_btc-flow]

Issuance-Funded Security Budget

• +approximately $150–180 — newly issued BTC distributed to miners alongside that same $1 fee[^s3_btc16]. More than 99% of miner income is issuance, not user payment. This is the security budget, by design — but at this ratio, fees are nowhere near replacing it.

Hidden Extraction / Off-Protocol Costs

• An estimated approximately $14.8B/yr in real-world energy, ASIC, and facility spend backstops the hashrate[^s3_btc-cost] — a cost that exists whether or not a single user transacts. Development is funded off-protocol via grants (approximately $12–15M/yr from OpenSats, Spiral, Chaincode)[^s3_btc10].

Total Ecosystem Value Flow: approximately $150–180 per $1 of user fees — the security budget is almost entirely issuance-funded.

Bitcoin is the cleanest expression of the report's framing. There are no token unlocks, no VC vesting cliffs, no foundation treasury — the entire security budget is protocol-level issuance, and it dwarfs fee revenue by two orders of magnitude. The long-running "security budget" debate is no longer academic: the day the subsidy halves to a number fees cannot replace is now closer than it is far.[^s3_btc12]

---

Solana: Strip Out the Meme Mania, and the Subsidy Remains

Solana's issuance machine minted approximately $1.57B in validator subsidies over the past year against $304.9M in user fees[^s3_sol2][^s3_sol4] — a 5.2x issuance-to-fee ratio measured on gross fees. SOL trades at $71.53, down approximately 44% since the October 2025 report[^s3_sol1]. The memecoin frenzy that made Solana's DEX volume look like it was challenging Ethereum has cooled — monthly DEX volume fell from a $145B October 2025 peak to approximately $42B by April 2026[^s3_sol15] — leaving a structurally issuance-dependent network underneath.

Inflation sits at 3.788% on the unchanged 15%-per-year disinflation schedule; SIMD-0411, which would have doubled the disinflation rate, was withdrawn without a vote in early 2026[^s3_sol2][^s3_sol10]. With 67.7% of supply staked[^s3_sol2], that mints approximately 22.0M SOL/year. A denominator note: Solana's gross fees are $304.9M, but DeFiLlama's retained-revenue line is just $35.7M[^s3_sol4] (50% of base fees are burned and priority fees pass through to validators) — so the multiplier below uses gross fees, the larger and more conservative base. On top of issuance, Jito MEV tips ran $295.0M over the trailing year[^s3_sol5] — a near-1:1 match with organic fees, and a vivid measure of how much value extraction rides alongside every transaction.

When a user pays $1.00 in Solana network fees:

Direct Fee Recipients

• approximately $0.95–0.99 — validators, via priority fees (100% to validators post-SIMD-0096; priority fees dominate fee volume)[^s3_sol9]
• approximately $0.01–0.05 — burned (50% of base fees only)[^s3_sol18]

Issuance-Funded Security Budget

• +$5.16 — inflationary issuance ($1.57B ÷ $304.9M gross fees)[^s3_sol4]. The validator security budget is majority issuance-funded.

Hidden Extraction (disaggregated)

• +$0.97 — Jito MEV tips ($295.0M ÷ $304.9M fees)[^s3_sol5]. As on Ethereum, a meaningful share is efficiency MEV (arbitrage, liquidations) rather than pure extraction; Jito's auction routes much of it back to stakers, making part of this protocol-captured rather than lost to users.[^s3_sol5]

Total Ecosystem Value Flow: approximately $7.12 per $1 of gross user fees — approximately 86% issuance plus extraction.

The structural story held even as the dollars fell. The Alpenglow consensus redesign entered community testnet on 11 May 2026, targeting 100–150ms finality versus today's approximately 12.8s[^s3_sol11], and Firedancer reached mainnet block production[^s3_sol12]. US spot SOL ETFs absorbed approximately $1.1B in cumulative inflows since their October 2025 launch[^s3_sol14] — institutions buying a 6–7% staking yield even as price fell. None of it changes the core arithmetic: for every visible dollar of fees, roughly $6 of issuance and extraction moves in the background.

---

BNB Chain: $214M in Fees, Billions in Corporate Burns

BNB Chain collects approximately $214.5M in trailing-twelve-month user fees[^s3_bnb3] while its issuer destroys an estimated $3.4–4.7B a year in corporate auto-burns[^s3_bnb6]. The value moving around the chain is roughly 16x what users pay for it at the current price — and unlike issuance-funded chains, this is deflationary corporate capital, not protocol inflation. That distinction matters: it is a value transfer funded by Binance Group's balance sheet, not a security budget and not user demand. BNB trades at $586.48, down approximately 57% from its $1,370 October 2025 ATH[^s3_bnb1].

The Fermi hard fork (14 January 2026) cut block time to 0.45 seconds[^s3_bnb7], making BSC the fastest EVM L1 by block interval. Three quarterly burns frame the corporate subsidy: the 33rd (Oct 2025, approximately $1.24B), 34th (Jan 2026, 1,371,803 BNB), and 35th (Apr 2026, 1,569,307 BNB)[^s3_bnb5][^s3_bnb6]. Annualizing the recent cadence at the current $586 price yields approximately $3.45B/year; at burn-time prices the figure was nearer $4.7B. Either way it towers over the $214.5M fee base. The retained-revenue line is smaller still — DeFiLlama reports $21.4M in BSC protocol revenue (the 10% BEP-95 burn share)[^s3_bnb3], so the gross-fee figure anchors the multiplier. On the demand side, BSC's RWA tokenization jumped 60% QoQ to $3.6B in Q1 2026 and stablecoin supply reached $17.9B[^s3_bnb10], repositioning the chain as an institutional settlement rail.

When a user pays $1.00 in BSC gas fees:

Direct Fee Recipients

• $0.90 — validators and delegators (90% of gas, via the ValidatorSet contract to 45 active PoSA validators)[^s3_bnb4]
• $0.10 — burned in real time via BEP-95 (approximately 286,000 BNB destroyed cumulatively)[^s3_bnb6]

Insider / Corporate Capital (not user-funded)

• +approximately $16.1 — corporate quarterly auto-burns ($3.45B annualized at the current price ÷ $214.5M fees)[^s3_bnb-flow]. Binance Group capital, not user payment and not protocol issuance.
• +approximately $0.47 — YZi Labs / builder-fund ecosystem grants ($100M Hash Global commitment atop an ongoing $1B builder fund)[^s3_bnb11]

Hidden Extraction

• Goodwill Alliance MEV protection holds sandwich attacks below 1K/day, versus a 140K/day pre-GWA baseline — extraction suppressed rather than monetized.[^s3_bnb-flow]

Total Ecosystem Value Flow: approximately $16–17 per $1 of user fees — backed by Binance Group capital, not organic revenue.

