Calculator Updated 2026-08-29 Runs in your browser By Ivo Renner

Weekend gap exposure

A perpetual can move while its underlying cash market is closed. Enter a hypothetical adverse price move to see its effect on notional and initial isolated margin. This is a scenario calculation, not a forecast of the next opening gap or an estimate of your exchange liquidation price.

The arithmetic, written out

Three lines:

isolated margin = notional ÷ leverage
loss = notional × (gap% ÷ 100)
share of margin consumed = (loss ÷ isolated margin) × 100
maximum survivable move = (1 ÷ leverage) × 100

The last line is the one worth memorising, because it collapses the entire leverage question into a single percentage. At 10× you have posted a tenth of the position's value, so a 10% adverse move consumes all of it. At 20× the figure is 5%. At 25×, the highest cap on the roster, it is 4%. At 5× — the cap Binance assigned to TSLAUSDT, the lowest on the roster — it is 20%.

Leverage caps shown are the maximums stated at launch; Binance adjusts tiers after listing. The full roster is on the contract archive.

A worked example

The defaults above: a 1,000 USDT long at 10×, facing a 6% adverse gap.

• Isolated margin = 1,000 ÷ 10 = 100 USDT.
• Loss = 1,000 × 0.06 = 60 USDT.
• Share of margin consumed = 60 ÷ 100 = 60%.
• Maximum survivable move at 10× = 10.00%.

On paper the position survives: 6% is less than 10%. In practice it is in serious trouble, because 60% of the margin has gone in a single print and the maintenance requirement is now uncomfortably close. Take the same 6% gap at 20× and the loss is still 60 USDT but the margin behind it was only 50 — the position is gone before you have read the news that caused it.

That asymmetry is the reason the third readout exists. A gap does not need to reach your theoretical wipe-out level to end the position, and it does not need to be large to be fatal at the top of the leverage range.

What actually happens over a weekend

Three things happen at once, and they compound.

The underlying stops printing. From Friday's US close to Monday's open there is no constituent price. Whatever the world learns in those roughly 65 hours (Friday’s 16:00 ET close to Monday’s 09:30 ET open) — an earnings pre-announcement, a regulatory decision, a weekend headline — is not in any stock price, because no stock price exists.

The equity index can move during a closure. The current Binance FAQ describes Orderbook EWMA for equity closures, replacing Fixed mode from 16 May 2026. It smooths order-book impact prices. In this mode the mark still uses the standard median calculation; a closure does not leave the index pinned to Friday’s cash close. Rules checked 11 September 2026.

Funding can continue while cash trading is closed. The number of payments depends on the contract’s interval and the settlement times crossed. Two full days at eight-hour intervals give six settlements; four-hour intervals give twelve. Use the funding calculator with current contract settings. These costs are separate from the price shock calculated here.

Then the market reopens and the position re-prices. Not gradually — the gap arrives as the index resumes tracking real constituent prices, and the mark price follows it in.

There is no opportunity to exit between Friday's level and Monday's, because the price that would have let you exit never traded.

Why the deviation band will not save you

Binance applies a deviation band that limits how far the mark price can travel from the index, and it tightens when the market is shut. Readers reasonably assume this caps weekend damage. It does not, and the reason is worth one paragraph.

The deviation band constrains the mark relative to the index, not relative to your entry or Friday’s price. The index itself can move through order-book pricing and when external prices return. The band is not a maximum position loss, an execution-price guarantee or a cap on the gap you should consider.

The two official Binance pages state different figures for this band. That discrepancy is set out in full, with both sets of numbers, in how Binance builds the price index and the mark price — it is that page’s to own, and repeating the table here would only give you two places to keep in sync.

Three levers, and only three

Run your actual notional and leverage against a gap the size of a bad but unremarkable day in the underlying — not a crash, just an earnings reaction. If the readout says a large share of your margin is gone, you have three levers, and only three. Reduce size, which reduces the loss in USDT. Reduce leverage, which widens the survivable move. Or close before Friday's close and reopen on Monday, accepting two round-trip costs in exchange for not holding a position through a period in which you cannot be right or wrong, only exposed.

Adding margin over the weekend is the tempting fourth option and it is the weakest one: it improves the ratio in the second readout without changing the loss in the first, and it commits more capital to a position you have already identified as fragile.

Index ETFs gap less than single names, for the obvious reason. Leveraged ETF perps gap more, and they carry a second layer of leverage on top of yours — see ETF perps and the double leverage nobody mentions. For the mechanics of what the index is doing while you sleep, read trading outside US market hours and check the session clock for which mode is plausibly live right now.