Issue 11 documented the mechanics — Ordinals, BRC-20, and Runes turned Bitcoin's block space into contested real estate. Each protocol generated dramatic fee spikes. Each spike decayed rapidly. The mechanics work. The demand is real.
This issue asks the question that connects everything back to Issue 9: does that demand actually solve the security budget problem?
The answer requires looking at fee revenue like an investor examines earnings — focusing on the baseline rather than peaks. The question centers on what the market generates on an average day when nobody is minting and the mempool sits nearly empty.
The Fee Data, Honestly
January 2023 — Ordinals Launch
Within three weeks of launch, inscriptions consumed approximately 50% of Bitcoin block space. Average block size climbed from roughly 1.4 MB toward 2.5 MB. Fees rose modestly as initial curiosity drove adoption rather than congestion.
May 2023 — BRC-20 Mania
This represented the first genuine stress test. Average transaction fees surged to $16–$31 per transaction from a baseline of $1–4. BRC-20 inscription activity represented more than half of all transaction fees on May 7–8. The mempool backlog exceeded 465,000 unconfirmed transactions. On May 8, total fees per block exceeded the 6.25 BTC subsidy for the first time since 2017, persisting roughly ten days before the minting frenzy exhausted per-ticker limits and collapsed. Within weeks, fees returned to baseline levels.
December 2023 — Ordinals Peak
Average fee rates reached 245 sats/byte on December 15, the highest since the 2017 bull run. Inscription fees comprised 28% of total fee revenue that day, with transactions per block reaching an all-time high of 4,688. Activity persisted through early 2024, pushing average block size to 2.293 MB by March. This elevated state lasted roughly a month before following the same trajectory: up, then down.
April 20, 2024 — Runes Launch/Halving Day
This marked the all-time high. Daily fee revenue hit $80 million — a record by a significant margin. Total miner revenue that day reached $107.8 million. The average fee per transaction stood at $128. Each block earned 18.62 BTC in fees against the newly halved subsidy of 3.125 BTC. Fees were six times the subsidy. Within the first two months, Runes generated 2,513 BTC in fees (approximately $163 million), with 53% accruing on day one alone. The next day brought $22 million; by April 28, this declined to $1.03 million — a 98% drop in eight days.
August 22, 2024 — Babylon Staking Launch
The Babylon protocol opened Phase 1 mainnet for native BTC staking. Its 1,000 BTC capacity filled within 74 minutes. Fees spiked from 0.14 BTC to 9.52 BTC in two blocks as 12,720 stakers competed for confirmation. This lasted hours; by the next day, fees normalized.
Late 2024 Through 2025 — The Floor Collapses
Inscription and Runes activity faded substantially. Daily Ordinals inscriptions dropped from over 400,000 at peak to under 3,000. Fee contribution fell below 1% of miner revenue — as low as 1.25% in March 2025. Each block earned roughly 0.025 BTC in fees against the 3.125 BTC subsidy. Daily fee revenue sank to approximately $300,000–$500,000 against roughly $30–45 million from the subsidy. Near-free blocks became common with average fees at 1 sat/vbyte during quiet periods. Total Bitcoin transaction fees for 2024 were approximately $920 million for the full year — inflated by the Runes ATH. Despite that, 2025 came in below it, making it one of the weakest fee years in recent memory.
Early 2026 — Still at the Floor
As of March 2026, fees remain near trough levels at approximately $300,000 per day — less than 1% of total miner revenue. There is no recovery. The 2025 low has simply extended into 2026. Whether a new demand catalyst will emerge is this issue's central question.
Block Space as a Market
Bitcoin produces approximately 4 MB of block space every ten minutes. That supply doesn't change. There is no central bank of block space that can expand issuance during demand surges or contract it during lulls. This fixed supply is the constraint enabling a fee market.
Huberman, Leshno, and Moallemi's 2021 paper "Monopoly Without a Monopolist" established a central finding: Bitcoin's fee market only generates meaningful revenue when demand for block space exceeds supply — when the mempool is congested and transactions must compete for inclusion. Without congestion, fees approach zero because there's no scarcity to bid against. The 1 sat/vbyte periods in 2025, when blocks had spare capacity, prove this empirically.
Ordinals and successors proved the theory from the opposite direction. When inscription demand filled blocks to capacity, fees rose dramatically. The market cleared at $128 per transaction on halving day. Users valuing block space most — Runes minters willing to pay premium rates — achieved confirmation. Low-value transactions were priced out. This is exactly how a competitive market for scarce resources should behave.
However, there's a difference between proving the mechanism works and proving demand is durable. The block space market has demand from several sources behaving very differently.
Financial transactions — payments, exchange withdrawals, consolidations — form the baseline. They're relatively stable but generate modest fees because most payment transactions aren't time-sensitive enough to bid aggressively. Before Ordinals, this baseline kept fees at 1–3% of miner revenue on non-congested days.
