Why your order gets rejected in a listing's first minutes

SpotRules EditorialUpdated 2026-09~2,100 wordsSafety
Article cover: cream background with a red Safety tag and the headline Why the order keeps getting rejected, two red bars forming a price band with one blue dot inside it and one yellow dot above it

Fair warning before you read on: this page will not help you get filled at the open, and there is no clever trick further down. It answers one thing only — what the system is doing with the numbers you typed during those first minutes, and why the same price is accepted one second and thrown back the next. Knowing it won't get you the coin. It will stop you blaming your connection and firing the same order eight times.

Rejected and unfilled are not the same thing

On the order form both feel identical: your money is still there and the coin is not. They happen at completely different points.

Binance describes a limit order as one placed on the order book with a specific limit price, which executes only if the market price reaches that limit price or better. That is the unfilled case. The order was accepted, it is sitting in the book, and the market simply hasn't come to it. The usual reasons an order sits there are already covered in why a limit order won't fill, so we won't repeat them here.

Rejected happens a step earlier. You hit confirm, and before the order is allowed anywhere near the book it has to pass a set of checks. Fail any one of them and it never lands: nothing in open orders, nothing in trade history, just a message box. In the first minutes of a new listing, most of what people call "I couldn't buy" is stuck right there, not in a queue.

The price field is checked against an average of past trades

The Filters page of Binance's spot API documentation lists, one by one, the checks an order has to clear before the pair will take it. Two of them look at the price field.

The first is PRICE_FILTER, which deals in absolutes: the price has to be at least minPrice, no more than maxPrice, and an exact multiple of tickSize. Typing an extra zero or one digit too many in the decimals usually dies here. That particular slip has its own article, what a fat-fingered limit price does, and it has nothing to do with new listings, so one sentence is enough.

The second one is why you're here: PERCENT_PRICE_BY_SIDE. In the documentation's own words, this filter "defines the valid range for an order price based on an average of previous trade prices". A buy has to clear both ends — no higher than that average multiplied by the buy-side upper multiplier, no lower than the average multiplied by the buy-side lower multiplier. Sells get a different pair of multipliers; the docs state plainly that the range differs depending on the side.

What matters is what it uses as the base. The documentation is specific. When a non-null reference price exists for the symbol, that is used. When one does not, the volume-weighted average price over the preceding avgPriceMins minutes is used instead. If avgPriceMins is 0, the last price is used. All three are trades that have just happened.

So the range you are allowed to type into isn't a band carved into the pair. Trades move, the band slides with them. You never notice on a mature pair because it slides slowly.

At the open, that average is still being made

Take a pair that has been listed for six months. Its average trade price barely shifts over a few minutes, the band stands still, and anything you type near the top of the book is comfortably inside it. That is why most people have never once seen a price rejected for being out of range.

A brand-new pair turns that upside down. The trades the band leans on barely exist before the open; they are built one fill at a time after it. And those same first minutes are when the price moves hardest. The base moves, the band follows — so the price you read off the screen and typed with your own fingers may be measured against a base that has already moved on by the time the order arrives. The number sits inside the range one second and outside it the next. You changed nothing. Neither did the reason for the rejection.

There's an asymmetry on top of that. The documentation says the two sides don't share a range, and when the price is running upward at the open, a buy chasing the last trade is the one that hits its ceiling first. You think you're typing "the current price". The system is comparing it against "the average of the last little while". Those two agree in normal markets. In the first minutes of a listing they can be far apart.

How big the multipliers are, and how many minutes avgPriceMins covers, the documentation returns per symbol — a new pair won't necessarily match the majors you're used to. These values get adjusted, so treat whatever your order page and the pair's trading parameters show at the time as the authority (checked September 2026).

Our own habit, for what it's worth: we don't open the order form during the first minutes of a listing. Not out of fear of typos — it's that "getting it right" isn't a stable target in those minutes, and being careful doesn't buy you much.

Quantity and notional each get their own check

Clear the price field and two more are waiting.

LOT_SIZE governs quantity: at least minQty, no more than maxQty, and an exact multiple of stepSize. A new pair's stepSize often isn't in the same ballpark as the coins you usually buy, so a quantity you've typed a hundred times before can simply bounce.

Market orders get their quantity judged separately. The documentation defines MARKET_LOT_SIZE specifically for MARKET orders on a symbol, with its own minimum, maximum and step — not the same parameters as the limit-order rule. The same quantity can be fine as a limit and refused as a market order.

Third is NOTIONAL, which looks at money. It doesn't judge the quantity you typed; it judges price times quantity, which has to be at least minNotional and no more than maxNotional. The upper end is the part people forget exists. Two switches, applyMinToMarket and applyMaxToMarket, decide whether each end applies to market orders at all.

