- What a grid actually is, and who it's for
- How it makes money: off volatility, not direction
- Spot grid vs. futures grid
- The three grid types: long, short, neutral
- Setting the parameters: range, grid count, per-grid size, trigger price
- Don't ignore fees: you pay both ways
- When you lose, and how to cut it
- We walked through creating a spot grid
- FAQ
Grid trading has taken off on Binance over the last few years, and the reason isn't complicated: it turns something that used to need constant screen-watching and judgment into a bot that just runs once you set the parameters. You don't have to sit on the chart all day buying low and selling high by hand — the bot does it for you. Sounds relaxing, and that's exactly where the trouble hides. Plenty of people never work out what the bot actually earns on or when it loses money; they watch a row of small profits pile up and assume it's safe, then the market runs one direction, the unrealized loss swallows those small profits, and they discover they're stuck holding a pile of coins bought near the top.
This is our long read that walks through grids end to end: how they work, the difference between spot and futures, which of the three directions fits which kind of market, how to actually set the parameters, the fee cost everyone forgets, and — the part that matters most — when a grid loses money and how you should cut it. By the end you should be able to judge whether this particular pile of money, in this particular market, has any business being run as a grid.
1. What a grid actually is, and who it's for
In one sentence: grid trading means picking a price range and, inside it, buying a little every time price drops one grid level and selling a little every time it rises one, over and over — buy low, sell high, collect the spread on the round trips. It doesn't predict up or down. It doesn't care whether tomorrow is green or red; it only cares whether price keeps bouncing around inside the range you drew.
Think of it as a vending machine. Across a band from, say, 2,000 to 3,000, you place a resting order every so often: when price drops to a level, a buy order sits there; the moment it fills, a sell order goes up one grid above it, waiting; when price rises and that sell fills, a buy goes back down a grid. The bot just loops through this, grid by grid, and as price sloshes around inside the range, it pockets the spread one fill at a time.
So a grid feeds on chop. Its happiest market is sideways back-and-forth with no clear direction: price grinds around inside a box and the bot keeps filling. What it fears is a one-way move. On the way up, your coins get sold off early and you're left holding cash, watching the move run away without you (missing out). On the way down, you buy at every level, more and more the further it falls, until price breaks below your lower bound and you're stuck holding high.
So who's it for? People who are willing to hand a chunk of money to a choppy market for a while and can't be bothered to trade it by hand. If your read on a coin is "it won't rip or crash in the near term, it'll just wobble in some range," a grid is the tool that turns that read into money. If you genuinely can't call the direction and don't want to watch, DCA suits you better — it isn't picky about the market. And if you're convinced something's about to run hard, just buying and holding beats a grid, because the grid will sell your coins off too early.
2. How it makes money: off volatility, not direction
This is the one thing to get straight in the whole piece. A grid's profit comes from volatility, not from which way price goes. Start at 2,500, and a month later you're back at 2,500, but along the way price swung between 2,200 and 2,800 a dozen-plus times — over that stretch a buy-and-hold makes nothing, while the grid banks a spread on every one of those dozen-plus round trips. The more often price swings and the more round trips it makes, the more grid levels fill, and the higher the grid's cumulative return.
How does the spread add up, grid by grid? Say you split 2,000 to 3,000 into 20 grids on a geometric grid — the gap between two neighboring levels is a bit over 2%. The bot buys a slice at 2,400 and sells it when price rises to the level around 2,450; that one round trip earns roughly 2% gross (subtract the two fees and you get the net). One grid isn't much, but a day of wobbling inside the range can trigger several, and over a few dozen days the count adds up. Mind the gap between "gross" and "net": every buy and every sell pays a fee — one buy, one sell, two charges — and both come out of that 2% before what's left is what you actually keep.
Here's the counterintuitive but important takeaway: the tighter you slice the grids, the thinner the per-grid profit, and the easier it is for fees to eat it whole. Take the same range from 20 grids to 100 and each grid is only about 0.4% apart, while two fees for the round trip can swallow most or all of it. So "more grids = more money" is wrong; the density has to match the fees and the swing size you actually expect. To feel the trade-off directly, drop the range, grid count, capital and fee rate into our grid profit range simulator and see whether the per-grid net comes out positive or negative.
3. Spot grid vs. futures grid
Binance grids come in two flavors, spot and futures, and the first thing a beginner should sort out is that the risk between them is night and day. A spot grid uses the real coins and USDT in your account, with no leverage — if price breaks below your range, you're holding a stack of spot bought high, at an unrealized loss, but the coins are still yours. There's no such thing as "liquidation" here; the worst case is holding and waiting to break even, or selling at a loss.
A futures grid runs the same buy-low-sell-high logic on futures, with leverage and a liquidation price. The upside is capital efficiency and the ability to short (you can profit when price falls too); the downside is that if price moves far enough against you it triggers forced liquidation, and at that point you're losing real margin, not paper. You also have to account for funding rates, a cost spot doesn't have — hold long enough and that adds up too.
