What capital costs, in bitcoin
Every decision a node operator makes about liquidity is a decision about the cost of capital, and almost nobody states what that cost is. This lesson states it, from a curve this operator computes daily, and then shows what it does to a channel-lease decision.
- 01 Channels are not accounts
- — Comfort reading a rate as an annual percentage
- — What a channel lease is
- Say what a lease costs you against the alternative use of the same capital
- Reject a liquidity offer priced above the measured rate without arguing
- Reproduce the fixing from the published method
Ask a routing node operator what their capital costs and you will usually get a shrug, or a number from a blog post. It is the input every liquidity decision depends on and the one nobody measures, because measuring it means computing a curve rather than repeating one.
This operator computes one daily. Here is today’s.
What three-month bitcoin-denominated capital cost on 2026-09-11, across 3 venues. At this rate, 10 BTC lent for a quarter returns 0.0164 BTC less than it started with.
- Venues
- 3
- Confidence
- interpolated
- Custody tier
- T4
- USD curve
- 2026-09-10
Read the sign before the magnitude. A negative bitcoin-denominated rate means the market prices bitcoin forward below spot: someone holding bitcoin and willing to sell it forward is paid to do so, and someone wanting bitcoin later pays less than someone wanting it now. That is the ordinary state of this market and not an anomaly.1
What that does to a lease
A channel lease is a purchase of inbound liquidity for a term. The seller locks capital in a channel with you and charges for it; the question is whether their price beats what you would otherwise do with the same capital.
Most operators answer that against zero. If bitcoin held in a channel earns nothing, then any lease fee is a pure cost and the instinct is to minimise it. But zero is the wrong baseline, and the measured rate is the reason: capital held at tier 1, in cold storage, is not earning zero in bitcoin terms — it is tracking a rate that is usually slightly negative.
A model, not a measurement. Every row below is arithmetic on a 5 000 000 sat lease held for one quarter, against a 450 ppm forwarding fee; the slider interpolates between them and computes nothing of its own. The cost of capital it is measured against is the figure above.
| Lease price, per quarter | Fee paid | Annualised | Break even at |
|---|---|---|---|
| 100 ppm | 500 sats | 0.04 % | 1 111 k sats forwarded |
| 500 ppm | 2 500 sats | 0.20 % | 5 556 k sats forwarded |
| 1 500 ppm | 7 500 sats | 0.60 % | 16 667 k sats forwarded |
| 3 000 ppm | 15 000 sats | 1.20 % | 33 333 k sats forwarded |
| 5 000 ppm | 25 000 sats | 2.00 % | 55 556 k sats forwarded |
Without JavaScript this block is the table itself — the same five rows the slider interpolates between, which is also what prints.
The arithmetic is not subtle and that is the point. A lease priced at 500 ppm for a quarter is 5 000 sats on five million — less than a single on-chain channel open at almost any fee environment, and cheaper than the opportunity cost of the same capital sitting somewhere it can lose value. A lease at 3 000 ppm is a different proposition, and now the channel has to actually route to justify itself.
What the measured rate gives you is the ability to say which of those you are looking at without arguing about it.
What is not here yet
This block will carry the median clearing price for a rebalance matched through Boltring — the price of the thing a lease substitutes for.Reason: only 0 of 0 buckets cleared k-anonymity (minimum 5 participants); a median over fewer than 3 is not a measurement.
That gap matters more than it looks. Leasing inbound liquidity and rebalancing into it are substitutes: they buy the same thing by different means, and the sensible choice is whichever is cheaper today. With one price measured and the other missing, this page can tell you what a lease costs against your capital but not what it costs against the alternative.
It will say so until it can say otherwise. A page that filled the gap with a figure from a blog post would be the content farm this site exists not to be.
Reproduce it
The fixing is computed from public venue data every day and the method is published. Two commands get you the same numbers:
curl -s https://satrates.com/data/curve.json | jq '.headline'
curl -s https://satrates.com/data/curve.json | jq '.venues[] | {venue, ok, n_points}'
The second is the one worth running. It tells you how many venues answered and how many points each contributed, which is what separates a fixing from a quote — and on a day when only one venue answers, this page shows no figure at all rather than a number with a footnote.
- 01 A provider offers you 5 000 000 sats of inbound liquidity for three months at 2 500 ppm. The measured three-month rate is negative. Is the offer expensive?
2 500 ppm over three months is roughly 1% annualised on the amount leased, which sounds small and is not the comparison that matters. The right comparison is against what that capital would otherwise earn you — and if the bitcoin-denominated rate is negative, the alternative use is losing money slowly, so the lease clears a lower bar than it first appears. Expensive against what is the whole question.
- 02 The block above says the rate is negative. Does that mean you are paid to borrow bitcoin?
Not by anyone you can reach. It means the market prices bitcoin forward below spot at that tenor, which is what a lender who wants bitcoin rather than dollars faces. Realising it requires posting collateral at a venue, which is custody tier 4 — you have swapped a rate for a counterparty. The figure is a measurement of a market, not an offer to you.
- 03 Two of the blocks on this page are missing. Should you trust the one that is not?
More than you would if the missing ones were quietly absent. The two gaps each say which source is not producing and why, so you can see what the figure that is here does and does not depend on. A page that renders three numbers and tells you nothing about where they came from is the one to distrust.
Notes
- The rate is measured across venues at custody tier 4 — funds at an exchange, keys not yours. A tier 1 or tier 2 operator cannot realise it without changing tiers, and that change is the cost the headline figure does not include.Sat Rates methodology
Source & licence
This page adapts no licensed material. It is original work by Delta Gamma LLC and is published under CC BY-SA 4.0; the figures it reports are separately dedicated to the public domain, below.
Measurements are the author's own and are dedicated to the public domain (CC0). Cite as: ln.university, LN 300 lesson 4, "What capital costs, in bitcoin", figures as of the date shown on each block.