---

Cardano: The Treasury That Runs on Invisible Money

Cardano is the starkest issuance case among smart-contract chains in this report. It collected just $1.84M in user fees over the trailing twelve months[^s3_ada2]. The Ouroboros issuance engine simultaneously distributed an estimated $247M in new ADA to stake-pool operators and the on-chain treasury[^s3_ada3][^s3_ada-iss] — an approximately 134x issuance-to-fee ratio. ADA trades at $0.1615, down approximately 73% year-on-year and roughly 95% below its 2021 ATH[^s3_ada1].

The issuance is funded entirely from the unminted reserve pool (rho approximately 0.003/epoch on approximately 7.79B ADA of remaining reserves), split 80% to validators / 20% to the on-chain treasury[^s3_ada3]. Native DeFi TVL stands at $90.6M[^s3_ada2], with Minswap the largest protocol at $23.6M[^s3_ada13]. Retained protocol revenue is effectively a rounding error — DeFiLlama reports $89K for the trailing year[^s3_ada2b]. At current fee rates it would take roughly 134 years of user fees to match a single year of issuance.

When a user pays $1.00 in Cardano transaction fees:

Direct Fee Recipients

• $1.00 — to stake-pool operators (100% of fees; Cardano burns nothing and has no fee-funded protocol revenue)[^s3_ada-flow]

Issuance-Funded Security Budget + Treasury

• +approximately $107 — concurrent issuance to stake-pool operators (80% of approximately $134/$1 in issuance)[^s3_ada-flow]
• +approximately $27 — concurrent issuance to the on-chain treasury (20% share)[^s3_ada-flow]

Hidden Extraction

• $0 — no MEV layer of consequence, no burns; the entire developer-and-ecosystem apparatus (Project Catalyst Fund 15 at approximately $2.9M, Leios at approximately $4.4M) is issuance-funded, not fee-funded[^s3_ada10][^s3_ada5].

Total Ecosystem Value Flow: approximately $135 per $1 of user fees — almost entirely issuance-funded.

What's quietly notable is the governance. IOG's 2026 treasury ask was $46.8M — roughly half its 2025 figure[^s3_ada4] — and faced a real vote from approximately 1,000 elected DReps: six of nine proposals passed, one (Pogun, Bitcoin DeFi) was rejected at 32.4% support[^s3_ada12], and the community even vetoed Cardano Summit 2026[^s3_ada11]. The van Rossem hard fork (Plutus v11) was enacted 18 June 2026 — the first hard fork in Cardano's history initiated through on-chain governance[^s3_ada6]. The spending is more disciplined than it has ever been. It is still, almost in its entirety, invisible money.

---

Avalanche: "Deflationary" on Paper, Issuance-Funded in Fact

Avalanche burns 100% of its fees — and that fact is economically misleading. On a 30-day run-rate basis the chain is burning roughly $1.26M in fees per year (DeFiLlama's trailing-12-month figure is higher at $6.47M, inflated by busier earlier months)[^s3_avax3], while issuing an estimated $79M/year in new AVAX to validators[^s3_avax5]. On the run-rate basis, for every $1 a user burns, validators receive roughly $63 in fresh issuance; on the trailing-year fee base the multiple is closer to 12x. AVAX trades at $6.13, at multi-year lows[^s3_avax1]. Because Avalanche burns 100% of fees, gross fees and retained revenue are identical here — there is no separate revenue line to reconcile.

The "institutional honeymoon" met a reality check. Avalanche Treasury Co. (AVAT) listed on Nasdaq on 11 June 2026 via a $675M SPAC — and fell 16% on debut as the market confronted the gap between merger valuation and the approximately $90M in AVAX actually held[^s3_avax7]. Three spot AVAX ETFs (VanEck, Bitwise, Grayscale) launched and CME added futures[^s3_avax8][^s3_avax9], but the most credible demand driver was RWA: BlackRock BUIDL helped push tokenized assets to a record $1.16B in May 2026[^s3_avax10].

When a user pays $1.00 in Avalanche fees (all burned):

Direct Fee Recipients

• $1.00 — burned, permanently removed from supply (benefits all holders via deflation; no direct cash payment)[^s3_avax-flow]

Issuance-Funded Security Budget

• +approximately $63 (run-rate fee basis; approximately $12 on the trailing-year fee basis) — validators simultaneously receive newly issued AVAX from the 360M-token staking-reward allocation, entirely separate from and unfunded by user fees[^s3_avax6]
• Foundation grants (Retro9000's $40M pool, research grants, AVAT's approximately $90M treasury) underwrite ecosystem growth that organic fees cover none of[^s3_avax-flow]

Hidden Extraction / Off-Protocol Value

• Stablecoins and RWA assets sit atop the chain's approximately $461M of tracked DeFi TVL[^s3_avax-flow] — most dollar value on Avalanche lives outside the protocols that generate fees.

Total Ecosystem Value Flow: approximately $63 per $1 of run-rate fees (issuance-to-burn) — directionally an order-of-magnitude estimate, not a precise multiple.

At approximately $1.3M/year of run-rate fees, Avalanche's entire annual fee burn is dwarfed by a single mid-tier VC round. The deflationary label is technically true and economically secondary: the issuance subsidy is roughly 63x the burn at current activity.

---

Hyperliquid: The Fee-Funded Exception With an Insider-Supply Overhang

And then there's the exception. Hyperliquid runs a $1.063B trailing-twelve-month fee base[^s3_hype2] — among the top revenue-generating chains on earth — and recycles approximately 97% of it into HYPE buybacks via the Assistance Fund[^s3_hype5]. This is the one network in the section where users genuinely pay for what they get, and the retained-revenue line proves it: DeFiLlama reports $880M in trailing-year revenue[^s3_hype2], roughly 83% of gross fees — the inverse of the issuance-funded chains above. HYPE trades at $69.89, having set a fresh $76.70 ATH on 16 June 2026[^s3_hype1].

The fee engine is real: $81.5M in 30-day fees[^s3_hype2], $1.37B all-time, approximately $9.6B open interest, and approximately 40–44% of on-chain DEX-perp volume[^s3_hype2][^s3_hype7]. The Assistance Fund has accumulated approximately 44.4M HYPE (worth approximately $3.1B at the live price)[^s3_hype-af], and cumulative buybacks have crossed $1.5B[^s3_hype-bb]. The AQA v2 governance vote layered a second buyback stream — 90% of the yield on approximately $6.2B of on-platform USDC, an estimated $135–160M/year from October 2026[^s3_hype9].

But Hyperliquid's non-fee flow isn't issuance or VC — it is insider supply/value transfer via the team-unlock schedule. Since the November 2025 cliff, 9.92M HYPE unlocks on the 6th of every month through approximately November 2027; the 6 June 2026 tranche released approximately $693M in notional at the current price[^s3_hype12][^s3_hype-unlock]. Against approximately $81.5M in monthly fees, the buyback fund absorbs only approximately 11% of what the unlock schedule releases each month[^s3_hype-cover]. This is value transferred from new market buyers to insiders on a vesting clock — marked at market price, not a cash flow.

