Daily Fees

Daily Fees plots the network’s aggregate fee revenue in USD per day on a logarithmic right axis spanning $100 to $100M. The rust-tinted area is the raw daily total; the white line on top is the 7-day simple moving average that damps single-day inscription spikes. Price runs muted on the log left axis for cycle context; halving markers mark each subsidy-cut step. The log axis is mandatory here — on a linear scale the 2010–2014 floor and the $81M Runes spike cannot share a chart.

Today’s reading is $173.3K; the 7-day average is $209.3K; the live-recomputed fee share at today’s $65,837.03 spot is 0.6%. The fee market is in the Quiescent regime — one of four named bands defined in the next section.

Fee data is too volatile to read off the latest day. A useful regime read needs both a level (the 7-day average) and a context (the share of miner revenue). Crossing either threshold in either dimension is enough to move the regime label; both thresholds in both dimensions confirm it. Four named regimes cover the post-2017 history.

Quiescent. 7-day SMA below $1M and fee share below 5%. The mempool is cleared most blocks, fees compete on the floor, and miners are paid almost entirely from the subsidy. Fits the bulk of 2018–2019, the mid-2025 lull, and 2026 to date.

Active. 7-day SMA between $1M and $5M and fee share between 5% and 15%. The default steady-state regime in the late-2020s — some blocks fill fully, fee competition is real but not severe. Fits most of 2024 outside event days.

Congested. 7-day SMA between $5M and $25M and fee share between 15% and 35%. Mempool backlog is sustained; fee competition is meaningful enough to deter low-priority transactions. Fits the late-2017 mempool shake-out, the 2021 spring run, and the 2023 Ordinals waves outside their peaks.

Eruption. 7-day SMA above $25M and/or fee share above 35%. Fees are within an order of magnitude of the subsidy, sometimes above it. Has fired at the December 2017 cycle-top mempool wave, the 2023 Ordinals second wave, and most distinctly around the April 2024 halving when the Runes launch pushed a single-day share to 73.8%. Eruption regimes lasting more than a week are rare; lasting more than a month, unprecedented.

The regime read is descriptive, not normative. A Quiescent regime says nothing about whether prices will rally or fade; an Eruption regime says nothing about whether the underlying activity is wallet payments or inscription metadata. It is a frame for what kind of fee-market environment the network is in.

Three identities, one schedule. For every day d:

fees(d) = sum of (inputs − outputs) across all confirmed transactions in d
subsidy_usd(d) = reward(d) × 144 × price(d)
share(d) = fees(d) / (fees(d) + subsidy_usd(d))

The fee surplus is the protocol-level definition: every transaction’s inputs minus outputs, summed over the block, paid to the miner alongside the block subsidy. The 7-day SMA is a simple unweighted mean over the trailing seven days; the share denominator uses the protocol’s halving schedule (the same one that drives the daily issuance chart) and the day’s closing price. The reading row above re-prices the subsidy denominator against live spot so the share ticks with the BTC price through the day; the chart and the cycle-anchor table use the daily-close pricing for cross-cycle consistency.

The series spans 17 Jan 2009 to 14 Jun 2026, sourced from blockchain.info’s transaction-fees-usd daily chart. Pre-2020 daily resolution is sparser than the recent tail by source design — closer to a four-day cadence — so the SMA windows treat the available points as the trend signal. The rolling-window choices are documented on the methodology page.

Three useful lenses, in increasing order of analytic value. Level reading. The 7-day SMA against the four regime bands; the live regime label is shown in the reading row. Share reading. Fee share of miner revenue at today’s spot — the second axis of the regime map and the long-run security-budget gauge. Cycle-aligned reading. The cycle-anchor table below puts each anchor in price context; this is where the cross-cycle comparisons earn their keep.

