Funding Rates Explained: The Real Cost of Perpetual Futures
A funding rate is not a fee paid to the exchange. It is a payment between the holders of the long side and the short side, settled at fixed intervals for as long as your position stays open. That is precisely why a perpetual trade whose price moved slightly in your favour can still end up negative.
What is a funding rate?
A recurring balancing payment between the two sides of a perpetual contract. When the rate is positive, holders of long positions pay holders of short positions. When it is negative, the payment runs the other way.
The exchange acts as the settlement venue rather than the recipient. That distinguishes funding fundamentally from trading fees, which you pay regardless of your direction. The size depends on how far the contract price sits from its reference price, and it can change from one interval to the next.
Each exchange sets its own calculation and its own intervals. Many work in eight hour intervals, meaning three payments a day. What binds is always your exchange's contract specification, and that page is worth reading before you hold a position overnight.
Why does funding exist at all?
Because a perpetual contract has no expiry date. In a conventional future, expiry is what forces contract price and spot price to meet at the end. Without expiry that mechanism is missing.
The funding payment replaces it. When the contract price runs above the reference price, holding the long side becomes more expensive and holding the short side more rewarding, which narrows the gap. When it runs below, the effect reverses.
The practical consequence: the rate is not an arbitrary figure but a reflection of positioning. It tells you which side is currently willing to pay for its exposure. What it does not tell you is the direction of the next price move.
When is it charged and when is it not?
Only at each settlement timestamp, and there in full for the interval. This is where most misunderstandings come from.
Two things follow. Someone who closes a position shortly before the timestamp pays nothing for that interval, even having been exposed for almost all of it. And someone who opens a minute before the timestamp and holds through it pays the full amount for an interval they barely used. Nothing is prorated.
So in practice: your exchange's settlement times belong in your trading plan whenever you hold positions across hours. For an intraday trader who closes beforehand, funding is irrelevant. For everything beyond that, it is a cost block that grows with holding time.
What does this look like as arithmetic?
An example with freely chosen numbers, meant only to show the order of magnitude. Actual rates fluctuate and are published by your exchange.
Suppose you hold a position with a notional value of 10,000 euros, the rate is 0.01 percent against you in each interval, and the exchange settles three times a day. You then pay 1 euro per interval, 3 euros a day, and 30 euros over ten days.
What matters is the reference point. Those 30 euros are 0.3 percent of the notional, but if you hold the position on 2,000 euros of margin they are 1.5 percent of the capital you committed. Costs attach to notional size while your risk and your result attach to margin, and that gap is why funding bites so much harder on leveraged positions.
If the price moves 0.2 percent in your favour over that period, the trade was positive in price terms and negative in result. How to keep notional and margin cleanly apart is covered under trading crypto volatility.
Why does this turn a good trade negative?
Because the cost grows with time while your price gain does not. An idea that works over a day can be consumed by its own holding costs across two weeks without anything being wrong with the idea itself.
Two effects amplify that. First, the payment falls due whether or not the price moves at all, which makes a sideways phase not free but the most expensive state a perpetual position can be in. Second, rates are often highest when one side is heavily crowded, meaning exactly in the phases where many people hold the same position.
A sober planning consequence follows: with perpetuals, expected holding time is part of the trade idea. An idea without a time horizon cannot be tested against its costs.
How do you estimate holding costs before entering?
With a back of the envelope calculation that takes under thirty seconds and still answers the most important question: how far does the price have to move for this trade to earn its holding costs at all?
The calculation has three inputs: the current rate, the number of settlements until your planned exit, and your notional size. Their product is your expected holding cost. Set that against your planned price target and you see immediately whether the idea has room or whether it was already tight before you entered.
Two caveats belong with it. First, the rate is not guaranteed and changes from interval to interval, so the calculation stays an estimate rather than a commitment. Second, your notional moves with the price, which means the actual payment will differ from the estimate.
The order of magnitude is still solid enough to decide on. Running the calculation once for your typical holding period gives you a feel for when extending a position stops paying, after which you no longer need to repeat it before every trade.
How do you record funding costs in your journal?
As a separate item, kept apart from the price result. That is the central practical recommendation here, and the reason is simple: mixing the payments into your result makes it permanently impossible to tell whether an idea was poor or merely held too long.
Three fields are enough:
- Price result from entry and exit, with no ancillary costs in it.
- Trading fees for opening and closing.
- Sum of funding payments across the full holding period, with a sign, because it can also fall in your favour.
The sign matters. Standing on the receiving side means negative holding costs, and that case disappears entirely once everything is collapsed into a single result figure. The groundwork for this separation sits in the crypto trading journal guide.
If the data arrives through an exchange connection, check deliberately once where the payments land. Some interfaces deliver them as their own event, others net them off directly. How such a connection works in practice is covered under connecting the Bitget API.
What changes in your analysis?
Three figures gain a second reading once funding costs sit in their own column.
Result per trade gets calculated twice, once from pure price movement and once after all costs. The distance between the two is your cost ratio, and on long held perpetuals it is often larger than expected.
Holding time turns from a description into a lever. Sort your trades by holding time and compare the net result, because only then do you see where holding stops paying for you.
Splitting by direction is worth adding, since the payment is direction dependent. If your long trades systematically underperform your short trades despite similar price moves, you have probably found the reason in this column. A properly kept crypto trading journal produces that analysis without extra work.
Where does this topic end?
At the question of which instrument suits your trade idea in the first place. Funding is an argument for expressing short holding periods in perpetuals and solving long ones differently, but it is only one argument among several.
What else differs between spot and derivatives, particularly for journaling, is covered under crypto futures vs spot. This article covers the cost mechanics, that one the differences in how you record trades.
A closing note: this text explains a cost mechanism and is not investment advice. Whether an instrument carrying ongoing holding costs fits your strategy depends on your trade idea and your risk, not on the level of any single rate.