Put simply: if you're unsure and want to play it safer, start with a spot grid; only touch a futures grid once you want to short or chase capital efficiency and you understand how liquidation and funding work. The finer differences and when to use each get their own piece — how to choose between a spot grid and a futures grid goes deeper; here we'll leave it at that.
4. The three grid types: long, short, neutral
Beyond the spot-vs-futures split, a grid itself comes in three flavors by "direction bias," each matched to a kind of market. Pick the wrong direction and no matter how busy the bot is, it's running against the market.
| Type | Roughly which market | What it does |
|---|---|---|
| Long grid | Leaning bullish, or a box grinding upward | Buy low, sell high; the range usually sits from around the current price and up, buying more the lower it goes, betting price bounces around and slowly climbs |
| Short grid | Leaning bearish, or grinding downward (mostly futures) | Sell high, buy low; sell as price rises, buy back as it falls, betting price bounces around and slowly drifts lower |
| Neutral grid | A pure chop box with no clear direction | Assumes no direction; price sits mid-range and orders are placed both ways, purely harvesting the back-and-forth |
Most beginners on spot are really running a long-grid mindset: they like a coin, expect it to chop within some range while staying strong over time, buy more as it dips to lower the average, and scale out as it rises. Short grids live mostly on futures, since shorting on spot is awkward. The neutral grid is the purest — it only bets "it'll chop," not "which way" — and it makes the biggest demand on your range call: draw the range wrong, price walks straight out, and a neutral grid gets stuck or misses out just the same. How to pick each direction and what differs in the parameters is covered in how to use long, short, and neutral grids.
5. Setting the parameters: range, grid count, per-grid size, trigger price
Eighty percent of whether a grid works or fails is in the parameters. The bot just faithfully executes the rules you set; set the rules wrong and it faithfully loses money for you. There are only a handful of core parameters:
Upper and lower bounds
This is the single most important parameter, no contest. The range is where you think price will keep wobbling next. The only rule for these two numbers: make the range cover the swing you expect — set the upper bound at a high you figure it probably won't reach in the near term, and the lower bound at a low that would still be reasonable and where you'd be happy to keep buying. Too narrow, and price leaves the range quickly, so the grid either misses out or gets stuck; too wide, and the grids are spread thin, with small per-grid spreads, few fills, and your money working inefficiently. Most people err on the narrow side, because a narrow range "looks" busy and high-return in a backtest — but in real life, the moment the market leaves that narrow band, it falls apart.
Grid count / density
Grid count sets how big the spread is between levels. More grids means a smaller spread per grid and more frequent fills, but as we said, too dense and fees eat it. Fewer grids means a bigger spread per grid and less frequent fills, but each one earns something real. How to decide? The floor is making the per-grid spread clearly bigger than two fees for the round trip. To work out how many grids a given range should have and the gap per grid, use the grid spacing / count calculator, then use the grid profit simulator to check that the per-grid net actually comes out positive.
Per-grid size
Per-grid size = total capital ÷ grid count. It decides the absolute amount each fill earns. With fixed capital, more grids means a smaller per-grid size, and that thin per-grid spread in dollar terms can shrink to where it's not worth it — and you still pay two fees. So when capital is small, better to use fewer grids and make each one thicker than to spread a tiny amount across a dense grid.
Trigger price
The trigger price is optional: you can hold the grid off instead of starting it right away, waiting until price pulls back (or spikes) to a level before it kicks in. The upside is that you avoid laying out all your grids at a poor entry. For example, if you like a coin's chop but the current price is on the high side, you can set the trigger lower and let the bot enter only if it really drops. No trigger price means it builds the position at the current price immediately.
How these parameters pull against each other and a typical order to set them in gets a finer breakdown in how to set grid parameters.
6. Don't ignore fees: you pay both ways
This is the hidden pit beginners fall into most. Every fill in a grid pays a fee, and a grid is by nature a "fill often" machine — you think the bot is diligently earning you spread, but every move it makes also hands the exchange a fee. One round trip per grid is a buy and a sell, two charges. The denser the grids and the thinner the spread, the bigger the share fees take, and past a certain density each grid is actually negative — the busier the bot, the faster you lose.
Binance charges spot fees separately for maker (resting) and taker (aggressive) orders, and the rate also varies with whether you pay in BNB and with your VIP tier, so go by the current number your account shows on Binance's page — we won't hard-code it here. Grids mostly rest limit orders, which in theory makes maker fills more likely, but not guaranteed. What you can do is run the math with your real fee rate before setting the grid count: does the per-grid spread clearly cover two fees? For how the math works and how to read each tier, see how Binance fees are calculated, and to quickly estimate the cost of a single trade use the fee calculator.
7. When you lose, and how to cut it
A grid was never a can't-lose tool. It has very clear ways to lose money, and knowing them is what lets you cut losses in time.
A one-way drop through your lower bound is the classic way to lose. Price heads straight down, the bot buys at every level, more the further it falls, and once it breaks the lower bound you set, the bot stops buying — leaving you holding coins bought all across the range, at an average cost well above the current price, and the unrealized loss can be big. At that point the small realized profits from earlier don't come close to covering it. On a spot grid this means you're stuck holding (the coins are still there); on a futures grid it can mean liquidation the moment price hits the liquidation level, and losing the margin outright.