When a user pays $1.00 in Hyperliquid trading fees:

Direct Fee Recipients

• $0.97 — Assistance Fund, which buys HYPE on the open market (held, not burned)[^s3_hype-flow]
• $0.01–0.02 — HLP vault liquidity providers[^s3_hype-flow]
• $0.01–0.02 — HyperEVM gas and protocol operations[^s3_hype-flow]

Insider Supply / Value Transfer

• +$8.51 — team token unlock value released monthly (approximately $693M ÷ approximately $81.5M monthly fees)[^s3_hype-flow]. Not a subsidy to users — a supply overhang against them, and a notional mark, not cash.
• +approximately $0.15 — AQA v2 USDC reserve-yield buyback (an interest-rate transfer from Circle/Coinbase to HYPE holders)[^s3_hype9]

Off-Protocol Value

• A $6.0B HyperEVM ecosystem of 175+ dApps and an approximately $3.1B mark-to-market Assistance Fund treasury amplify every price move into billions of latent impact[^s3_hype4][^s3_hype-af].

Total Ecosystem Value Flow: approximately $10–11 per $1 of user fees — but inverted: the protocol is structurally fee-funded; the overhang is the risk, not the revenue.

Hyperliquid breaks the section's pattern in the most interesting way. It is not subsidy-dependent — it is overhang-exposed. The fees are real and the buyback is real, but the insider supply still entering the market each month is roughly 9x what the buyback absorbs. Whether the market absorbs the rest is a question of sentiment, not protocol mechanics.

---

L1 Networks: Patterns and Limitations

Seven chains, one verdict for six of them: the user-funded layer is a small fraction of total value flow. Across every issuance-secured network we measured, gross fees cover a single-digit-to-low-double-digit share of total value movement. The corrected, live-data multiples as of 20 June 2026:

| Chain | Dominant non-fee mechanism | $1 gross fee -> total flow | Non-fee share | |---|---|---|---| | Bitcoin | Issuance-funded security budget | approximately $160 | >99% | | Cardano | Reserve-pool issuance | approximately $135 | approximately 99% | | Avalanche | Validator issuance (vs 100% burn) | approximately $63 | approximately 98% (run-rate basis) | | BNB Chain | Corporate quarterly auto-burns | approximately $16 | approximately 94% | | Ethereum | Staking issuance + MEV | approximately $9.2 | approximately 89% | | Solana | Issuance + Jito MEV | approximately $7.1 | approximately 86% | | Hyperliquid | Insider unlock overhang (fee-funded) | approximately $10.5 | inverted — fee-funded |

A few patterns, and the limits of reading them too literally:

• The cleanest cases show the widest gap. Bitcoin and Cardano have no MEV, no VC schedule, no foundation cut from fees — and that simplicity is exactly why their non-fee ratio is so extreme. A pure issuance-funded security budget is, by construction, almost entirely non-fee. This is a designed property, not necessarily a defect — but it means fees are nowhere near covering the security spend.
• The "5–15% user-fee share" claim is scoped to this issuance-funded L1 sample. It is a statement about the seven networks here, not a universal law of blockchains. Fee-real exceptions exist and are growing: Hyperliquid funds its buyback entirely from trading revenue; Base retains a large share of sequencer fees as genuine protocol revenue. And a rising share of on-chain value is stablecoin settlement — dollars moving over rails like Tron and Ethereum that generate real, fee-bearing demand independent of the issuing chain's token economics. The subsidy framing applies to token-secured L1 economics, not to every dollar that touches a blockchain.
• Price compression flattered nobody and exposed everybody. Across the cohort, dollar-denominated issuance fell hard with token prices (ETH approximately −64%, SOL approximately −44%, AVAX at multi-year lows, ADA approximately −73% YoY), but the structural ratios held — because both issuance and fees deflate together. The machine got cheaper to run in dollars; it did not get more self-funding.
• "Deflationary" is a marketing word. Avalanche and BNB both burn fees, and both have a wider non-fee gap than Ethereum. A burn mechanism tells you nothing about whether users pay for the network — only about who benefits from the issuance or corporate capital that does the actual funding.
• Hyperliquid is the proof that fee-funded L1s can exist — and the proof of how rare it is in this cohort. Its risk is the opposite of everyone else's: not too little organic revenue, but too much insider supply still to vest.

Limitations. Three of the largest inputs are estimates, not hard data, and we flag them as such. Validator/staking issuance is calculated from published inflation parameters and live staking ratios (🔷 only on the price and supply inputs, not the derived totals). MEV (approximately $550M on Ethereum, $295M Jito on Solana) is sourced from research estimates and relay data, not a clean on-chain meter, and is partly efficiency and protocol-captured rather than pure extraction. Corporate/foundation/unlock totals (BNB burns at burn-time vs current prices, Avalanche grants, Hyperliquid unlocks) depend on price assumptions and partial disclosure; unlock and burn values are marked-to-market notional, not cash flows. The fee and revenue figures themselves are 🔷 HARD DATA from DeFiLlama. The multiples built on estimated numerators should be read as orders of magnitude, not decimals. The direction is unambiguous in every case; the precise multiple is not.

---

[^s3_eth1]: CoinGecko — Ethereum — ETH price $1,763.58, market cap approximately $212.8B, ATH $4,946.05 (Aug 24, 2025), now approximately −64% from ATH. Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_eth3]: Etherscan — Gas Tracker — Safe gas price approximately 0.19 Gwei (0.188 Gwei live); simple transfer approximately $0.006 at the live ETH price. Gas has roughly doubled off its sub-0.1 Gwei trough. Retrieved via Etherscan gas oracle API, June 20, 2026. 🔷 HARD DATA. [^s3_eth4]: DefiLlama — Ethereum Fees — Trailing-12-month gross fees $302.7M; 30d $11.14M; 24h $229,633. Retrieved via DefiLlama fees API, June 20, 2026. 🔷 HARD DATA. [^s3_eth4b]: DefiLlama — Ethereum Revenue — Trailing-12-month validator-retained revenue $115.3M (priority tips plus MEV-related tips; distinct from and smaller than the $302.7M gross-fee figure). Retrieved via DefiLlama revenue API, June 20, 2026. 🔷 HARD DATA. [^s3_eth5]: DefiLlama — Ethereum Chain TVL — DeFi TVL $39.0B. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_eth9]: Beaconcha.in — Staked Ether — Annual gross consensus issuance calculated as approximately 1,102,922 ETH/yr on approximately 39.67M staked ETH; at the live $1,763.58 price this is approximately $1.95B. Estimate derived from live staking ratio and the consensus reward curve; not 🔷. [^s3_eth10]: CoinLedger — Ultrasound Money Explained — Net ETH inflation positive at current activity; EIP-1559 burn (approximately 40K ETH/yr at the current fee rate) runs well below gross issuance (approximately 1.1M ETH/yr). Burn is highly variable with blob/L1 demand. June 2026. [^s3_eth11]: Blockworks — Fusaka Upgrade — Fusaka deployed Dec 5, 2025; EIP-7918 set a minimum blob base-fee floor. [^s3_eth13]: KuCoin Research — Ethereum Staking & MEV 2026 — Annual Ethereum MEV estimated approximately $550M. Expert/secondary estimate, not on-chain hard data; treat as order-of-magnitude. [^s3_eth14]: Flashbots — MEV-Boost and MEV Taxonomy — MEV disaggregates into extractive (sandwich/frontrunning, a user cost), efficiency (arbitrage/liquidations, supporting price discovery and protocol solvency), and protocol-captured (relay payments routed to validators). Used to qualify the MEV line rather than treat the full estimate as net user harm. [^s3_eth16]: CoinLaw — Ethereum Gas Fee Statistics — L2 networks handle approximately 95% of Ethereum transaction throughput. 2026. [^s3_eth17]: Ethereum.org — Gas and Fees — EIP-1559 base fee burned; priority-fee tip to validators. Base fee approximately 85% of total at current conditions.

