what negative funding means and how much it actually pays

negative funding means the short side pays the long side at the funding settlement. if you are long and still hold the position across that timestamp, you receive funding; if you are short, you pay it. positive funding reverses the flow: longs pay and shorts receive. the sign describes the direction of the transfer, not whether either side has positive PnL on the position.

the payment is calculated from position notional, not the margin you posted. a displayed −0.05% input on $10,000 notional therefore means $10,000 × 0.0005 = $5 transferred from the short side to the long side for that interval. the minus sign does not mean the receiver loses $5; it identifies shorts as the paying side under the usual perp convention.

a negative print is also a crowding signal: short positioning is carrying the payment. it is not free money because the rate can change by the next settlement, price can move against the receiving long, and taker fees can consume the credits. funding PnL and position PnL remain separate lines in the economics of the carry.

both Binance and Bybit use an 8-hour funding interval for the products covered here. that gives 3 possible settlements per day, 21 across a week, and 90 across a 30-day month. at an illustrative −0.05% input, multiply notional by 0.0005 for one interval, then multiply that result by the number of settlements actually crossed.

the table assumes the rate remains exactly −0.05% at every settlement and that the long remains open across each timestamp. it is scenario arithmetic, not a forecast or an observed historical rate. actual funding can change from one interval to the next, so a monthly projection must use each applicable rate rather than extending one print automatically.

a long receives when the funding rate is negative and pays when it is positive. a short has the equal directional case: it pays at a negative rate and receives at a positive rate. on $10,000 notional, the absolute transfer at either +0.05% or −0.05% is $5 because $10,000 × 0.0005 = $5; only the payer changes.

the symmetry applies to funding, not to total trade PnL. a long collecting $5 can still lose more from a falling mark, while a short paying $5 can remain profitable if its position PnL exceeds the charge. the displayed rate is also not locked for the holding period. each crossed settlement uses the rate applicable to that interval.

funding is charged on notional rather than your posted margin. keeping notional fixed at $10,000 therefore keeps one −0.05% credit fixed at $5, whether the position uses 5x, 20x or 100x. leverage changes the capital supporting that notional: $10,000 ÷ 5 = $2,000, $10,000 ÷ 20 = $500, and $10,000 ÷ 100 = $100.

relative to those initial-margin figures, one $5 credit is 0.25% at 5x, 1% at 20x and 5% at 100x. if the same −0.05% input persisted for all 90 monthly settlements, $450 would equal 22.5%, 90% or 450% of those respective margin amounts. that comparison does not include adverse PnL, fees, funding-rate changes or liquidation.

no. funding is settled only when your position remains open across the applicable funding timestamp. opening after one timestamp and closing before the next means no funding transfer for that interval, even if the displayed rate was negative while you held. conversely, crossing one timestamp can create one payment even when the position was not held for the full 8-hour interval.

this makes holding time discontinuous rather than a smooth hourly accrual for settlement purposes. on $10,000 notional at the illustrative −0.05% input, crossing one timestamp produces $5; crossing none produces $0. holding across 3 timestamps produces $15 if the same rate applies each time. the exact result must use each venue settlement record.

an order intended to exit before funding still carries execution risk. a resting order may not fill, and a taker exit introduces its own fee and fill price. trigger type for stop, take-profit and conditional orders is selectable and varies by product and interface, so it must be checked on the account rather than assumed from a venue default.

yes. Binance’s supplied taker fee is 0.05% per fill, or 0.1% for an assumed taker entry and taker exit. Bybit’s supplied figures are 0.055% per fill and 0.11% for the same round trip. at −0.05%, 2 unchanged funding credits gross-match the Binance 0.1% round-trip fee; at −0.055%, 2 credits gross-match Bybit’s 0.11%. funding can change before either second credit arrives.

leverage magnifies fee drag against margin in the same way it magnifies funding. at 20x, the supplied round-trip taker cost equals 2% of margin on Binance and 2.2% on Bybit. at 100x, those figures become 10% and 11%. these comparisons exclude slippage, funding paid in another interval and any position loss between entry and exit.

liq room can be thinner than the funding being pursued. for the Binance BTC example with a $97,400 entry and $10,000 notional, the estimated 20x long liq is $92,919.60, leaving 4.6% or $4,480.40. at 100x, the estimate is $96,815.60, only 0.6% or $584.40 away. estimate — exchanges compute exact liq from live margin; always double-check on your venue.

start with funding cash flow: notional × absolute funding rate × crossed settlements. then keep position PnL, entry and exit fees, likely taker fill, and liq room separate. for a $10,000 long at −0.05%, one settlement contributes $5 and 90 unchanged settlements contribute $450, but that funding line alone says nothing about what happens to the mark.

the liqsiren funding-cost calculator takes size, rate and holding time and returns funding paid or received per interval, day, week and month. the useful output is the full carrying-cost line rather than the rate in isolation. a persistently negative rate can identify crowded shorts, but a funding edge that does not survive fees and the liquidation risk of the carrying position is not an edge.

• check the funding sign and which side pays at the next settlement.
• use position notional, not posted margin, in the funding multiplication.
• count only the 8-hour timestamps the position actually crosses.
• compare total credits with the 0.1% Binance or 0.11% Bybit assumed taker round trip where those supplied fees apply.
• stress the position against its MMR bracket and estimated liq, especially when leverage compresses room to 0.6% or less in the worked BTC case.