A one-way rally you miss out on is the other kind of loss — strictly speaking, it's earning less. Price runs straight above your upper bound, your coins were sold off grid by grid, you're left holding cash, and you watch the move fly away with no way to get back in. Here the grid's return is far worse than if you'd just bought and held.
So how do you cut it? Set yourself a few hard rules instead of judging by emotion in the moment. First, watch total equity, not realized profit — Binance's grid screen shows a total return (including unrealized P/L), and that number is the real ledger; the row of small realized profits creates an illusion. Second, when price nears or breaks the lower bound, it's decision time — either take the loss and stop, or genuinely commit to holding these coins long-term at a lower price; if it's the latter, you've effectively turned the grid into passive DCA, so be clear about what you're doing. Third, when the market clearly turns one-way, against the chop assumption you started with, don't white-knuckle it — stopping and re-planning is more rational than stubbornly waiting to break even. The full set of scenarios and responses is in why grids lose money and how to cut it in time.
8. We walked through creating a spot grid
On the spot-grid creation page, the first thing to pick is the trading pair — choose a liquid, mainstream coin you'd be willing to hold through some chop; don't run a grid on an obscure small-cap, where the book is thin and the spread is wide and slippage eats your edge before you've earned any. Once you've picked the pair, the page lets you choose between "manual" and "smart/recommended" parameters. Beginners often just take the recommended settings to save effort, but the recommended range is usually computed from recent price action — it assumes the market keeps chopping the way it did, which may not match your own read. Our advice is to treat the recommended values as a reference and still call the final range yourself.
The next three boxes are the ones to fill carefully: upper and lower bounds, grid count, and amount. The easiest slip here is flipping the bounds or adding/dropping a zero, so before you submit, go back and sanity-check the order of magnitude. After you fill the grid count, the page usually shows the spacing percentage per grid — pause and look at whether that percentage is enough to cover two fees. That's the most direct on-the-spot signal for whether this grid is worth running; if the spacing is too small, dial the grid count down. After you enter the amount, watch whether the minimum per-grid size it shows drops below the platform threshold — too low and it won't build, or each grid is too thin to mean anything.
Below that are a few optional fields: trigger price (whether to wait for price to reach a level before it starts) and take-profit / stop-loss price (the whole grid ends automatically when price hits a line). Neither is required, but at least think through the stop-loss line — it's your safety rope for when you're not watching. The last step is confirming the order; Binance makes you double-check the parameters, which is your last chance to fix a mistake, so scan the range, grid count and amount once more and you can build it. Once it's live, the thing to actually watch isn't that constantly ticking realized profit but the total return and how far price is from the edges of your range. The flow itself isn't complicated; the hard part was never which button to click, it's whether those few numbers up front are right.
9. FAQ
Is grid trading good for beginners?
The setup is easy enough — starting a grid is simpler than learning every order type — but the way it "looks like it's making money" is exactly what fools beginners. It fills often in a choppy market and the row of small profits looks great, but the moment price drops in one direction and breaks below your lower bound, the unrealized loss buries those little profits, and beginners usually don't notice until the pain sets in. Start with a small amount, a wide range and few grids, watch total equity rather than realized profit, and then decide whether to add more.
What's the minimum to start a grid?
Binance spot grids have a minimum investment, and the threshold varies by trading pair and by period — go by the minimum amount shown on the page when you create it. The real constraint isn't the platform minimum, it's the grid count: capital ÷ grid count is your per-grid size, and if that's too small, the price gap per grid won't even cover the two fees for a round trip, so the grid bleeds. If your capital is small, use fewer grids rather than a very small amount across a dense grid.
Can a spot grid get liquidated?
A spot grid uses your own capital with no leverage, so there's no liquidation — the worst case is price breaks below your lower bound and you're stuck holding coins bought high, an unrealized loss, but the coins are still yours. A futures grid is different: it uses leverage and has a liquidation price, and if price moves far enough against you it triggers liquidation, meaning you actually lose the margin. That's the most important difference between the two, so if you're unsure, start with a spot grid.
What if I set a parameter wrong?
The most common mistake is setting the range too narrow, so price quickly leaves it and the grid spins idle or gets stuck; the next is packing in too many grids, so fees eat the per-grid spread. You generally can't change the range or grid count on a running grid, so if you realize it's wrong you usually have to stop it, re-plan and build a new one. When you stop it, check whether the coins you already bought sell at market or as limit orders, so you don't eat another round of slippage on the way out.
Grid vs. DCA — which should I pick?
It depends on your read of the market and whether you want to watch it. If you think price will chop up and down for a while and you can roughly draw a range, a grid fits better — it earns the spread off the back-and-forth; if you can't call the direction, just want to build a position in pieces over time to lower your average cost, and don't want to fuss over it, DCA is easier — it doesn't need chop and can't get stuck outside a range. The two don't conflict; plenty of people DCA the core position and run a grid with a small slice to catch the swings.
This article is not investment advice, and the products carry risk. Reference: Binance Help Center.