Funding Rates: What They Price and When They Mislead

A perpetual futures contract never expires, which removes the mechanism that normally pulls a derivative toward the price of the thing it tracks. Funding replaces it: at regular intervals, one side of the market pays the other, and the direction and size of that payment are set by how far the contract has strayed from the underlying.

The tether

If the contract trades above spot, longs pay shorts. Holding a long becomes continuously expensive and holding a short is paid, which encourages selling the contract and buying the underlying until the gap closes.

If it trades below, the flow reverses. In both cases the payment moves between traders rather than to the venue — this is a redistribution, not a fee, though venues charge fees separately.

The rate is calculated from the observed premium, usually with an interest component reflecting the cost of holding the underlying versus the contract. Both parts are published in the contract specification, and both differ between venues in ways that matter for anyone comparing.

What a persistent rate actually says

Sustained positive funding means the contract has been trading above spot for a while, which means demand for leveraged long exposure exceeds the supply of people willing to take the other side at a lower price.

That is a real fact about positioning. What it is not is a forecast. Funding has been strongly positive through sustained rallies and strongly positive immediately before sharp reversals, and the rate itself does not distinguish between those.

The more defensible reading is as a measure of crowding and of cost. Persistent high funding tells you that one side is paying a lot to stay in the trade, which means the position is expensive to hold and populated by people with a continuing outflow. Crowded, expensive positions unwind faster when they start to hurt — not because funding predicted anything, but because the participants were already paying to be there.

The compounding cost people ignore

Funding is charged at intervals, often several times a day, and the quoted rate is usually per interval rather than annualised. A number that looks trivial per period is not trivial per year, and traders holding directional positions for weeks routinely pay more in funding than in trading fees without ever adding it up.

For a leveraged position, the funding cost is charged on notional rather than on margin, which means it scales with leverage. At high leverage, funding alone can consume the margin over a sufficiently long hold even if the price never moves against you.

This makes funding the dominant cost of a long-held perpetual position and the one least often modelled before entering.

Where the signal breaks down

Basis trading distorts it. A trader who is long spot and short the perpetual is collecting funding while carrying no directional exposure. When that trade is popular, funding reflects the size of the basis trade rather than directional sentiment, and reading it as bullishness is simply wrong.

Venue differences distort it. Funding is venue-specific, and a rate on a thin venue with concentrated positioning says nothing about the wider market. Comparing across venues requires knowing their formulas differ.

And extreme readings mean less than they appear to. During a dislocation, funding can spike because arbitrage capital is unavailable rather than because conviction is extreme — the tether is straining, not signalling.

Using it properly

• Annualise the rate before deciding whether it is expensive. Per-interval figures systematically understate the cost of a long hold.
• Model funding on notional rather than margin, so the cost of leverage is visible.
• Read persistent funding as crowding and carry cost rather than as a directional forecast.
• Check whether a large basis trade explains the rate before interpreting it as sentiment.
• Compare rates across venues only after reading both formulas, since the inputs differ.