[^s3_btc1]: CoinGecko — Bitcoin — BTC $63,951; market cap approximately $1.282T; ATH $126,080 (Oct 6, 2025), now approximately −49% from ATH. Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_btc2]: mempool.space — Block height 954,576; 3.125 BTC subsidy; trailing-144-block avg fees approximately 0.0172 BTC/block; daily fee revenue approximately $159K; daily issuance 450 BTC = approximately $28.8M; fees approximately 0.55% of miner revenue. Retrieved via mempool.space API, June 20, 2026. 🔷 HARD DATA. [^s3_btc3]: mempool.space — Hashrate & Difficulty — Hashrate approximately 937 EH/s; difficulty 124.9T. Retrieved via mempool.space API, June 20, 2026. 🔷 HARD DATA. [^s3_btc7]: BTC.network — Block Space Report, Mar 13–19, 2026 — Fee-to-revenue ratio well under 1%; block fullness high. ⏳ HISTORICAL (Mar 2026); used for trend context, superseded by live mempool data for current figures. [^s3_btc8]: CoinDesk — Bitcoin Hashrate Posts First Quarterly Drop in Six Years — Production cost approximately $90K/BTC; first quarterly hashrate decline since 2020; miners pivoting to AI. March 30, 2026. ⏳ HISTORICAL (Mar 2026). [^s3_btc10]: OpenSats — Bitcoin Core LTS Grant Program — OpenSats distributes approximately $1M/month in development grants; total ecosystem dev funding approximately $12–15M/yr (estimate, incl. Spiral, Chaincode). [^s3_btc12]: Cointelegraph — Bitcoin's Long-Term Security Budget Problem — Analysis of the fee-only security model as the issuance subsidy declines. [^s3_btc13]: Cointelegraph — Bitcoin Mining Outlook 2026 — Cipher Mining 15-year 300 MW AWS deal (approximately $5.5B projected); Core Scientific, IREN, TeraWulf pivoting to AI compute. [^s3_btc16]: mempool.space — Subsidy-to-fee ratio approximately 150–180:1: $10.5B annual block issuance (164,250 BTC × $63,951) vs approximately $58–69M annual fees at current throughput. Issuance and price inputs 🔷 HARD DATA; the ratio is derived. June 20, 2026. [^s3_btc-cost]: Estimated annual mining industry cost approximately $14.8B (approximately $90K/BTC production cost × 164,250 BTC mined/yr). Expert estimate combining the CoinDesk March 2026 cost figure and issuance volume; not audited, not 🔷. CoinDesk. [^s3_btc-flow]: Bitcoin fee flow: 100% of fees to the block-winning miner; no burn, no treasury, no fee-funded development. Per mempool.space block data and Bitcoin protocol design.

[^s3_sol1]: CoinGecko — Solana — SOL $71.53; market cap approximately $41.5B; circulating approximately 580.2M SOL; approximately −44% since the Oct 2025 report. Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_sol2]: Solana Compass — Tokenomics — Inflation 3.788%; staked approximately 67.7%; annual disinflation 15%. Annual issuance approximately 22.0M SOL (approximately $1.57B at the live price). Inflation parameters and staking ratio 🔷 HARD DATA; the USD total is derived. Retrieved June 20, 2026. [^s3_sol4]: DefiLlama — Solana Fees — Trailing-1y gross fees $304.9M; 30d $11.49M; 24h $359,958; trailing-1y revenue $35.7M. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_sol5]: DefiLlama — Jito — Jito MEV tips trailing-1y $295.0M; 30d $6.22M; protocol revenue 1y $18.1M. Jito's auction routes much of the tip value back to stakers, making part of this protocol-captured rather than lost to users. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_sol9]: The Block — SIMD-0096 — Validators receive 100% of priority fees; 50% of base fees burned. Live since Feb 2025. [^s3_sol10]: CoinPaper / Galaxy Research — SIMD-0411 Withdrawal — SIMD-0411 (double disinflation) withdrawn without a vote; 15%/yr schedule unchanged. [^s3_sol11]: CoinDesk — Alpenglow Consensus Testnet — Alpenglow entered community testnet May 11, 2026; targets 100–150ms finality vs approximately 12.8s. [^s3_sol12]: The Block — Firedancer Mainnet — Firedancer producing blocks on mainnet as of May 2026. [^s3_sol14]: KuCoin — Solana ETF Inflows — US spot SOL ETF cumulative inflows approximately $1.06–1.13B as of June 2026. [^s3_sol15]: BlockEden / CCN — Solana Metrics 2026 — Monthly DEX volume fell from $145B (Oct 2025 peak) to approximately $42B (Apr 2026); memecoin normalization. ⏳ HISTORICAL (Mar 2026) for the volume series. [^s3_sol18]: Solana Docs — Transaction Fees — Base fee 50% burned / 50% validator; priority fees 100% to validator post-SIMD-0096.