The shape of the chart on a log axis is the entire fee-market history at readable amplitude: a near-zero floor through 2010–2014, a step up in the 2017 mempool wave, a noisier post-SegWit middle through 2018–2022, and the 2023–24 inscription/Runes era pushing fees within an order of magnitude of the subsidy on event days. The trend is non-monotonic; reading it as “fees grow each cycle” is a stronger claim than the data supports.

Reading fees at canonical cycle dates is the cleanest summary of how the chart moves with both price and protocol-level demand events. The 2017 mempool peak, the 2021 cycle top, the 2023 inscription waves, and the April 2024 Runes launch are the four event clusters that account for most of the chart’s vertical range over the last decade. The 2025–2026 stretch has been defined by the absence of an event cluster.

The single rarest event in this chart is a day on which fees exceeded the block subsidy — share above 50%. It has happened 1 time in seventeen years of daily data, on the day the Runes protocol launched. That day, 20 April 2024, block 840,000 was mined at the halving boundary, fees came in at $81.1M against a subsidy of roughly $29M — a fee share of 73.8%. For one day the post-2140 economy was a working preview; no other single day in the record has come close to crossing the line.

The pattern matters because Bitcoin’s long-run security model rests on that share rising structurally, not just spiking on event days. A subsidy halving every four years means the absolute USD figure of the subsidy compresses unless price is rising; if price stagnates, fees are the only remaining miner-revenue lever. The April 2024 cluster is the first time the chart has stress-tested that thesis at non-trivial scale — and it lasted a day, not a quarter.

A rally fakes a fee surge. The headline is in dollars, so it reflects both sat-per-vByte fee pressure and BTC price. In a flat fee market, a price doubling prints a doubled USD-fee line with no extra transactions sent — the 2021 spring run shows the trap: on 10 November 2021, at the all-time price high, daily fees were only $1.2M and the fee share a mere 2.0%. The dollar line was near its cycle peak; the fee market was quiet. Read the share alongside the level, never the level alone.

One block can move the day. A coordinated wallet movement, an inscription mint, or a single high-fee transaction can swing the daily total by a factor of three. The 7-day SMA exists for exactly this — the Eruption threshold is set on the SMA precisely so a lone bad day cannot trip it. The April 2024 Runes day is the cleanest example: the raw total hit $81M, but the SMA that week sat near $18M. Use the raw daily number for event reading and the SMA for regime calls.

The share rises mechanically as the subsidy falls. Fee share has a moving denominator: every halving cuts the subsidy in half, so the share can drift up over years even if absolute fee demand is flat. Reading the long-run share trend as proof of “structural” fee growth overstates the case — some of that rise is just the denominator shrinking. The daily issuance companion chart is the right cross-reference to separate the two.

The early history is thin. The blockchain.info source drops to roughly a four-day cadence before 2020, so SMA windows in that era are effectively wider than their names. Treat the 2009–2019 fee floor as approximate context, not a precise series; the post-2020 daily resolution is reliable.

If you hold for the security-budget thesis, the fee share is the cleanest live read. A long-run drift toward higher steady-state fees would confirm that fee revenue can carry security through to 2140; a flat or declining calm-period floor would undercut it. The chart to date supports neither cleanly — event spikes are real but transient, and the quiet-period floor has not lifted materially across the last three cycles. Pair it with daily issuance to watch the subsidy half of the equation step down.

If you run hash rate, the 7-day SMA is the planning number. Sustained Active or Congested regimes change the marginal economics of deployment; sustained Quiescent regimes squeeze marginal operators and show up first in the hash-ribbon capitulation read. The four regimes map loosely to four operating stances: defensive (Quiescent), neutral (Active), opportunistic (Congested), and exceptional (Eruption).

If you are decomposing blockspace demand, read fees alongside daily transactions: the count tells you whether fee pressure is capacity-bound (count flat, fees rising) or demand-bound (count rising, fees rising in lockstep). And compare fee share, not absolute fees, across cycles — the latter conflates sat-per-vByte pressure with BTC price.