Inscription and token transactions — Ordinals, BRC-20, Runes — are event-driven. They create acute congestion during launches and speculative waves, then retreat. The first economic analysis of Ordinals found that inscription transactions individually tend to carry lower fees than payment transactions in the same block — but their sheer volume reduces available space and forces everyone else to bid higher. The fee revenue from inscriptions is partly direct payment and partly an externality imposed on other users.
Staking and protocol transactions — like Babylon — generate intense but extremely brief demand. A few hours of fee congestion represents real revenue, but not a business model.
Roughgarden's 2024 Journal of the ACM paper on transaction fee mechanism design adds another layer. Even if demand grows, the fee mechanism's design determines how much that demand translates into miner revenue. Bitcoin's first-price auction mechanism — where miners simply include highest-bidding transactions — captures all fees directly. But Roughgarden showed that alternative designs (like Ethereum's EIP-1559) burn significant fee portions rather than paying miners — trading miner revenue for price stability and collusion-resistance. Bitcoin lacks EIP-1559, so miners keep everything. The trade-off is that Bitcoin's fee revenue depends entirely on raw congestion dynamics. There's no base fee creating a revenue floor regardless of demand levels.
Compare this to Ethereum, where EIP-1559 burned 2.6 million ETH in its first year after activation — destroyed entirely, creating deflationary pressure tied to ongoing usage from DeFi, stablecoins, NFTs, and staking. Ethereum's fee revenue reflects persistent, diversified demand. Bitcoin's reflects episodic, event-driven demand. This is a structural difference, not cyclical, and Bitcoin's architecture doesn't create recurring demand floors the way Ethereum's does.
The Sustainability Test
Applying Issue 9's security budget framework to actual fee data:
Current state (March 2026): Bitcoin trades at approximately $74,000. The block subsidy is 3.125 BTC per block, roughly 450 BTC per day, worth approximately $33 million daily. Fee revenue sits at approximately $300,000 per day, bringing total miner revenue to roughly $33–34 million daily — almost entirely from the subsidy. Hashrate sits at 924 EH/s. Fees represent less than 1% of miner revenue.
At the 2028 halving, the subsidy drops to 1.5625 BTC per block — approximately 225 BTC per day. At the current $74,000 price, that's roughly $16.7 million daily from the subsidy alone. To maintain today's ~$33–34 million daily security spend, fees would need to cover approximately $16–17 million per day.
From the current $300,000 daily fee baseline, that's a roughly 55x increase — not a modest gap, but a structural chasm.
If Bitcoin's price rises to $100,000 by 2028, the subsidy provides ~$22.5 million daily. The fee gap narrows to ~$11 million — still a 37x increase needed from today's baseline. At $150,000, the subsidy covers ~$33.75 million and the fee gap nearly closes without any fee growth. At $200,000, the subsidy alone exceeds current security spend.
This is the optimist's argument, and it's not wrong on its own terms: if BTC price appreciation outpaces the subsidy decline, the dollar-denominated security budget can hold or grow even as the block reward shrinks. Every halving cycle so far has seen sufficient price appreciation to more than offset the subsidy cut.
But there are three problems with relying on it.
First, past price appreciation is not a mechanism. The difficulty adjustment guarantees scarcity. Nothing guarantees price growth. Extrapolating four halvings of price appreciation into a permanent security model is not engineering — it's faith.
Second, even at higher BTC prices, the ratio of security spend to secured value matters. Budish's framework showed that attack costs must remain high relative to the value the network secures. If BTC triples in price while security spend stays constant, the secured value grows faster than the attack cost. The ratio deteriorates even as absolute numbers look fine.
Third, the 2032 halving makes the arithmetic much harder. The subsidy drops to 0.78125 BTC per block — roughly 112.5 BTC per day. At $100,000, that's $11.25 million from the subsidy. At $74,000, it's $8.3 million. Fees would need to be not just a supplement but the majority of miner revenue. From the current daily baseline, fees would need to grow roughly 80x just to maintain 2026 security levels at 2032 — and the subsidy will have dropped another 50%.
What the Fee Spikes Actually Tell Us
It's tempting to look at spike data and see either hopeless volatility or proof of concept. Both readings miss what the spikes actually demonstrate.
The positive case is real. Before Ordinals, the strongest security budget bear case was that Bitcoin block space had no demand beyond payments — and payment demand was migrating to Layer 2. Ordinals, BRC-20, and Runes disproved this. People paid $128 per transaction to mint Runes. During the Babylon staking launch, two blocks collectively generated 9.52 BTC in fees as stakers competed for the limited cap. The fee market cleared at prices that would have seemed absurd in 2022. Latent demand for Bitcoin block space — for data availability, cultural artifacts, token issuance, and staking coordination — exists and is willing to pay.
