COST Updated 2026-09-12 By Ivo Renner

Funding on equity perps: why a stock perp drifts

Without an expiry date there is nothing to force one of these contracts back towards the index it tracks, so periodic funding does that work instead. The interval and cap now differ by contract: some settle every eight hours with a ±2% cap, while newer and revised contracts settle every four hours with a ±1% cap.

Fact check: 11 September 2026. Binance changed nine existing TradFi contracts from eight-hour funding with a ±2% cap to four-hour funding with a ±1% cap on 10 September, and some contracts launched with the newer parameters. The contract panel and latest announcement govern; this article uses both schedules as explicit examples.

What funding is, and who pays whom

A perpetual contract never settles. There is no expiry date at which the contract price and the underlying price are forced to converge, which is the mechanism that keeps a normal futures contract honest. Something has to stand in for that mechanism, and on Binance that something is funding.

Funding is a periodic transfer between the two sides of the contract. It does not go to the exchange. When the funding rate is positive, longs pay shorts. When it is negative, shorts pay longs. The payment is calculated on your position's notional value, not on the margin you put up, and that distinction is the source of most of the surprise people feel when they read their first funding statement.

The logic is a feedback loop. When a contract trades persistently above the index it tracks, funding turns positive, longs get more expensive to hold and shorts get more attractive, and that asymmetry is meant to draw in enough sellers to close the gap. It is a price signal, not a rule — nothing obliges anyone to take the other side.

The 0% interest rate, and what it changes

On Binance's crypto perpetuals, the funding rate has two ingredients: a premium component measuring how far the contract trades from its index, and an interest rate component representing the cost of financing the two currencies in the pair. For most crypto perps that interest term is a small, constant, structurally positive number — the reason funding there has a mild permanent tilt in favour of shorts even when the book is balanced.

The TradFi launch and parameter announcements reviewed for this article set the funding interest rate to 0%. That removes the fixed interest input used by the formula; it does not reveal how many traders sit on either side of the book.

That zero narrows the interpretation, but not as far as a headcount. If NVDAUSDT prints a positive rate, longs pay shorts at the next settlement if the rate remains positive. The rate is derived from the contract's premium mechanics; it does not prove that there are more long traders than short traders, because positions differ in size and every open long notional has matching short notional.

Two practical consequences follow.

A funding rate near zero means the calculated transfer is near zero. It does not reveal trader headcount, intent or how concentrated positions are. Use it as the price of holding through the next settlement, then read the premium, order book and history separately.

The sign tells you the payment direction at the next settlement if it remains unchanged. A positive rate means longs pay shorts; a negative rate means shorts pay longs. Because the rate is derived from the contract’s premium mechanism, it can differ from the direction of the underlying share. By itself it does not establish trader headcount, crowding, positioning concentration or sentiment.

The contract's funding clock, and the instant that counts

The interval is a contract parameter. The original eight-hour schedule produces three settlement windows a day, commonly 00:00, 08:00 and 16:00 UTC. A four-hour schedule produces six. Binance launched HK0625USDT, BYDUSDT and HK0992USDT with four-hour funding, then changed nine existing contracts to four hours on 10 September. Read the live interval and countdown for the symbol instead of inheriting the schedule from another contract.

The detail that costs people money is this: you pay or receive funding only if you hold the position at the applicable settlement instant. Funding is not continuously pro-rated over the preceding interval. The interface countdown is the useful control because the interval and next boundary are contract-specific; closing around that boundary still leaves trading fees, spread, slippage and price risk.

A week held continuously crosses 21 settlements on an eight-hour schedule and 42 on a four-hour schedule. The realised cost still depends on the rate at each boundary and its sign, so multiplying one current print by either count is a scenario, not a forecast.

The cap, the interval exemption and later revisions

Launch specifications state an interval, a maximum funding rate and whether the generic funding-interval adjustment rule applies. Those values must be read together and treated as changeable contract parameters.

Many earlier contracts launched with a +2.00% / −2.00% cap per eight-hour settlement. Newer contracts and the nine revised on 10 September use +1.00% / −1.00% per four-hour settlement. Both combinations imply an extreme theoretical six percentage points of notional across a full day if every settlement remained at the cap; that is limit arithmetic, not a typical or predicted rate.

The TradFi specifications reviewed also exempt these contracts from the automatic mechanism described as "8.1 Adjustment of Funding Interval". That exemption does not promise an eight-hour interval forever. Binance can revise a contract's stated interval and cap by announcement, as the 10 September change demonstrates.

The practical rule is simpler: record the current rate, interval, cap and countdown from the contract panel before calculating a holding-cost scenario. Do not infer one field from another, and recheck them after a parameter announcement.

Why funding behaves differently here than on a crypto perp

The mechanics are identical to a crypto perpetual. The behaviour is not, and the reason is arbitrage.

On BTCUSDT, when funding goes persistently positive, a well-capitalised trader can buy spot bitcoin, short the perpetual against it, and collect the funding with essentially no directional exposure. That trade is available to thousands of participants, it is cheap to put on, and it is precisely what stops crypto funding staying elevated for long. The arbitrage is the enforcement mechanism; funding is just the price at which it becomes worth doing.

On AAPLUSDT, that trade is hard and often impossible. The arbitrageur would need to hold actual Apple shares, in a brokerage account, in a jurisdiction where they can also access Binance Futures, and they would need the equity leg to be tradeable at the moment the perp is dislocated — which, for a substantial part of every week, it is not. The share cannot be delivered against the contract in any case. There is no settlement date that will ever bring the two together.