[^s3_bnb1]: CoinGecko — BNB — BNB $586.48; market cap approximately $79.0B (rank #4); ATH $1,369.99 (Oct 13, 2025), now approximately −57% from ATH. Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_bnb3]: DefiLlama — BSC Fees — Trailing-1y gross fees $214.5M; 30d $10.40M; 24h $220,861; trailing-1y revenue (10% BEP-95 burn share) $21.4M. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_bnb4]: BNB Chain — Introducing BEP-95 — 90% of gas fees to validators/delegators, 10% to the real-time burn address. [^s3_bnb5]: Crypto Economy — BNB 34th Quarterly Burn — 34th burn (Jan 15, 2026): 1,371,803.77 BNB. [^s3_bnb6]: CryptoSlate — BNB 35th Quarterly Burn — 35th burn (Apr 15, 2026): 1,569,307.34 BNB; cumulative BEP-95 burn approximately 286,000 BNB. Annualized recent cadence approximately $3.4–4.7B (estimate; $3.45B at the current $586 price, higher at burn-time prices). Not 🔷. [^s3_bnb7]: BNB Chain — Fermi Hard Fork — Fermi (Jan 14, 2026) cut block time to 0.45s. [^s3_bnb10]: Bitcoin.com — BNB Chain RWA Q1 2026 — RWA grew 60% QoQ to $3.6B; stablecoin supply approximately $17.9B (May 2026). [^s3_bnb11]: CryptoBriefing — YZi Labs BNB Holdings Fund — YZi Labs committed $100M to Hash Global's BNB Holdings Fund (2026), atop an ongoing $1B builder fund. [^s3_bnb-flow]: BNB dollar-flow multiple approximately 16.1x: $3.45B annualized quarterly auto-burns (at the current $586 price) ÷ $214.5M trailing gross fees. Estimate, not 🔷; at burn-time prices the burn approached $4.7B. GWA MEV protection suppresses sandwich attacks to <1K/day. Sources: DefiLlama — BSC, CryptoSlate.

[^s3_ada1]: CoinGecko — Cardano — ADA $0.1615; market cap approximately $6.01B (rank #20); approximately −73% YoY; ATH $3.09 (Sep 2, 2021). Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_ada2]: DefiLlama — Cardano — DeFi TVL $90.6M; trailing-1y gross fees $1.84M; 30d $57,760; 24h $1,198. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_ada2b]: DefiLlama — Cardano Revenue — Trailing-1y retained protocol revenue $89K — effectively a rounding error against issuance. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_ada3]: Cardano — Monetary Policy — rho approximately 0.003/epoch; tau (treasury fraction) = 0.20; 80% of issuance to stake-pool operators. [^s3_ada4]: CoinDesk — IOG Seeks $46.8M — IOG 2026 treasury request $46.8M, down approximately 52% from $97.5M in 2025. April 23, 2026. [^s3_ada5]: CryptoTimes — Cardano Leios Governance Vote — Leios approved at 84% DRep support; 27.7M ADA (approximately $4.4M) funded. May 25, 2026. [^s3_ada6]: Yahoo Finance — Cardano van Rossem Hard Fork — van Rossem (Plutus v11) enacted June 18, 2026; first governance-initiated hard fork. [^s3_ada10]: Project Catalyst — Fund 15 — 18.5M ADA (approximately $2.9M) + 250K USDM budget. [^s3_ada11]: CoinDesk — Cardano Governance Kills Summit 2026 — Summit proposal failed at 65.2% (needed 66.67%). June 1, 2026. [^s3_ada12]: IOHK Blog — IO Treasury Proposals Overview — Six of nine proposals approved; Pogun (Bitcoin DeFi) rejected at 32.4% DRep support. [^s3_ada13]: DefiLlama — Cardano Protocols — Top native protocol Minswap DEX $23.6M TVL. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_ada-iss]: Estimated annual ADA issuance approximately 1.53B ADA (approximately $247M at $0.1615): rho 0.003/epoch × 73 epochs × approximately 7.79B ADA reserves. Split 80% validators (approximately $197M) / 20% treasury (approximately $49M). Issuance-to-fee ratio approximately $247M ÷ $1.84M ≈ 134x. Derived from published protocol parameters; epoch amounts vary. Not 🔷. Cardano Monetary Policy. [^s3_ada-flow]: Cardano fee flow: 100% of fees to stake-pool operators, no burns. Concurrent issuance per $1 fee approximately $134 (approximately $107 to SPOs, approximately $27 to treasury); total approximately $135/$1. Derived from DefiLlama fees and protocol parameters.

[^s3_avax1]: CoinGecko — Avalanche — AVAX $6.13; market cap approximately $2.65B; multi-year lows. Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_avax3]: DefiLlama — Avalanche Fees — 30d fees $103,431 (approximately $1.26M annualized at this run-rate); trailing-1y $6.47M; all-time approximately $91M. All fees burned (gross fees equal revenue). Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_avax5]: Annual validator issuance estimate approximately $79M: approximately 3.0% inflation × 431.77M AVAX × $6.13. Inflation rate per Messari State of Avalanche Q4 2025. ⏳ HISTORICAL (Q4 2025 inflation rate); derived estimate, not 🔷. [^s3_avax6]: Issuance-to-burn multiplier approximately 63x on the 30-day run-rate fee basis (approximately $79M issuance ÷ approximately $1.26M annualized burn), or approximately 12x against trailing-1y fees of $6.47M. Validator rewards funded by the 360M-AVAX staking allocation, separate from user fees. DefiLlama — Avalanche Fees. [^s3_avax7]: CryptoBriefing — AVAT Nasdaq Debut Decline — Avalanche Treasury Co. (AVAT) listed June 11, 2026 via a $675M SPAC; holds approximately 15M AVAX (approximately $90M at spot); stock fell 16% on debut. [^s3_avax8]: The Defiant — Bitwise Launches Avalanche ETF — VanEck VAVX (Jan 26, 2026), Bitwise BAVA (Apr 15, 2026), Grayscale GAVA (Mar 12, 2026); stake up to 70–87% of AUM. [^s3_avax9]: CME Group — Crypto Suite Expansion — CME AVAX futures launched May 5–6, 2026. [^s3_avax10]: CoinJournal — Avalanche RWA Milestone — Tokenized assets hit a record $1.16B (May 2026); BlackRock BUIDL $625M on Avalanche. [^s3_avax-flow]: Avalanche fee flow: 100% of fees burned (deflation, no cash payment); validators receive approximately $63 of fresh issuance per $1 burned on the run-rate basis. Stablecoins and RWA sit above approximately $461M of DeFi TVL. Foundation Retro9000 ($40M pool) and grants underwrite ecosystem growth unfunded by fees. Estimate, not 🔷. Sources: DefiLlama — Avalanche Fees, avax.network — Retro9000.