Here's the sentence that ties this back to the open. A market order has no price, so, as the documentation puts it, an average of previous trade prices is used instead — same order of preference as before: the reference price if there is one, otherwise the volume-weighted average over the preceding avgPriceMins minutes, or the last price when avgPriceMins is 0. The least settled number on the pair is once again the one deciding whether your order is big enough.

Minimum order values differ from pair to pair, so go by what the order page shows you at the time. A greyed-out buy button telling you the amount is too small is a separate path, walked through in why the buy button is greyed out.

A market order on a thin book fills, just at a bad price

After a few rejections most people switch to a market order, on the reasonable theory that if you don't name a price, something has to happen.

Binance's help page on market and limit orders says a market order is executed at the current market price as quickly as possible when a user places the order, and that a market order lets you buy or sell at the best current price. Its worked example uses 2,400 as the current price and notes that the average filled price of your order may not be exactly that number: depending on the direction of your order, it could be slightly below or slightly above it.

"Slightly" is describing a normal book. In the first minutes of a listing, each level holds very little, so your order walks up through them until it has what it asked for. How far the average fill ends up from the number you saw when you pressed the button depends on how empty that stretch of sell orders was — and that gap is one you will actually feel. It is the same mechanism described in what slippage is, and how to lose less to it, with the volume dialled up.

So the usual outcome for a market order here isn't "couldn't buy". It's bought, at a price you'd rather not look at. Whether it clears MARKET_LOT_SIZE and NOTIONAL in the first place is the previous section's problem.

Missing the open is not the same as losing

Getting rejected and slipped in the opening minutes doesn't sting because of the money. It stings because everyone else appears to have got in and you are staring at an error box.

Put the pieces together, though. During those minutes, the judgements the system makes about your price and your order value rest on a number that hasn't settled yet. Not getting in means you didn't take on the fill price that instability produces. It doesn't mean you missed something certain.

Two small practical notes to close. A rejected order never reaches the book, so by that mechanic there is nothing filled and nothing resting to tie up your balance; fix the field and place it again. And if you do want to take part, waiting until a real run of trades has accumulated and the band has settled means the range you're allowed to type into widens back out. At least then you're dealing with a check that holds still.

Heads upThis page explains what the system checks during the first minutes of a listing. It doesn't teach you how to get filled at an open, it doesn't recommend coins, and it doesn't predict prices. Newly listed pairs typically move more violently than established ones; crypto assets are volatile and can go to zero, and nothing here is investment advice.

FAQ

The same price worked a second ago and was refused the next. Is my connection the problem?

Not necessarily. Binance's spot API documentation says the valid range for an order price is based on an average of previous trade prices: the pair's reference price when a non-null one exists, otherwise the volume-weighted average price over the preceding avgPriceMins minutes. In the first minutes of a new listing that base is being rewritten trade by trade, so the range shifts under you and the same number can land inside it and then outside it. Whatever message you get is whatever the page shows at the time.

If I switch to a market order at the open, can it still be rejected?

Yes. The documentation gives market orders their own quantity rule, MARKET_LOT_SIZE, and the NOTIONAL filter has applyMinToMarket and applyMaxToMarket switches that decide whether its minimum and maximum apply to market orders. Because a market order has no price, an average of previous trade prices is used in its place for that calculation. The practical difference is in the outcome: a rejected limit order means you bought nothing, while a market order on a very thin book usually means you bought something at an average price you won't enjoy.

Sources

The definitions of a limit order ("placed on the order book with a specific limit price", executing "only if the market price reaches your limit price (or better)") and of a market order ("lets you quickly buy or sell an asset at the best current price") come from Binance's help page on the different order types in spot trading (English page, checked 2026-09-20). "A market order is executed at the current market price as quickly as possible when a user places the order", and the worked example in which the average filled price of an order may not be exactly 2,400 and could be slightly below or above it depending on direction, come from Binance's help page on market and limit orders (English page, checked 2026-09-20); 2,400 is that page's own example figure, not ours. PRICE_FILTER, PERCENT_PRICE_BY_SIDE, LOT_SIZE, MARKET_LOT_SIZE and NOTIONAL — their definitions and pass conditions, the phrase "an average of previous trade prices", the reference-price and avgPriceMins order of preference, the separate ranges for the buy and sell sides, and the use of an average of previous trade prices in place of a price for MARKET orders — are all from the Filters page of the Binance spot API documentation (checked 2026-09-20). This page was written from those public documents and contains no screenshots of our own trading; multipliers, step sizes, notional limits and real depth vary by pair and by market, so go by what your page shows when you order.

SR
SpotRules Editorial
A pen-name team · we explain how to read the rules, we don't shill coins or call trades · Editorial principles