Perpetual Futures Explained: Funding Rate, Mark Price and Liquidation Price

Perpetual futures (perps for short) are futures contracts with no expiry date, and they are the most traded product in crypto. You can go long or short, use leverage, and hold for as long as you like. But there are three mechanisms you need to understand before using real money: the funding rate, the mark price and the liquidation price. This article explains each one, with numbers worked out step by step. If you’re not yet sure how crypto differs from Forex, read Crypto vs Forex first.

No expiry, so what keeps the price near spot?

A regular futures contract has a delivery date. On that date its price must equal the spot price (the price of the actual coin), which anchors the market. A perpetual has no such date, so its price can drift away from spot if too many traders want to hold one side.

What pulls it back is funding: payments that longs and shorts make to each other at regular intervals, based on how the perp price compares with the index price, an average price of the coin across several spot exchanges.

• Perp price above the index → funding rate positive → longs pay shorts
• Perp price below the index → funding rate negative → shorts pay longs

When holding the crowded side costs money and the other side gets paid, traders have a reason to close the expensive side or open the paid one, and the perp price gets pulled back toward the index.

How funding is charged, with a cost example

Funding per interval = position value × funding rate

Position value (notional) is the number of coins × the price, usually the mark price at the funding time. The point people miss is that funding is charged on the full position value, not on the margin you put up.

A common interval is every 8 hours (three times a day), but not everywhere. Some exchanges or contracts use 1 or 4 hours, and the rate changes every interval with the gap between the perp and the index. When the two are close, many exchanges sit at a base rate of around 0.01% per 8 hours. Check the real figures in the contract specification and the countdown on the trading screen.

Example: long 0.5 BTC at 60,000 USDT with 10x leverage.

1. Position value = 0.5 × 60,000 = 30,000 USDT
2. Margin = 30,000 ÷ 10 = 3,000 USDT
3. Funding +0.01% per 8-hour interval → you pay 30,000 × 0.0001 = 3 USDT per interval, or 9 USDT a day
4. If that rate held for 30 days (90 intervals) → 3 × 90 = 270 USDT, which is 9% of the margin
5. If a hot market pushes funding to 0.05% per interval → 15 USDT per interval, 1,350 USDT over 30 days, or 45% of the margin

In reality the rate is not constant, and it can turn negative (then longs get paid). But the example shows that the higher the leverage, the bigger the bite funding takes out of your margin, because it is charged on the full value. It works much like the overnight swap in Forex, covered in Swap and overnight fees. If you hold for weeks, add funding to your costs.

Mark price vs last price

• Last price is the price of the most recent trade on that exchange. A long wick lasting a second can happen when a large order sweeps a thin order book
• Mark price is a reference price the exchange calculates from the index price (prices from several exchanges), adjusted by a smoothed premium. The exact formula differs by exchange

Most exchanges use the mark price for unrealized PnL and to decide liquidation, so a wick on one exchange doesn’t wipe out positions across the market. For example, if the last price spikes down to 53,500 for a few seconds while the mark price stays at 55,000, a position with a liquidation price of 54,271 (from the example below) is not closed.

Your own Stop loss is another matter. Some exchanges let you choose whether the last price or the mark price triggers it. With last price, that same wick can fill your Stop loss. Check this option every time you set the order (Order types explained covers the basics).

Isolated vs cross margin

• Isolated keeps each position’s margin separate. If it is liquidated, you lose only the margin assigned to that position, and the rest of the account is untouched. The downside is a closer liquidation price
• Cross uses the whole account balance (in the same currency) as a shared buffer for every position. The liquidation price is further away, but if price really runs against you, you can lose more than you meant to, and one losing position can drag others down with it

Cross margin doesn’t make you safer. It just pushes the liquidation point further away by putting the whole account behind it. Your real risk is still set by position size and the Stop loss.

