What Is the Funding Rate in Perpetual Futures — and What Does Holding Cost?

Why perpetuals pay funding at all

A perpetual future never expires, so nothing forces its price to converge with the spot market. Funding is the mechanism that does it instead: at fixed intervals — commonly every eight hours — one side of the market pays the other, longs paying shorts when the perp trades above spot and shorts paying longs when it trades below. The exchange is not charging you a fee; the other side of the market is, and which side you are on decides whether you pay or receive.

Small rate, real money

The advertised number looks harmless because it is quoted per interval. A 0.01% rate every eight hours is three payments a day — roughly 0.03% of position notional daily, about 0.9% a month, near 11% annualized. On a $10,000 position that is about $3 a day and $90 a month.

Leverage is what turns that from a rounding error into a position cost. The payment is charged on notional, but your capital is the margin: at 10x, that ~0.9% of notional per month is roughly 9% of your margin per month. In stretched markets funding has printed several times the 0.01% baseline for days at a time — and it settles out of the same margin that is holding your liquidation line away.

When funding decides the trade

For a position closed within hours, funding is noise. It becomes the decision on three kinds of trade:

• Multi-day holds at leverage — a week of elevated funding can cost more than the stop distance you were protecting.
• Range strategies with thin expected edge — an edge of 0.5% per trade does not survive 0.9% per month of holding cost.
• Crowded direction trades — heavily positive funding means the crowd is long and paying for the privilege; the cost itself is telling you how consensus the position is.

Put it in the plan, not the postmortem

The honest way to hold a perpetual is to price the hold before entry: expected days held × intervals per day × an assumed rate × notional, subtracted from the trade's expected value, and checked against the margin buffer. A scenario is not a forecast — the point is that a trade you would not take at an assumed funding cost is a trade whose thesis depends on the market staying cheap to hold.

Common questions