[^s3_hype1]: CoinGecko — Hyperliquid — HYPE $69.89; market cap approximately $15.55B (rank #10); ATH $76.70 (Jun 16, 2026). Retrieved via CoinGecko API, June 20, 2026. 🔷 HARD DATA. [^s3_hype2]: DefiLlama — Hyperliquid Fees — 24h fees $1.57M; 7d $15.6M; 30d $81.5M; trailing-1y gross fees $1.063B; trailing-1y revenue $880M; all-time $1.37B. Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_hype4]: DefiLlama — Hyperliquid Protocol TVL — Ecosystem TVL approximately $6.0B (Hyperliquid L1 + Arbitrum). Retrieved via DefiLlama API, June 20, 2026. 🔷 HARD DATA. [^s3_hype5]: CoinShares Research — Hyperliquid Primer & 5-Year Valuation Framework — 97–99% of fees to the Assistance Fund for HYPE buybacks; approximately 44.4M HYPE accumulated; approximately 6–7% of all perps volume. June 2026. [^s3_hype7]: Coin Bureau — Aster vs Hyperliquid 2026 — Hyperliquid approximately $9.6B open interest, approximately 40–44% on-chain DEX-perp share. June 16, 2026. [^s3_hype9]: Crypto Briefing — Hyperliquid USDC Yield Buybacks (AQA v2) — AQA v2: 90% of yield on approximately $6.2B on-platform USDC to buybacks, approximately $135–160M/yr from Oct 2026. [^s3_hype12]: Yahoo Finance — Hyperliquid June Token Unlock — June 6, 2026 unlock 9.92M HYPE (approximately $693M notional at the live price); monthly cadence on the 6th through approximately Nov 2027. [^s3_hype-af]: Assistance Fund holds approximately 44.4M HYPE (approximately $3.1B at $69.89). Per CoinShares Research (June 2026) and Binance Square (40M+ confirmed Feb 2, 2026). Token count is reported; the USD value is marked at the live price. Estimate, not 🔷. [^s3_hype-bb]: Cumulative HYPE buyback spending >$1.5B since launch. Per CryptoTimes (June 2, 2026). Estimate, not 🔷. [^s3_hype-unlock]: Tokenomist — Hyperliquid Vesting Schedule — Core-contributor cliff Nov 2025; approximately 9.92M HYPE/month thereafter through approximately Nov 2027. [^s3_hype-cover]: Monthly unlock vs buyback coverage approximately 11%: 9.92M HYPE × $69.89 ≈ $693M unlocked vs approximately $79M absorbed (97% of 30d fees plus AQA v2/12). Unlock value is marked-to-market notional, not cash. Calculated June 20, 2026. DefiLlama — Hyperliquid Fees. [^s3_hype-flow]: Hyperliquid fee flow: $0.97 to Assistance Fund buybacks (held, not burned), approximately $0.01–0.02 each to HLP vault LPs and HyperEVM operations. The non-fee flow is insider supply/value transfer via the team-unlock schedule (approximately $693M/month ≈ $8.51 per $1 fee), a notional mark rather than issuance, VC, or cash. Sources: CoinShares Research, DefiLlama.

---

Layer 2 Networks: Fee Distribution

Layer 2 rollups were sold as the engine that would make Ethereum cheap, fast, and self-funding. As of 20 June 2026, the four most-watched rollups collectively bill users a few million dollars a month in sequencer fees — and run economies an order of magnitude larger on token issuance, insider unlocks, and corporate or VC subsidy. Where a base layer like Bitcoin runs an issuance-funded security budget, an L2 funds its own existence: on three of the four chains below, sequencer revenue cannot cover the operation. Each subsection traces a single user dollar, then names the multiple of non-fee-funded value flowing underneath it — keeping the four buckets distinct: organic user fees, consensus/issuance, external VC capital, and insider supply transferred through token unlocks.

A note on the macro: the crypto market these L2s settle into has compressed hard since the October 2025 baseline. ETH trades around $1,732 (up roughly 2% in a week, and itself a volatile input that moves every dollar figure below), and the data-availability cost of posting an L2 batch to Ethereum has been gutted twice — first by Pectra (May 2025), then by Fusaka/PeerDAS (December 2025) — cutting L1 settlement costs by a further 40–60% on top of the post-Dencun collapse.[^s4_1] Cheaper settlement is good for users and brutal for L2 income statements: the one cost that used to justify the toll is now a rounding error, and so is the toll. One framing caveat applies throughout: the "hidden multiples" below mix realized cash flows (fees, VC) with mark-to-market notional supply (issuance, unlocks), and notional value is endogenous to the same token price that deflates the fees. They are directional measures of value-at-stake, not like-for-like cash comparisons.[^s4_44]

---

Base — Coinbase's Corporate Toll Road

Base is the outlier that proves the rule: it is the only major L2 here that behaves like a profitable business, because a roughly $60B public company runs the sequencer and keeps the change. Base collected $77.5M in sequencer fees in full-year 2025 — down approximately 13% from 2024's $88.9M as trading volumes softened, but still enough to make Base the #1 L2 by fees with an estimated 62% of all L2 fee revenue.[^s4_2][^s4_3] Over the trailing 30 days it booked $5.10M in fees (gross, what users paid) against $5.10M in revenue (net, after L1 costs) — implying just $9,010 in L1 blob costs, a settlement bill equal to 0.18% of fees after Pectra expanded blob capacity.[^s4_2][^s4_4] TVL sits at $4.2B, off the approximately $4.4B January 2026 peak but still the largest L2 by a wide margin.[^s4_5] All-time sequencer fees since the August 2023 launch now total $205.9M.[^s4_2]

The structural event of 2026 was the divorce. In February 2026 Base announced it was leaving the OP Stack, ending the revenue-share arrangement that fed the Optimism Collective.[^s4_6] Over the 2.5-year partnership Base paid Optimism 8,387 ETH — roughly 41% of the Collective's lifetime revenue and over 90% of its monthly revenue right before the exit.[^s4_7][^s4_8] In dollar terms that is approximately $14.5M at today's ETH price (and approximately $14.2M at the slightly lower price used elsewhere in this report); the higher "$16.4M" figure sometimes quoted implies ETH near $1,955, which is the partnership-period average across August 2023–February 2026 — a historical price, not a current mark.[^s4_7] Post-divorce, Coinbase keeps essentially everything.

When a user pays $1 in Base sequencer fees (post-OP departure):

• $0.998 → Coinbase sequencer profit. Near-total capture by the corporate parent. No more Optimism cut since February 2026.[^s4_6]
• $0.002 → Ethereum L1 blob fees. ETH burned for data availability, collapsed to near-zero post-Pectra (the pre-Pectra rate was approximately 5%).[^s4_4]
• $0.00 → Optimism Collective. Was approximately 14.3 cents under the old deal; now zero.[^s4_7]

The hidden multiple: roughly $5–7 per $1 of sequencer fee. This is the rare case where the multiple isn't a subsidy indictment — it's app-layer economics. Apps on Base generated an estimated $369.9M in 2025 revenue (Aerodrome alone approximately $160.5M) against $77.5M in sequencer fees, a 4.8x ratio of protocol economy to toll.[^s4_9] One clarification the headline ratio hides: both numbers are gross — app-layer revenue is not net profit, and sequencer fees are not net sequencer margin — so 4.8x understates how different the underlying operating economics are; it compares two top-lines, not two bottom-lines.[^s4_9] Layer on undisclosed sequencer MEV and Coinbase's stablecoin float income, and the visible sequencer fee is roughly the top 15–20% of what actually moves.[^s4_10][^s4_11]