Working out the liquidation price (isolated long)

A position is liquidated when the margin left (margin put up + PnL) falls to the maintenance margin, the minimum needed to keep the position open. As a share of position value it is called the MMR (maintenance margin rate). For USDT-margined contracts, leaving out fees and funding, the formula is approximately:

Liquidation price (long) ≈ entry × (1 − 1/leverage) ÷ (1 − MMR)

Same example, long 0.5 BTC at 60,000 with 10x, and assume MMR = 0.5%.

1. 1 − 1/10 = 0.9
2. 60,000 × 0.9 = 54,000
3. 54,000 ÷ (1 − 0.005) = 54,000 ÷ 0.995 = 54,271.36
4. Distance from entry = (60,000 − 54,271.36) ÷ 60,000 = 9.55%

Check: at 54,271.36 the loss is 0.5 × (60,000 − 54,271.36) = 2,864.32, leaving 3,000 − 2,864.32 = 135.68 of margin. The maintenance margin is 0.5 × 54,271.36 × 0.5% = 135.68. They match exactly.

With cross margin and 5,000 USDT in the account, with this the only open position, the whole balance is the buffer. The liquidation price moves to (30,000 − 5,000) ÷ (0.5 × 0.995) ≈ 50,251, 16.25% from entry. That’s further than isolated, but if it’s hit you lose almost the full 5,000.

Limits of this formula: real exchanges also deduct a closing fee, some calculate maintenance differently, and the MMR isn’t fixed. It comes in tiers: the bigger the position, the higher the MMR and the lower the maximum leverage allowed. So the real price usually sits slightly closer than the formula says. Use the formula to understand the picture, then check the number the exchange shows before you open the trade. A short works the same way in reverse: price rising toward the liquidation price.

Leverage myths

The same rule as in Forex, explained in Leverage, margin and stop out: leverage doesn’t change how much a 1% move costs you. Position size does.

Hold a long of 0.5 BTC at 60,000. If price falls 1% to 59,400, you lose 0.5 × 600 = 300 USDT, whether you used 10x (3,000 margin) or 20x (1,500 margin). The only differences are:

• The margin tied up in the position
• The distance to liquidation: 9.55% at 10x versus 4.52% at 20x (same formula)

The right way is to start from what you’re willing to lose, then size from the Stop loss distance. For example, a 5,000 USDT account risking 1% = 50 USDT, with a Stop loss 2% below entry (60,000 → 58,800, a 1,200 gap), gives 50 ÷ 1,200 ≈ 0.0417 BTC, worth about 2,500 USDT. Only then pick a leverage that covers the margin and keeps the liquidation price far beyond the Stop loss. At 5x the margin is about 500 USDT and the liquidation price is about 48,241, far below the 58,800 Stop loss. Sizing for crypto volatility is covered in Crypto position sizing by volatility, or use the crypto calculator.

A simple rule: if the liquidation price comes before your Stop loss, the Stop loss is meaningless. The position is closed first, with a liquidation fee on top.

Insurance fund and ADL

In a very fast market, a liquidated position may close at a worse price than the point where its margin runs out. Exchanges cover that shortfall from an insurance fund. If the fund isn’t enough, some exchanges use ADL (auto-deleveraging), reducing profitable positions on the opposite side in an order the exchange sets. So even on the right side of a trade, you can be partly closed early on a chaotic day. Check the documentation of the exchange you use for the details.

Common mistakes

• Thinking funding is charged on margin. It is charged on the full position value, so high leverage lets funding eat your margin fast
• Holding for a long time without counting funding. A small rate per interval is not small after weeks or months
• Using leverage instead of position sizing. The loss per 1% depends on position size, not the multiplier
• No Stop loss because liquidation “is” the Stop loss. Liquidation takes almost all the margin plus a fee, at a point you didn’t choose
• Switching to cross to escape liquidation. The liquidation point moves away, but when it’s hit you lose more
• Not checking which price triggers your Stop loss. Last price and mark price behave differently on a day with long wicks

A perpetual is a tool, neither good nor bad in itself. If you know the funding cost, know where your liquidation price is, and size from what you’re willing to lose, you can manage the risk with the same principles as any other market.