Fees Fund Three of the Ten Modelled Years

BSAHI — interpretation note Produced: 2026-09-23

What the SCCR does and does not say

The Storage Cost Coverage Ratio is 0.4307 — an upper bound. That number is easy to over-read. It does not say "Bitcoin is 57% underfunded," and it does not say "nobody is paying for nodes."

What it says, in one sentence: on the current fee basis, fees fund roughly the first 3 of the 10 modelled years of the node-cost commitment, and the remainder is carried by whoever runs a node.

What the model actually claims

The SCCR is a ratio of one fee reading to a multi-year commitment:

SCCR  = fees / L_net        L_net = N · C · T

where N = the reachable-node count (a floor), C = annual node operating cost ($925/yr), and T = 10 years. So the denominator is the network-wide node cost over a decade, and the numerator is one fee reading. The ratio is therefore a coverage fraction of that commitment, not a solvency verdict.

Per block, 2026, all on one fee basis (measured 30-day mean, ~$1,417/block):

At 30.3% of a ten-year commitment, fees cover ~3.0 years. That is the finding.

A trap in that table — and why it is worth naming

Those last two rows are the same cost at different horizons. L_net is exactly node_opex × T (measured ratio 10.001), so their coverage percentages differ by exactly T (30.3% × 10 = 303.0%). They are one fact, not two.

We nearly published them as two. The version of this note I first drafted argued "fees cover node operation but not permanence." That is an artifact. There is no separate permanence term in the model — permanence is the multi-year node cost. Reading the pair as "opex covered, permanence not" is a horizon illusion, and the table invited it by showing a 1-year and a 10-year figure side by side unlabelled.

Both the table and the headline now say so explicitly, because the next person to read those two rows will make the same mistake.

Who carries the remainder

The uncovered ~70% is not absorbed by the fee market. It lands on the node classes that actually hold the data — public and private/non-listening nodes alike. Every hidden node independently bears the replication and validation burden, so the measured floor makes the true coverage lower, not higher.

Two other classes receive the verification guarantee without carrying it: institutional infrastructure (exchanges, custodians — partially, and internally) and delegated/lightweight users (not at all, and not observable in the P2P layer). See the verification-population observatory for why those two cannot be measured from this vantage.

Two caveats that must travel with the number

• The fee window matters, and it is named. The SCCR is a live reading on its own block window (137 blocks, 2026-09-23), which implies ~$2,014/block. On the 30-day mean (~$1,417/block) the storage coverage is 30.3%; on the SCCR's own window it is 43.1%. Both are true. Quoting one as "the" coverage while the other is the headline is a category error, so the coverage table uses one basis and the SCCR is reported separately.
• This is time-varying, not a constant. SCCR is driven by the fee market, which moves. The 3-of-10 figure is the current reading, not a fixed property.

What this does not say

It is not a claim that Bitcoin is failing, nor that anyone should pay differently. It is a measurement of which layer the fee market currently funds and which it leaves to node operators. The interesting question is the second one — and it is answerable, which is why the distinction is worth keeping sharp.

Data

• data/fee_allocation.json — the three claims, one fee basis, and the horizon note.
• data/sccr.json — the live SCCR reading and its own window.
• research/sccr-sensitivity.md — why N dominates the uncertainty and why the baseline is a strict floor.

← All research · ← Back to Learn