Most traders check the funding rate, see a bigger number than yesterday, and assume the exchange changed something. It didn't. Funding is a formula with two inputs, and once you can read it, a "sudden" spike stops being a surprise and becomes a signal you can act on.
Here is a real case from 6 August, with the actual numbers.
## The number that looked wrong
A perpetual contract that had been settling at 0.0050% every four hours suddenly printed 0.06199%. That is 12.4 times higher. Nothing was announced. Nothing was changed.
Pulling the contract's own parameters shows why 0.0050% had been so sticky: settlement interval four hours, funding cap ±2%, base interest rate 0.01% per eight hours. Convert that base rate to a four-hour window and you get exactly 0.0050%.
That number was not a coincidence. It was a floor.
## The formula, and the clamp that hides inside it
Binance calculates funding as:
funding = Premium + clamp(InterestRate − Premium, ±0.05%)
The premium index P measures how far the perpetual's mark price sits above or below the underlying spot index. The clamp is the part people miss.
When P is small, the clamp term absorbs the difference completely and funding lands on the interest rate — 0.0050%. This holds for any P below roughly 0.055%. That is why a quiet market produces the identical number, settlement after settlement, for days.
Once P pushes past that threshold, the clamp saturates and the formula collapses to:
funding ≈ P − 0.05%
From that point on, funding tracks the premium directly. Working backwards from 0.06199% gives P ≈ 0.112%. Checking the live prices confirmed it: mark 0.011504 against index 0.011467, an instantaneous premium of 0.32%.
So the rate did not jump because of a rule change. It jumped because the contract crossed out of the flat zone into the linear zone.
## What the history actually showed
The previous thirty settlements averaged 0.0161%, and most of them were the 0.0050% floor exactly. The break came at 20:00 on 5 August, printing 0.0803%. Then 0.0749%, 0.0231%, 0.0290%, 0.0477%.
Read that sequence properly. It is not noise. It is the premium repeatedly clearing the threshold, which means perpetual buyers kept paying above spot rather than waiting for spot to catch up.
The single most useful habit here: when funding leaves the floor, stop treating it as a fee and start treating it as positioning data.
## Confirming it with open interest, not vibes
A funding spike alone is ambiguous. It could be longs piling in, or it could be shorts capitulating. Three cross-checks separate the two.
Open interest over the same 24 hours rose 12.5%, from $3.78M to $4.25M. Rising price with rising open interest means new money opening longs — not shorts closing. Had shorts been covering, open interest would have fallen while price rose.
The global long/short account ratio sat at 2.37 to 2.57. Roughly 71% of accounts were positioned long.
The top-trader position ratio was 1.44, about 59% long.
That gap is the interesting part. Retail was far more one-sided than the larger accounts. Crowded retail positioning with a premium that will not close is a recognisable late-stage pattern, not an early one.
## Translating funding into an actual cost
Percentages per four hours feel harmless. Annualise them and they stop feeling that way.
0.062% every four hours is 0.372% per day, because there are six settlements. That is 11.2% per month, or roughly 136% annualised.
Now apply leverage, since funding is charged on position notional rather than on your margin. At 3x, a month of holding that long costs about 33.5% of your capital in funding alone. At 5x it is 55.8%.
Sit with that for a second. At those levels, a market that goes perfectly sideways still drains the position. You are not waiting for free — you are paying rent, and the rent resets every four hours.
## How to use this before you enter
Four checks, none of which take more than a minute.
Check the settlement interval first. Not every contract is on eight hours; this one was on four, which means the daily cost is six settlements rather than three. People carry over assumptions from BTC perps and get the daily figure wrong by double.
Compare current funding against the contract's floor. Convert the base interest rate to the settlement window. If funding equals that floor, the premium is quiet. If it is a multiple of the floor, positioning is stretched.
Check open interest direction alongside price. Price up with open interest up means fresh longs. Price up with open interest down means shorts covering, which is a different trade with a different ending.
Compare the retail long/short ratio to the top-trader ratio. When retail is materially more long than the large accounts, you are the crowd, and the crowd pays the funding.
## What this does not tell you
Funding is a positioning indicator, not a direction indicator. High funding means longs are crowded and paying; it does not mean the price must fall, and plenty of trends run for weeks with elevated funding. Traders who short purely because funding is high get run over regularly.
What it does tell you reliably is your carrying cost, and that number is not an estimate — it is arithmetic. A position that needs to move 11% in a month just to break even on funding is a different proposition from one that needs to move 1%, even if the chart setup looks identical.
Know which one you are holding before you size it. None of this removes risk; it just stops you from paying a cost you never calculated.
Register: https://www.bsmkweb.cc/zh-CN/join?ref=BNAPP
#Binance #FundingRate #Perpetuals #TradingEducation #DerivativesTrading