What funding rate and open interest actually say

A futures contract with an expiry date converges on the spot price because it has to: at expiry it settles. A perpetual contract never expires, so nothing forces it back. Funding is the mechanism put in its place — a periodic payment from one side of the contract to the other, sized by how far the contract has drifted from spot, whose stated purpose is to keep the two from separating.

The direction follows the premium rather than anybody's opinion. When the contract trades above spot, longs pay shorts; when it trades below, shorts pay longs. The payment is between the traders themselves: Hyperliquid states that funding is purely peer-to-peer and that no fees are collected on the payments, and Binance states that it does not charge fees on funding payments and that they transfer directly between traders holding opposing positions. This is the first thing usually got wrong about it — it is not a charge the venue levies.

The intervals differ and are worth knowing before reading any figure. Binance settles every eight hours by default, at 00:00, 08:00 and 16:00 UTC, and may switch to hourly settlement in extreme volatility. Hyperliquid settles every hour, at one eighth of a rate computed over an eight-hour window. In both cases only a position still open at the settlement moment pays: closing beforehand ends the obligation, which is why a rate quoted without its schedule says less than it appears to.

A funding rate measures one thing: the distance between the perpetual price and the spot price it is anchored to. A large positive rate means the contract is trading above spot and the longs are paying to keep it there. That is a statement about two prices right now, and it is a fact. It is not a statement about what happens next.

It does not count anybody. The most common reading — a high positive rate means there are more longs than shorts — is not possible in the terms of the market. Every open contract has a long side and a short side, which is exactly why funding can flow from one to the other at all. The two sides are always the same size. What differs between them is not their number but their urgency: which side is willing to pay to hold its position, and how much.

It is a cost that compounds while nothing happens. A position held through many settlements pays each time, and the payment comes out of the collateral backing it. That matters for the arithmetic on this site: it moves a position closer to its liquidation threshold without the price having moved at all. A quiet market is not a free one for whoever is paying.

Open interest is the amount of contracts currently open and not yet closed out. It rises when positions are opened and falls when they are closed, which makes it different in kind from volume: volume counts trading, open interest counts what is left standing after the trading. A day of heavy volume that opened and closed the same positions leaves open interest exactly where it started.

The limit is the part worth carrying away, and it is the reason the cascade simulator on this site observes nothing. Open interest is one number for a whole market. It does not say at what leverage those positions are held, at what entry prices, or how close any of them sit to a maintenance threshold — and no venue publishes that distribution. So the quantity that would actually tell you what a move would trigger is precisely the quantity nobody can read.

That is why the simulator asks you to state position buckets rather than fetching them. A stated assumption you can inspect and disagree with is a more honest input than a fetched number that does not contain the information the calculation needs.

The liquidation arithmetic on this site leaves fees and funding out, and says so where it is stated rather than in a footnote. The consequence is stated too and points one way: a real position is closed slightly earlier than this arithmetic closes it, because the funding it has paid has already been taken out of the collateral the threshold is measured against.

Putting funding into the formula would mean inventing a holding period, a rate for each of its settlements, and a schedule for the venue in question — three numbers nobody supplied, to move a threshold by a small amount in a known direction. Naming the omission and its direction is the more useful answer, and it is the same choice made everywhere else here: state the assumption rather than smooth it over.