Decent Newsroom
Sign In

The Only Interest Rate Nobody Sets

By Roger November 2, 2026
Bitcoin sits 38 percent below its record one year on. Across 1,178 funding settlements the long side paid 3.70 percent of notional to stay in the trade, which annualises to 3.44 percent, below the rate the Federal Reserve charges for overnight money. What the price of borrowed conviction says about time preference, the halving, and a debt growing 6.71 percent a year.
The Only Interest Rate Nobody Sets

Bitcoin trades at 78,161 dollars this morning. The record was 126,080 on 6 October 2025, which leaves the price 38.0 percent below the high, 392 days later. That number is on every screen today and it is not the interesting one.

Here is the number that is missing from every screen. Over those 392 days, the people who stayed long with borrowed money paid 3.70 percent of their position for the privilege.

Rent, settled every eight hours

Perpetual futures have no expiry date, so the contract has to be tied to the spot market by something. On Binance that something is a funding payment, settled three times a day, moving between whoever is long and whoever is short. Positive means the long side pays. Negative means the short side pays. No committee meets, no statement is published, nobody votes. A few hundred thousand accounts disagree about the next eight hours and the disagreement is charged as a price.

I pulled all 1,178 settlements from the record high to this morning. Summed, the long side paid 3.70 percent of notional. Annualised over 392 days, that is 3.44 percent.

Now line up the other number. The average effective federal funds rate over the same window was 3.72 percent.

So borrowed exposure to bitcoin was priced at 3.44 percent a year, in the same 392 days in which overnight dollars cost 3.72. Calling that “being paid to be wrong” would be too strong, since the position still carries margin and collateral costs. Cheap is the right word for it. A position that is cheap to hold does not capitulate in an afternoon. It grinds.

Bitcoin perpetual funding, daily cost and cumulative, September 2025 to November 2026

Where the sign flipped

Funding is not a constant. Between February and April this year, across 267 settlements, the sign reversed. Cumulative funding for that stretch came to minus 0.33 percent, and 58.1 percent of the individual prints were negative. For three months the shorts were paying the longs, which is what a crowded short book looks like from the inside. The heaviest single day came on 6 February, when the three settlements together ran the other way and the short side paid 0.029 percent of notional. By May the monthly sum was positive again.

Then the crowd changed its mind further. Over the last thirty days the long side has paid 0.42 percent, which annualises to 5.11 percent, and the most recent print, at 08:00 UTC this morning, was positive as well.

Line the two rates up. Across the full 392 days, the market’s price of time sat 0.27 points below the committee’s. Across the last thirty, it sits 1.23 points above it. The gap moves every week, and what moves it is the crowd’s opinion about the next eight hours, not about the next decade.

Thirty-day funding annualised against the effective federal funds rate

Interest is a price. This one is discovered.

Böhm-Bawerk spent the 1880s on a question that looks trivial until you try to answer it: why does interest exist at all? His answer was that it is not a fee for money and not a reward for waiting in the moral sense. It is an agio, a premium that present goods carry over future goods, because the future is uncertain and because a person who has the thing now can put it to work now. Every interest rate in a market is a measurement of that premium at a particular place and time.

Mises drew the practical conclusion a generation later. If the rate charged for credit sits below the rate that reflects how people actually value present against future goods, then projects get financed that should never have been financed, and the bill arrives later as a correction. The correction is not a disease. It is the invoice.

Set that against a 3.44 percent cost of holding a long position through a year of falling prices, and the shape of this drawdown explains itself. Borrowed conviction was cheap. Cheap conviction does not capitulate quickly. It bleeds: position by position, week by week, at a rent that never quite hurts enough to force the sale in one afternoon.

None of this makes a funding rate into a pure interest rate. It is a positioning premium as much as a price of time, and it says more about who is crowded than about how a society discounts the future. Read it as the fee for a claim, not as the philosophy of the claim.

The other side of the ledger

There is a second set of numbers, and it is arithmetic rather than interpretation.

Since the halving on 19 April 2024 the subsidy has been 3.125 coins per block, which at 144 blocks a day comes to 450 coins, and 164,250 over twelve months, against a circulating supply of 20,105,365. The supply grew 0.82 percent.

The United States federal debt went from 38.04 trillion dollars at the end of October 2025 to 40.59 trillion on 29 October 2026. That is 2.55 trillion more, a growth rate of 6.71 percent.

M2, the broad money measure the Federal Reserve publishes itself, went from 22.15 trillion in September 2025 to 23.41 trillion in September 2026. Growth of 5.71 percent.

Annual growth of supply: bitcoin 0.82 percent, US M2 5.71 percent, federal debt 6.71 percent

Put those three bars next to each other and the entire Austrian case fits inside one image. The quantity of the thing being priced does not grow, except by a small amount that keeps getting smaller. The quantities it is priced against are neither. One of these numbers is produced by a schedule written in 2009 and enforced by a network that no one administers. The other two are produced by decisions taken in rooms.

At the next halving, expected in March 2028, the subsidy drops to 1.5625 coins per block and that 0.82 percent becomes roughly 0.40 percent. Nothing about the debt or the money supply is scheduled to do anything similar.

Cantillon, the miners, and the exit

Cantillon noticed in the 1730s that new money is never neutral, because it enters at a point. Whoever stands at that point spends before prices adjust, and whoever stands furthest from it receives the higher prices before the higher income. The effect is not a conspiracy. It is geometry.

Bitcoin’s issuance has an entry point too: whoever burns electricity to find a block. That is not a fair arrangement in any moral sense, and nobody serious claims it is. What it does have is visibility. The reward is paid at a market price, the recipient has to sell part of it to cover the power bill, and the sale happens continuously in public. There is no first recipient with the option to buy assets before the price has moved, because the sale is how the new supply reaches the market.

Thiers’ law completes the thought. Bad money drives out good only when the exchange rate between them is fixed by decree. Let people choose and the harder money tends to win, slowly and without drama. Bitcoin’s share of the whole crypto market sits at 60.0 percent this morning while the altcoin season index reads 47, the neutral middle of its range. Within this asset class, capital keeps arriving at the one constituent whose quantity cannot be expanded by anyone. That is what the choice looks like when people make it rather than a rule.

What this does not say

Supply arithmetic is a slow tide. Price is a vote taken every eight hours, and the vote can run against you for years. A 0.82 percent issuance rate did not prevent a 38 percent drawdown and it will not prevent the next one. Anyone who tells you the ratio determines the date is selling something.

Two things follow from the measurements above, and neither of them is a target.

If funding keeps running above the policy rate while the price grinds sideways, longs keep paying rent for a claim that is not moving, and the exit keeps happening the way it has all year: not a crash, a slow transfer of patience from one set of hands to another.

If funding turns negative for several weeks again, the short side is the one paying, and the cost of forcing it to close becomes cheap. February to April this year was that configuration, with three negative monthly sums in a row, and by May the monthly figure had turned positive again.

The sentiment reading this morning is 37, which is fear. Turnover was 17.1 billion dollars against a market value of 1.57 trillion, about 1.1 percent, which is a market that has stopped arguing with itself. Nothing in that is a signal. It is a description.

The one claim I will defend without hedging is narrower than a forecast and harder to argue with. The quantity of bitcoin is fixed, the quantity of the dollars it trades against is not, and the difference between 0.82 and 5.71 is not an opinion about the future. It is what happened in the last twelve months. The rent on a long position, annualised, came to 3.44 percent against a policy rate that averaged 3.72. The long side paid in 76.8 percent of the settlements, and what it was paying for fell 38 percent. One of those rates is decided in a room. The other is decided by everyone who shows up.

Sources

Write a comment