"Start by adjusting the leverage." That's the first instruction in Binance's own walkthrough for setting up a futures grid. How much is right comes down to how much of that margin you can afford to lose, so we won't hand you a number. What we can show is exactly what the setting does inside the bot. Your initial margin stays the same; the help page's formula is Total Investment = Initial Margin × Leverage, and the quantity in every grid order scales with it. Put 300 USDT in at 3x and each order is three times the size it would be at 1x. The position a one-way move piles up, and the unrealized loss on it, scale by the same three.
Two quieter things also hang on that setting: how little money it takes to start, and whether the grid survives a change to the pair's leverage limits. The sections below follow the bot's own order of operations. All examples are hypothetical. How a grid actually gets liquidated is covered in our piece on futures grid liquidation, so it only gets a mention here.
1. Three formulas with leverage in them
Leverage gets set first, before you decide how much to commit. The help page then gives the arithmetic, and the multiplier turns up in each line:
- Total Investment = Initial Margin × Leverage. This is the size the bot actually spreads across its orders. The page also writes it the other way round, Investment = initial_value / leverage. We read that as the same relationship from the opposite end.
- grid_qty = adjust_coef × initial_margin × leverage / sum(price). grid_qty is the quantity per grid order. On this neutral-grid formula, keep the margin and the range where they are, double the leverage, and every order doubles.
- min_initial_margin = min grid qty × sum(price) / (leverage × adjust_coef). Here leverage sits underneath the line, so a higher multiplier means a lower minimum.
One caveat on the last two: the help page gives them as the Neutral version for USDⓈ-M futures grids. Long and Short grids get longer formulas with an extra term involving leverage and the mark price, so the clean ratios here don't carry over. Our example uses the neutral version only; for a long or short grid, go by the figures the creation screen shows.
On adjust_coef, the page says it is currently defaulted to 0.8 (checked October 2026; treat whatever the page shows when you create the grid as the live value). sum(price) depends on your range and number of grids; the example below just assigns it a value.
Follow the money and it changes shape twice. You commit margin, the bot multiplies it into a total investment, then slices that into orders at each level. Leverage controls the middle step.
Margin that stays tied up while the grid runs depends on leverage as well. In Cross Margin, occupied margin = present notional / current leverage. Isolated Margin mode is currently not supported for COIN-M Futures grid strategies.
2. One 300 USDT grid at 1x, 3x and 5x
Example (neutral grid, hypothetical numbers, not tied to any real coin, ignoring fees, funding and any initial position): 300 USDT initial margin, adjust_coef at the current default of 0.8, and a range and grid count that make sum(price) come to 44,000. Price slides one way, all ten buy orders below fill at an average of 2,100 USDT, then price keeps going and ends 10% under that average.
| Leverage | Total investment | Qty per order | Position after 10 fills | Value at average | Loss after a further 10% drop | Loss vs 300 USDT |
|---|---|---|---|---|---|---|
| 1x | 300 USDT | ~0.00545 | ~0.0545 | ~114.5 USDT | ~11.5 USDT | ~3.8% |
| 3x | 900 USDT | ~0.01636 | ~0.1636 | ~343.6 USDT | ~34.4 USDT | ~11.5% |
| 5x | 1,500 USDT | ~0.02727 | ~0.2727 | ~572.7 USDT | ~57.3 USDT | ~19.1% |
Same money in every row. Only the leverage changed, and order size, position and loss all run 1 : 3 : 5. If price keeps falling past those ten levels, the gap keeps widening in that ratio.
You won't find a liquidation price in the table. That depends on maintenance margin tiers and the mark price, and a made-up figure would just get compared against real positions. Section 3 of the liquidation article walks through the mechanism.
Now the entry ticket. Keep sum(price) at 44,000 and assume the pair's minimum grid quantity is 0.001. Plugging into the minimum-margin formula gives about 55 USDT at 1x, about 18.3 USDT at 3x and 11 USDT at 5x. That's the pull of high leverage: a futures grid for roughly the price of lunch. The minimum drops because the same smallest order can be reached with less margin. Thinner margin also means the same price swing eats a bigger share of it. Spot grids work their minimum out differently; see how much money a grid bot needs to start.
3. Leverage caps, notional limits and other ways a grid stops
Leverage limits aren't fixed per pair forever. When Binance lowers a pair's maximum leverage, any running grid order set above the new cap is cancelled, and the help page ties that to the next new order being placed, so it can happen mid-run at a moment you didn't choose. Run a grid near a pair's ceiling and you're first in line when that ceiling comes down. This article doesn't list any pair's limits.
The second limit is size. Each leverage bracket comes with a maximum notional value, and once the grid's position grows past it, the grid is shut down. A higher multiplier puts more notional to work from the same margin, and a one-way move keeps adding to it, so a high-multiple grid riding a strong trend is the one most likely to hit that cap. The page adds a third route: a grid can also end when there isn't enough margin left to support it.
There's an Auto-add Margin option for the first case. It's off unless you switch it on, and per the help page it only steps in when the leverage bracket is cut. It doesn't top you up as a drawdown grows.
4. What a running grid lets you change, and how to back into a multiple
On a USDⓈ-M futures grid, the price range and grid count can still be changed while it runs; on a COIN-M futures grid, the page says both are locked from the moment the grid is created. Leverage isn't on the list of things you can modify, and the help page flags some creation settings (such as mode and investment) as "Cannot be modified after placing the grid order". We treat the multiplier as locked the moment you create the grid.
So the order we'd work in puts leverage last:
- Decide the most this grid is allowed to lose. The position size calculator works back from that figure to a position you can carry.
- Pick a stop below the range and work out what a table like the one above shows at that price. Whether your stop or your liquidation price comes first is covered in where to set a grid stop-loss.
- Divide the position you worked out by the margin you're willing to commit. That's your leverage ceiling. If it comes out at 1x, use 1x; if even 1x is too much, put in less. If you'd rather skip leverage altogether, spot grid vs futures grid sets the two side by side.
Nothing here is investment advice. Futures grids use leverage, can't guarantee profit, and you can lose your entire margin; decide for yourself and size accordingly. The margin, sum(price), average price, minimum quantity and price drop above are hypothetical figures for explaining the mechanism; they ignore fees and funding and are not a way to calculate a liquidation price. Formulas and rules are summarized from the English edition of Binance's What Is Futures Grid Trading? page (updated 2026-10-26), checked 2026-10-27. The Chinese and English editions of this page are not translations of each other, and this version uses the English one only. adjust_coef and leverage limits can change; go by what the page shows when you create your grid. References: What Is Futures Grid Trading?