SeriesTopical Lesson
Lesson2 of 3
TopicFunding Rate
Harmonic Academy
← Back to Academy
Topical Lesson 2 · Harmonic Academy
The Funding Rate
What it is, how it is calculated, and what it is telling you
Understand Apply Case Study
Most traders look at the funding rate and see one number. Positive means bullish, negative means bearish, job done. That reading misses two components, a structural bias, and the difference between a leveraged rally and a spot-driven one. This lesson shows you what the number is actually made of and what it is actually telling you.
From the Harmonic Weekly · June 15
The Trade of the Week Is Already in Motion →
How the Funding Rate Is Actually Calculated
A perpetual futures contract never expires. Because there is no settlement date to force the contract price back toward spot, exchanges needed a different mechanism to keep the two prices anchored. That mechanism is the funding rate: a periodic payment that flows between holders of long and short positions. When longs are paying shorts, longs are expensive relative to spot. When shorts are paying longs, shorts are. That payment pressure is what pulls the two prices back together over time.
The rate is not simply the gap between the perpetual price and spot. On every major exchange it is a formula with two distinct components, a premium index and an interest rate, and understanding each one separately is what makes the number legible.
The premium index measures the gap between where the perpetual contract is trading and where the underlying spot index is. But it does not use the last traded price or the simple midpoint of the order book. Binance, Bybit, and OKX all calculate it using impact prices, the average fill price to execute a standardized notional amount on each side of the order book.

Premium Index = [Max(0, Impact Bid Price − Index Price) − Max(0, Index Price − Impact Ask Price)] ÷ Index Price

Impact Bid Price = average fill price to execute a standard notional on the bid side

Impact Ask Price = average fill price to execute a standard notional on the ask side

Index Price = weighted average of BTC spot prices across multiple reference exchanges

Using impact prices instead of the quoted midpoint matters for two reasons. First, it reflects what you would actually pay or receive if you tried to trade a meaningful size: not just the best bid and offer. Second, it makes the index resistant to manipulation in thin markets where someone could temporarily move the quoted price with a small order.
The premium index is then averaged across the entire funding period using a time-weighted moving average: every minute for the last 8 hours, for example. The final premium index is not a snapshot of the last minute's spread; it is a rolling summary of where the contract has been trading relative to spot throughout the whole period.
This is the part most explanations skip entirely, and it is the reason the funding rate has a structural positive bias.
The interest rate component reflects the theoretical cost of capital, the difference in borrowing rates between holding dollars versus holding crypto. Binance and Bybit both set this at a fixed 0.01% per 8-hour period for BTC/USD contracts. It is small and it rarely changes, but it is always there. The practical effect is that even in a perfectly neutral market, where the perp is trading exactly at spot and the premium index is zero, the formula still produces a small positive funding rate equal to the interest rate. The rate gravitates toward 0.01% per period, not zero.

Funding Rate = Premium Index + clamp(Interest Rate − Premium Index, 0.05%, −0.05%)

The clamp limits how much the interest rate can push the final rate away from the premium index.

When the premium index is large, it dominates and the interest rate becomes negligible.

When the premium index is near zero, the interest rate adds ~0.01% positive bias to the result.

Deribit's formula does not include a fixed interest rate component. It uses a premium-based approach with a deadband: if the mark price is within 0.025% of the index price in either direction, the funding rate is set to exactly zero. Outside that band, the rate equals the premium minus the 0.025% buffer. A truly neutral market on Deribit produces a funding rate of zero: there is no structural positive bias.
Most BTC perpetuals settle every 8 hours, but not all. Binance moved certain contracts to 4-hour settlement in 2023. OKX offers contracts settling every 1, 2, 4, or 8 hours. This matters directly when annualizing the rate.

Annualized funding = per-period rate × (24 ÷ settlement interval in hours) × 365

8-hour settlement: rate × 3 × 365   |   4-hour: rate × 6 × 365   |   1-hour: rate × 24 × 365

0.01% per 8h = 10.95% annualized   |   0.03% per 8h = 32.85%   |   0.05% per 8h = 54.75%