Each spike is also a stress test of the fee mechanism itself, and it passed every time. When demand exceeded supply, fees rose. Low-value transactions were priced out. High-value transactions paid market rates. The mechanism Huberman et al. described theoretically — congestion-driven fee competition — worked exactly as modeled when real congestion materialized.
But the problem is equally real. Every spike was event-driven — a launch, a cap fill, a halving — rather than demand-floor-driven. After each event, demand evaporated because there was no recurring use case keeping inscription transactions in mempool competition. Compare again to Ethereum, where DeFi lending, stablecoin transfers, and NFT trading create daily block space demand persisting regardless of whether any particular protocol just launched. Bitcoin's block space demand has no equivalent structural floor.
There is no 2026 recovery to point to. Fees remain below $300,000 per day — structurally indistinguishable from the 2025 trough. The absence of a new demand catalyst is itself the data point. The block space market, absent a minting frenzy or protocol launch, defaults to near-zero congestion and near-zero fee income above the payment baseline.
A few weeks of congestion per year cannot substitute for sustained daily revenue. The security model requires the latter.
The Layer 2 Paradox, Revisited
Issue 9 flagged a tension in Bitcoin's scaling strategy: the Lightning Network routes transactions off-chain, reducing mempool competition and therefore L1 fee revenue. Ordinals and Runes create the opposite pressure — more on-chain activity, more fees. These forces are in direct opposition.
A 2024 paper on Lightning's impact (ResearchGate, August 2024) examined this tension directly. Its finding confirms the intuition: as small transactions migrate to Lightning, L1 loses exactly the volume it needs to maintain congestion. Lightning channels require opening and closing transactions on-chain, but these are infrequent relative to the volume of payments they facilitate. The net effect is a reduction in L1 fee pressure.
Lightning's own trajectory adds uncertainty. Network capacity peaked and then dropped approximately 20% by 2025. Whether this reflects a temporary plateau or a fundamental adoption ceiling is unclear, but it means that even the L2 pressure on fee revenue may be less severe than feared — because L2 adoption itself may be slower than projected.
The deeper tension is philosophical. Bitcoin's scaling roadmap has long been "small blocks, Layer 2 for payments." This makes engineering sense for throughput. But it creates a direct conflict with the security model: every transaction that moves to L2 is a transaction that doesn't compete for L1 block space, and therefore doesn't generate L1 fee revenue. The community has historically treated scaling and security as separate problems. They are not. They are the same problem viewed from different angles.
Inscription protocols push in the other direction — more L1 demand, more fees — but as the data shows, that demand is episodic. The net effect of Lightning reducing baseline fee demand and inscriptions adding spiky fee demand is a fee revenue profile that is volatile and unpredictable. For security budget forecasting, volatility is almost as bad as low revenue, because miners need to make capital allocation decisions — buying ASICs, securing power contracts — based on expected future income. A revenue stream that swings between $300,000 per day in 2025 and $80 million per day in 2024 is not one you can build a business on.
Wu, Shi, and Chung's 2024 paper at ITCS adds a mechanism design wrinkle: maximizing miner revenue requires fee market structures that Bitcoin currently doesn't have. Bitcoin's simple first-price auction captures fees efficiently when the mempool is congested, but it provides no revenue floor when it isn't. The question isn't just whether demand will grow, but whether Bitcoin's fee market is designed to convert that demand into reliable miner income. Right now, the honest answer is: only when blocks are full.
Where I Might Be Wrong
The strongest argument against everything written above is historical precedent. Bitcoin has survived every "existential threat" narrative for seventeen years. The security budget concern has been raised in academic literature since at least 2016 (Carlsten et al.), and nothing catastrophic has happened. Every halving has been followed by sufficient price appreciation to maintain or increase dollar-denominated security spend.
You may be overweighting the fee data's importance relative to price appreciation. If Bitcoin reaches $200,000 by 2028, the subsidy alone provides $45 million per day, matching current security levels without any fee growth at all. The price-driven security model has more empirical support than the fee-driven one, even if it's less satisfying as engineering.
You may also be underestimating how quickly new demand sources can emerge. BitVM-enabled smart contracts could bring DeFi functionality to Bitcoin's L1, generating structural demand resembling Ethereum's rather than episodic inscription waves. Institutional settlement — BlackRock and Fidelity ETF operations, nation-state reserve management — could create a baseline of high-value L1 transactions independent of speculative cycles. Neither exists at meaningful scale today, but neither did Ordinals in December 2022.
The speculative froth interpretation may be wrong in a subtler way. Ethereum's NFT market looked like pure speculation in 2021. Five years later it still generates meaningful daily volume and fee revenue. The Ordinals ecosystem is young. The decline from 400,000 daily inscriptions to under 3,000 looks like death, but it could also look like the trough before maturation — the way Ethereum's entire ecosystem had a total value locked under $1 billion in early 2020 before DeFi Summer sent it above $10 billion within months.