On MEV specifically, the report's taxonomy requires disaggregation rather than a single black box. Base's Flashblocks design gives Coinbase's centralised sequencer 200ms priority blocks, and the MEV captured within them spans three economically distinct flows: extractive (sandwich and front-running, a one-way transfer from users), efficiency-improving (arbitrage and liquidations, which keep prices and lending markets honest), and protocol-captured priority fees. All three are retained by Coinbase and none are separately disclosed in any public filing, so the total is an estimate, not hard data.[^s4_11] On the float income: Coinbase reported $305M in stablecoin revenue in Q1 2026 — up 55% year-on-year on a record approximately $19B average USDC balance held in Coinbase products — of which a material but undisclosed share is Base-driven.[^s4_14] That $305M is a verified line item from Coinbase's Q1 2026 10-Q; the "Base-driven portion" of it is an estimate.[^s4_14]

One caveat that cuts the other way: a native BASE token has not launched — exploration was announced in September 2025, and prediction markets assign roughly 69% odds to a launch before end-2026.[^s4_12][^s4_13] If it ships with typical insider/VC vesting allocations, an insider-supply/value-transfer column appears — the same "unlock" bucket that dominates zkSync below, and the one that would convert Base from a fee-real outlier into a subsidised chain. Footnote [^s4_8]'s original 118M OP token agreement is not an insider unlock — it was an inter-chain revenue-share commitment, now voided by the exit.[^s4_8]

---

Arbitrum — Break-Even Sequencer, Bottomless Treasury

Arbitrum is the anti-Base: nobody pockets the margin, because there is barely a margin to pocket. The sequencer runs at a break-even mandate, with all surplus routed to the Arbitrum DAO treasury.[^s4_15] The problem is the surplus has nearly vanished. Trailing 30-day fees are $383,724 — annualising to under $5M — while the Arbitrum Foundation asked its own DAO for $43.5M in a single funding request, roughly 1.85x the entire $23.49M gross revenue of 2025.[^s4_16][^s4_17] The chain that secures roughly $15.6B in value (the #1 L2 by total value secured) cannot pay its own staff out of its own fees.[^s4_18]

The token tells the rest. ARB trades at $0.0834, down approximately 96.5% from its $2.39 ATH, with a $531M market cap.[^s4_19] The DAO treasury is approximately 93% ARB — a position now worth roughly $224M, down from $651M in January 2026 — meaning the treasury's value collapses in lockstep with the token it is supposed to fund operations with.[^s4_20] Meanwhile ARB unlocks continue at roughly 92.65M tokens/month — approximately $7.7M of monthly insider supply/value transfer (team, investor, and DAO-tranche vesting marked at market price), outpacing monthly fee revenue by roughly 20x; the next DAO tranche unlocks 16 July 2026.[^s4_21]

When a user pays $1 in fees on Arbitrum One:

• $0.31 → Ethereum L1 data availability. Blob/calldata reimbursement; the L1 share of a much-smaller total post-Fusaka (midpoint estimate; July 2025 token-flow data showed approximately 4.6% direct sequencer reimbursement, but L1's share of the shrunken fee base now runs 25–35%).[^s4_22]
• $0.69 → Arbitrum DAO treasury. All sequencer surplus, denominated in ETH and stablecoins, per the official fee-distribution model.[^s4_15]
• $0.00 → sequencer operator. Offchain Labs takes no fee margin — unique among major L2s.[^s4_15]

On top of base fees sits Timeboost, the express-lane priority auction launched April 2025: $7.5M cumulative, annualising approximately $5.94M, roughly 25% of total DAO revenue — though its 30-day take has compressed to $155K as the novelty premium fades.[^s4_23]

The hidden multiple: roughly $8–12 per $1 of fees. Dividing annualised ARB unlock value (approximately $92M/year of insider vesting at current prices), the approximately $20M+ structural DAO deficit, and VC-funded Offchain Labs opex (the company raised $120M+ in 2021–22 to run the sequencer at zero margin) by approximately $4.6M of annualised fee revenue yields a chain where roughly eight to twelve dollars of non-fee-funded value — split across insider unlocks, treasury drawdown, and VC subsidy — move for every dollar a user actually pays.[^s4_21][^s4_24] The unlock value is mark-to-market notional, not cash; the VC opex is realized cash. Arbitrum is a venture- and issuance-funded public good, not a self-sustaining business.

---

Optimism — The Anchor Chain Becomes a Rounding Error

If Arbitrum can't fund itself, OP Mainnet barely registers. The chain that anchors the Superchain generated $56,377 in fees over the trailing 30 days — annualising to under $700K on a run-rate basis — against approximately $1.88M over the full prior year.[^s4_25] OP trades at $0.1012, down approximately 97.9% from its $4.84 ATH, with a $218M market cap.[^s4_26] The gap between OP's roughly $435M fully diluted valuation and its sub-$700K run-rate fee revenue now exceeds 600x.

Two events defined Optimism's 2026. First, Base walked out (February 2026), stripping the Superchain of the tenant that had supplied approximately 41% of all Collective revenue ever and approximately 87% of recent sequencer revenue; OP fell 28% in 48 hours.[^s4_27] Second, in January 2026 governance approved (84.4%) a buyback program redirecting 50% of net Superchain revenue to monthly OP purchases for a 12-month pilot — launched, with grim timing, just as the revenue base was about to exit through the front door.[^s4_28]

When a user pays $1 in gas on OP Mainnet:

• $0.03 → Ethereum L1 data costs. Blob/calldata posted to Ethereum validators, post-EIP-4844.[^s4_29]
• $0.97 → Optimism Collective treasury. OP Mainnet routes 100% of net sequencer profit to the public-goods engine — every cent above L1 cost.[^s4_30]

The hidden multiple: roughly $5.6 per $1 of fees, almost all of it issuance. Against approximately $1.88M of annualised fees, the chain prints approximately 85.9M new OP/year via 2% inflation — roughly $8.7M of fresh supply, a 4.6x issuance ratio.[^s4_31] This is consensus/governance issuance, not a temporary external subsidy — but on a chain whose fees cover well under a quarter of it, the directional point stands. The Feb-2026 buyback offsets part of it (approximately $4.97M/year, approximately 2.6x of fees) but offsets inflation, not the eroding fee base.[^s4_28] Roughly 2.135B OP (approximately $216M) remains locked through 2029 — a continuous insider supply/value-transfer overhang including the approximately 31M OP Core-Contributor unlock in May 2026.[^s4_32] RetroPGF — once the industry's flagship public-goods model — distributed 16M OP in 2025, worth approximately $1.62M today versus approximately $20M+ at 2024 prices; the model survives, but the token collapse gutted the real-dollar value of every grant.[^s4_33] The remaining Superchain (ex-Base) holds approximately $522M TVL across nine chains, with Unichain (approximately $23M DefiLlama TVL) nowhere near replacing Base's multi-billion footprint.[^s4_34]