Always check the settlement interval on the specific contract you are trading before annualizing. The same per-period rate on a 4-hour contract costs twice as much annually as on an 8-hour contract.
The funding rate is not a directional signal: high positive funding does not mean sell and negative funding does not mean buy. It describes the structural condition of the market your trade is entering. A long entry into a high-funding environment carries cascade risk because the market is already heavily leveraged in the same direction. A long entry into a suppressed-funding rally carries less of that risk because the move is being driven by spot buyers, not by longs who can be liquidated. That context changes your sizing and your stop placement even when it does not change your direction.
What Each Reading Means and What to Do Next
Regime 1 High positive Above 0.03% per 8h  /  >33% annualized
Leveraged long crowding
Perp has been trading persistently above spot. The premium index is driving the rate well above the interest rate floor. The market is carrying a structurally fragile long position.
→ Carry trade is open. Collect funding from longs.
Regime 2 Near baseline ~0.01% per 8h  /  ~11% annualized
Balanced market
Premium index near zero. The interest rate component dominates. This is the equilibrium the formula gravitates toward in neutral conditions: no meaningful crowding either way.
→ Carry is thin. Look to other yield sources.
Regime 3 Suppressed 0–0.004% per 8h  /  0–4% annualized (linear contracts)
Spot-driven buying (ETF inversion)
Funding is below the interest rate floor even as price rises. Buyers are entering at the spot level, ETFs, direct purchases: not through leveraged perp demand. Less leverage overhang than a normal rally.
→ Borrow environment. Low-cost dollar liquidity.
Regime 4 Negative Below 0.00% per 8h
Short crowding
Perp trading below spot. Shorts are paying longs. Historically short-lived: the formula's positive bias and arbitrage capital both push back toward positive territory quickly.
→ Contrarian long signal. Small defined-risk long only.
Reading the Rate Before Every Trade
Before entering any position in the perpetual market, run through this sequence. It takes under two minutes and tells you which regime you are in and what that means for the position you are about to put on.
01 Go to CoinGlass or the funding rate section of your exchange. Find the current BTC-USD perpetual funding rate. Note the per-period rate and check the settlement interval, 8-hour, 4-hour, or otherwise. Do not assume it is 8-hour.
02 Annualize it using the formula from the Understand section: per-period rate multiplied by (24 divided by the settlement interval in hours) multiplied by 365. A 0.03% rate on an 8-hour contract is 32.85% annualized. The same rate on a 4-hour contract is 65.7% annualized, not the same number.
03 Match the annualized rate to the four regime cards above. That tells you the structural condition of the market and the appropriate response, carry trade, borrow structure, or contrarian long.
04 Check whether you are on Deribit or a Binance/Bybit/OKX venue. On Deribit, a rate of exactly zero means the premium index is flat with no bias. On other exchanges, a rate near 0.01% per period means the interest rate floor is dominating and you are in Regime 2, a balanced market. A rate below that floor means the premium index has turned slightly negative and you are in Regime 3, a borrow environment. The same number means different things depending on the exchange and the direction of the premium index.
05 Log the regime alongside your trade entry. When you review the trade later, the funding regime at entry is part of the data, it tells you whether the structural conditions supported the position or worked against it.
Two Regimes, Same Asset: 2021 vs Mid-2026
Q1 2021: High Positive Funding, Regime 1
BTC: $29,000 → $58,000 Q1  /  $65,000 April peak
Through Q1 2021, as BTC moved from $29,000 to $58,000 through Q1 and on to $65,000 by April, the premium index ran persistently elevated. The perp was consistently trading above the spot index, retail traders were piling into leveraged long positions and paying a premium to do it. Funding rates reflected this: 0.05 to 0.10 percent per 8-hour settlement for extended periods, or 54 to 109 percent annualized. At those levels the interest rate component of the formula, 0.01% per period, was completely irrelevant. The premium index was dominating entirely.
This was Regime 1 sustained for weeks. The signal was clear: the market was carrying an enormous leveraged long position, and every leveraged long in the market was paying above 50 percent per year to hold it. That crowding is what makes a cascade possible. When the China mining ban announcement hit in May 2021, the premium index collapsed as long positions were liquidated en masse, the funding rate went briefly negative as sentiment flipped, and over $10 billion in positions were cleared in a week. The funding rate had been describing the structural fragility for months before the catalyst arrived.
Mid-2026: Suppressed Funding, Regime 3
BTC: $59,375 low → $66,500  /  June 5–17, 2026
Following the June 2026 liquidation cascade, the derivatives market reset. Sustained ETF outflows put consistent selling pressure on the market over several weeks, driving BTC through successive support levels. As price broke lower, leveraged long positions hit their liquidation thresholds. The cascade cleared an estimated $5.8 billion in long positions, driving BTC to a low of $59,375 on June 5. The premium index compressed toward zero as the perp stopped trading at a meaningful premium to spot. Funding rates on linear USDT contracts fell well below the interest rate floor. The 30-day actual average across Binance, OKX, and Bybit ran 2 to 4 percent annualized, with 1-day readings briefly going negative as shorts paid longs.
As BTC recovered through mid-June driven by the Iran deal resolution and returning ETF inflows, price was rising but the premium index was not expanding with it. This is Regime 3: spot-driven buying. ETF inflows push the spot index price up directly without a corresponding surge in perp demand that would push the contract above spot and lift the premium index. The formula had nothing significant to charge because there was no structural imbalance building.
The same asset recovering from a sharp sell-off looked completely different in the funding data across the two periods. In 2021 the premium index was elevated throughout the rally and a cascade was being loaded. In mid-2026 the premium index stayed flat as price recovered from the cascade low. Knowing which environment you are in changes how you size the trade, where you put the stop, and what you do next.
Key Takeaway
The Funding Rate Is a Premium Index Signal With a Built-In Floor
High positive rates mean leveraged long crowding, the market is loading a structure that can cascade. Suppressed rates during a rally mean spot-driven buying, structurally cleaner, less leverage overhang. Negative rates mean short crowding: historically brief. The same asset in two different funding environments tells a completely different structural story. You can now read that signal on any major exchange and know what it actually means.
Up next · Topical Lesson 3
Now you know how to read the funding rate. Learn what to do with it.
Run carry when funding is high · Borrow cheap when funding is low · Compare perp to expiry basis · Calculate BNOC to pick the right leg
Continue →