Finally, you're treating Bitcoin's fee market as if it must solve the security budget problem on its own. In practice, the solution might be hybrid: sufficient price appreciation to cover most of the gap, combined with periodic fee spikes during high-demand events, combined with gradual growth in baseline fee revenue from maturing Bitcoin-native protocols. No single factor closes the gap. All three together might.
The honest position: the data does not support confidence that fees will replace the subsidy on schedule. It also does not support confidence that the system will fail. The uncertainty itself is the finding — and for a $1.4 trillion network, uncertainty about the long-term security model is not a comfortable place to be.
The Question This Issue Doesn't Answer
The mechanics work. The demand is real but episodic. As of March 2026, fees remain near $300,000 per day — less than 1% of miner revenue, and roughly 55x below what the 2028 halving demands at current prices. The gap is not closing. It is sitting unchanged, waiting for the next halving to make it significantly worse.
Issue 13 puts all the pieces together. It builds scenario models for the 2028 and 2032 halvings — conservative, base, and optimistic — and makes falsifiable predictions. Not philosophy. Not price speculation. Specific, date-bound claims about fee revenue, pool concentration, miner AI-pivot rates, and the fee-to-subsidy ratio that you can come back and check.
If Issue 11 was about what moved in, and Issue 12 was about whether the rent covers the mortgage, Issue 13 asks: what does the building inspector's report actually say?
Sources & Further Reading
Fee data and market analysis
- IntoTheBlock / Daily Hodl, "Bitcoin Daily Fees Hit New All-Time High of $80,000,000 Following Launch of Runes" (April 2024)
- Glassnode, miner revenue and fee decomposition data (2023–2026)
- The Defiant, "Bitcoin Fees Normalize After Runes Coming-Out Party Abruptly Ends" (April 2024)
- CoinDesk, "Miners Reap Windfall as Runes Debut Sends Transaction Fees to Record Highs" (April 2024)
- Cointelegraph, "Bitcoin Runes Earn 2,500 BTC in Fees in 2 Months" (June 2024)
- The Block, "Bitcoin Miner Fees Fall to 12-Month Low, Underscoring Long-Term Reliance on Block Subsidies" (2025)
- The Block, "2026 Bitcoin Mining Outlook" (2026)
- CoinGecko Research (2024) — Bitcoin earned ~$920M in total transaction fees in 2024
- Crypto.com Research, "Bitcoin Ordinals" (March 2024) — 63 million inscriptions, ~6,370 BTC in fees
Staking and protocol events
- Decrypt, "Bitcoin Fees Skyrocket as Babylon Launches Native BTC Staking" (August 2024)
- The Miner Mag, "Bitcoin Fee Spike from Babylon Staking" (August 2024)
- PR Newswire, "Babylon Phase-1 Mainnet Staking Cap Extended" (August 2024)
Academic literature
- Huberman, G., Leshno, J., & Moallemi, C. (2021). "Monopoly Without a Monopolist: An Economic Analysis of the Bitcoin Payment System." The Review of Economic Studies, Vol. 88(6), pp. 3011–3040
- Roughgarden, T. (2024). "Transaction Fee Mechanism Design." Journal of the ACM
- Wu, X., Shi, E., & Chung, H. (2024). "Maximizing Miner Revenue in Transaction Fee Mechanism Design." ITCS 2024
- "On the Impact of the Lightning Network on Bitcoin Transaction Fees and Network Value." ResearchGate (August 2024)
- "Bitcoin Ordinals: Determinants and Impact on Total Transaction Fees." ScienceDirect (2024)
- Carlsten, M. et al. (2016). "On the Instability of Bitcoin Without the Block Reward." ACM CCS
- Budish, E. (2025). "Trust at Scale: The Economic Limits of Cryptocurrencies and Blockchains." Quarterly Journal of Economics 140(1)
Block space and inscriptions
- Investing.com, "Inscriptions Are Filling Bitcoin Block Space, Consumes 50% of Space" (February 2023)
- Glassnode, "The Week On-Chain, Week 20" (May 2023) — BRC-20 congestion analysis
- OBM.io, "Bitcoin Fee Market Dynamics" — BRC-20 fees exceeded 50% of total transaction fees on May 7–8, 2023
- Grayscale Research, "Market Byte: Awaiting Confirmations" (May 2023)
- Blockchain.news, "Bitcoin UTXO Set Jumps to 169M in Ordinals Era"
- NFT Evening, "Bitcoin Network Activity Hits Lowest" (2025)
Geo Nicolaidis
Builder, TrailBit.io
If you found this useful, subscribe to get the next issue in your inbox. Each issue breaks down a different heuristic used in Bitcoin forensics — what it assumes, where it breaks, and why it matters.