So the leash is weaker. Three consequences follow from that constraint:

• Premiums and discounts can persist longer. A gap between the perp and the index can survive while the equity hedge is unavailable. Funding responds at the contract's scheduled interval, which may be four or eight hours.
• Funding is doing more of the work, with less to work with. It is the primary tether rather than a secondary one, and it is capped and infrequent. That is a weak instrument for a job this size.
• Closed-market events can move the premium input. After-hours earnings, a weekend headline or an overnight macro release can move the perpetual while the equity hedge is unavailable. The resulting funding rate records the formula’s transfer for that premium; it does not reveal how many traders are on either side or what they believe.

The useful inversion: a large funding rate can coincide with a closed underlying market, but it remains a contract-level transfer calculated from the premium mechanism, not a measure of conviction. Read it alongside the session calendar.

What the same rate costs at eight hours and four

Abstract percentages are easy to wave through. Here is the arithmetic on a position size a real person might take.

Assume a 5,000 USDT notional long position, held continuously for 30 days, and a hypothetical positive funding rate fixed at 0.03% per settlement. This is a comparison input, not a claim about a typical or current rate. A positive rate means the long pays in this example.

Two things about that table should stop you.

The first is the interval effect. The same per-settlement rate costs twice as much on a four-hour schedule because it lands twice as often. Annualising a current print is only a comparison device: actual rates change at each boundary.

The second is what the eight-hour example's 135 USDT does to margin rather than notional; the four-hour example would double each figure below if the same rate persisted:

• At 5×, the 5,000 USDT position needs 1,000 USDT of initial margin. The month's funding is 13.5% of your margin.
• At 10×, margin is 500 USDT. Funding is 27% of it.
• At 25×, margin is 200 USDT. Funding is 67.5% of your margin — you have paid away two-thirds of the capital backing the trade before the stock has moved a cent.

This is the mechanism by which leveraged positions can lose margin without a move in the underlying. Funding does not liquidate you directly; repeated payments can erode the buffer between the position and its liquidation price. The interaction with leverage tiers and liquidation is the part people miss.

How to read a funding rate before you enter

The rate shown in the trading interface is the predicted rate for the next settlement, and it moves. It is not a quote you have locked in; it is an estimate based on where the premium sits right now.

A position opened at 15:58 on a predicted 0.01% can settle at something else at 16:00.

1. Check the history, not the snapshot. Binance publishes both the live rate and the historical series on a public page — no account required. One reading tells you nothing; a month of readings tells you whether the contract carries a structural tilt, and tilts are what kill multi-week positions. The route is below.
2. Multiply by the applicable boundaries. A continuous week crosses 21 settlements at eight hours or 42 at four hours. Use the calculator as a scenario, then compare it with the changing historical rates rather than assuming one print persists.
3. Check which side pays. A persistently positive rate is a cost if you are long and a receipt if you are short; a negative rate reverses that direction. Taking a side to collect funding still leaves the position exposed to price movement. The rate alone does not show that the contract is crowded.
4. Put the live countdown against your calendar. Do not assume every symbol uses the three legacy UTC times. Timing a close around funding also incurs trading costs and price risk, so treat the countdown as cost information rather than a free-money tactic.
5. Confirm the parameters in the interface. The contract panel governs your position. On 10 September 2026 Binance revised nine existing TradFi symbols to four-hour intervals and ±1% caps; that is why a launch specification alone is not enough.

What we cannot tell you: what funding has actually averaged on any specific contract, because that number changes daily and any figure printed here would be stale before you read it. What we can tell you is exactly where to read it yourself, which is the next section.

Reading the actual rate, without an account

Binance publishes funding data on a public page. You do not need to be logged in, and you do not need to have funded anything. This is the single most useful habit on this page, because every number in the arithmetic above is a worked illustration and the one on the screen is real.

1. Open the live table. Binance Futures → Data → Real-Time Funding Rate, or go directly to the URL in the sources below. Select the USDⓈ-M tab and search your symbol.
2. Read your side of it. A positive rate means longs pay shorts. Note the figure and the countdown in Time to Next Funding — opening minutes before a settlement you are on the wrong side of is an avoidable cost.
3. Switch to Funding Rate History. One live reading is noise. Pull roughly thirty days and look at the median and sign stability. A persistently positive history documents repeated transfers from longs to shorts; it does not reveal trader headcount.
4. Multiply before you commit. Take the median, not the current print, into the funding cost projector with your intended notional and holding period. That is the number that decides whether a multi-week thesis survives its own carry.

One caveat we have not resolved: we verified this page is public and that it lists USDⓈ-M perpetuals, but we have not confirmed how completely it covers the TradFi symbols specifically. If your contract is missing from the list, the contract’s own trading screen carries the same figures. Tell us if you find a gap.

The short version

Funding on a Binance TradFi perpetual is a scheduled transfer between longs and shorts. A 0% interest input does not prove who outnumbers whom. The interval and cap are contract parameters: reviewed contracts include eight-hour/±2% and four-hour/±1% combinations, and Binance can revise them by announcement. You pay or receive only when you hold through the applicable boundary.

In the explicit 0.03%-per-settlement scenario above, 5,000 USDT held for 30 days costs 135 USDT at eight hours or 270 USDT at four hours. Those are sensitivity figures, not expected returns or advice to shorten a position. Use the symbol's current parameters and your own intended holding period.