Funding and Carrying Costs

The short answer

Holding a leveraged position is renting exposure. The rent is funding on a perpetual, or roll cost on a dated contract, plus fees on each transaction.

None of it depends on being right. It accrues while you hold, in both directions, and it is the reason a correct directional view can still finish behind.

How funding works

A perpetual has no expiry, so it needs a mechanism to stay near spot. That mechanism is a periodic payment between the two sides — typically every eight hours, three times a day.

• Contract above spot → funding positive → longs pay shorts.
• Contract below spot → funding negative → shorts pay longs.

The effect is economic pressure toward spot: being on the crowded side becomes progressively more expensive, which incentivises the arbitrage that closes the gap.

The arithmetic that surprises people

Funding is charged on notional, not on your capital. That distinction is where the surprise lives.

Illustrative and invented for teaching.

You have $1,000 of capital and open a 10× long: $10,000 notional. Funding is 0.01% per 8-hour period.

• Per period: $10,000 × 0.0001 = $1.00
• Per day: $3.00
• Per 30 days: $90.00

That is 9% of your capital in a month, from funding alone, with price unchanged. Annualised at that rate it is roughly 110% of capital — because the 11% annualised figure applies to notional, and your notional is ten times your capital.

Now consider that funding rates in active markets frequently run well above 0.01% for extended periods. The drag becomes the dominant term.

This is the structural reason perpetuals suit short holding periods and suit long-term exposure badly. For long-term exposure, spot has no carry — you can hold it indefinitely at zero ongoing cost. See spot vs derivatives.

Roll cost on dated contracts

Dated futures do not charge funding. They expire, so continuous exposure requires rolling into the next contract.

If the next contract trades at a premium, you sell the cheaper expiring one and buy the more expensive next one. That difference is your roll cost, and repeated across many rolls it produces a drag structurally similar to funding.

The mechanism differs; the conclusion does not. Time costs money in leveraged products.

Include it before you hold

Before extending a position's holding period, estimate:

expected funding cost = notional × rate per period × periods held

Then compare that with the move you expect. If a week of funding costs a meaningful fraction of your target, the trade needs to work quickly or not at all — which is itself useful information about whether the instrument fits the idea.

Put the estimate in the "other costs" field of the fee calculator alongside entry and exit fees, so the break-even move you compute is the real one.

Funding as a positioning signal

Extreme funding tells you positioning is crowded on one side. That is genuinely informative: a crowded, leveraged book is a book with a lot of liquidation levels stacked in one direction, which makes a move against it faster and larger than the underlying flow would suggest.

What it does not tell you is when. Funding can stay extreme for weeks while price continues in the crowded direction, and every trader who faded it early paid for the privilege.

Treat it as context about fragility, not as a signal. Fading crowded positioning without an invalidation level is a way of joining the next cascade from the wrong side.

The summary that matters

Fees are paid twice. Funding is paid continuously. Slippage is paid on entry and exit. Add them together for a realistic holding period and you have the actual hurdle your idea has to clear.

Most people compute the target and stop. Fewer compute the hurdle. It is the same arithmetic, and it changes which trades are worth taking.