---

zkSync Era — A Fee Machine Running on Vesting

zkSync Era is the purest illustration of the L2 unlock problem because the fees are too small to round. Trailing 30-day fees are $14,371 — about $175K annualised.[^s4_35] TVL has cratered from an approximately $541M 2024 peak to $15.3M today, a 97% collapse.[^s4_36] ZK trades at $0.0116, down approximately 96% from its $0.321 ATH, $116M market cap, approximately $244M FDV.[^s4_37]

Matter Labs has effectively pivoted away from the public chain: it announced a second round of layoffs, committed the company to "Prividium" (a permissioned, privacy-focused L2 for regulated institutions), and sunset zkSync Lite in early 2026.[^s4_38] A November 2025 tokenomics overhaul redirects interop and licensing revenue — not Era transaction fees — to ZK buybacks, burns, and staking.[^s4_39]

When a user pays $1 in fees on zkSync Era:

• $0.30 → Ethereum L1 data + proof costs. Blob data availability plus proof verification, amortised across the batch (estimate; varies with congestion).[^s4_40]
• $0.70 → Matter Labs sequencer profit. Retained by the still-fully-centralised sequencer operator. The ZKnomics value-accrual mechanism explicitly excludes Era transaction fees.[^s4_41]

The hidden multiple: roughly $217 per $1 of fees — the most lopsided in this report, and almost entirely insider supply. Team (13.55%) and investor (17.19%) allocations total 33.33% of the 21B supply and, post-June-2025 cliff, unlock roughly 286.56M ZK/month — approximately $3.3M of monthly insider supply/value transfer marked at market against $14,371 of monthly user fees, an approximately 217:1 ratio.[^s4_42][^s4_43] This is the cleanest case in the report of value transferred from new market buyers to insiders by a vesting schedule, not earned from users; it is mark-to-market notional, but the selling pressure it represents is real. Behind it sit unrealised governance reserves (Token Assembly approximately $67.8M, Ecosystem Initiatives approximately $46.1M) and an estimated approximately $450M in VC funding subsidising Matter Labs off-chain — a cyclical, capital-formation flow distinct from the unlocks.[^s4_44] The fee revenue is economically immaterial; the ZK economy runs on vesting, not users.

---

L2 Networks: Patterns and Limitations

Step back from the four chains and a single structure repeats. Sequencer fees are trivial and shrinking; the real economy is issuance, insider unlocks, and subsidy. The numbers as of 20 June 2026:

| Chain | 30d fees | Annualised | TVL | Token vs ATH | Hidden multiple per $1 fee | |---|---|---|---|---|---| | Base | $5.10M[^s4_2] | approximately $61M[^s4_2] | $4.2B[^s4_5] | (no token) | approximately $5–7 (app economy, not subsidy)[^s4_9] | | Arbitrum | $384K[^s4_16] | approximately $4.6M[^s4_16] | $1.30B[^s4_18] | ARB −96.5%[^s4_19] | approximately $8–12 (insider unlocks + deficit)[^s4_24] | | Optimism | $56K[^s4_25] | <$0.7M[^s4_25] | $306M[^s4_34] | OP −97.9%[^s4_26] | approximately $5.6 (issuance)[^s4_31] | | zkSync Era | $14.4K[^s4_35] | approximately $175K[^s4_35] | $15.3M[^s4_36] | ZK −96%[^s4_37] | approximately $217 (insider unlocks)[^s4_42] |

Three patterns hold across all four:

1. The DA-cost collapse broke the toll model. Pectra and Fusaka cut L1 settlement to near-zero, which was meant to be the L2's margin. Instead it removed the cost the toll was justifying. Base monetises anyway because Coinbase owns the rail; the others collect fees that no longer cover operations.[^s4_4][^s4_1]

2. Issuance and insider unlocks, not user fees, fund the chain. Arbitrum's DAO requested approximately 1.85x its annual revenue; Optimism prints approximately 4.6x its fees in annual issuance; zkSync transfers approximately 217x its fees to insiders every month through vesting. In every case the visible user fee is a fraction of the non-fee-funded value flowing to token holders and future unlock recipients.[^s4_17][^s4_31][^s4_42]

3. Ownership decides who captures the dollar. A corporate sequencer (Base) keeps 99.8 cents; a break-even/public-goods model (Arbitrum, Optimism) keeps approximately 0 and routes everything to a treasury or the Collective; a centralised-but-tokenised model (zkSync) splits with L1 and lets insiders extract via vesting. Same toll, radically different beneficiaries.

The limitations of this framing are real and worth stating. "Total value secured" is not revenue — Arbitrum's $15.6B TVS and Base's $4.2B TVL represent user capital, not income, and an L2 captures only the thin fee layer on top.[^s4_18][^s4_5] The hidden multiples mix categories that are not equivalent — and not even the same unit: app-layer revenue (Base) and VC opex are realized cash, whereas issuance (Optimism) and insider unlock pressure (zkSync) are mark-to-market notional supply, endogenous to a token price that also deflates the fee denominator.[^s4_44] Both inflate the "$X per $1" headline, but app revenue signals economic activity while insider unlocks are a one-way wealth transfer; the buckets must be read separately, not summed into one undifferentiated "subsidy." And MEV, stablecoin float, and private corporate cross-sells are undisclosed estimates, not hard data — flagged as such throughout. The multiples are directional indictments, not audited income statements.

The L2 sustainability question, plainly. Outside the one chain with a corporate balance sheet behind it, no major L2 in this report earns enough to fund itself. The standard rollup pitch — cheap fees today, fee revenue scales with adoption tomorrow — has collided with two facts: adoption did not produce proportional fee revenue (Optimism's fees fell as the Superchain grew), and the DA-cost collapse means the per-transaction take keeps falling even when usage holds. What fills the gap is issuance (Optimism), treasury drawdowns funded by a token-heavy reserve that deflates with the token (Arbitrum), insider vesting (zkSync), or a corporate parent (Base). Three of those four are running down a finite resource. The rollup economy, stripped of narrative, is a set of public goods waiting to discover whether anyone will pay for them once the subsidy runs out — and on current numbers, the non-fee-funded value flows are winning by two-to-three orders of magnitude.

---

---

[CONTENT CONTINUES BELOW - See attached files for complete report]

The full 225KB report continues with:

• Detailed case studies for each blockchain (Bitcoin, Ethereum, Solana, etc.)
• Layer 2 networks analysis (Arbitrum, Optimism, Base, zkSync, etc.)
• Infrastructure provider economics
• Foundation and VC funding flows
• All visualizations and charts
• Complete footnotes and references

📥 DOWNLOAD COMPLETE REPORT:

• master.html (1.2MB) - Full interactive report with charts (open in browser)
• master.md (220KB) - Complete text
• PDF (2.8MB) - Print version

All files attached below as release assets.