What Is Slippage in Crypto?
The gap between the price you saw and the price you paid, taught with one worked order book example and the three places it bites hardest.
TL;DR
- Slippage is the gap between the price you expected and the price your order actually filled at.
- The maths: buy 10 units when only 5 sit at £100 and the next 5 at £101, and you pay an average of £100.50, 0.5 per cent over the screen price.
- It bites hardest in thin order books, in shallow DEX pools where your own trade moves the price, and in fast markets where quotes age in seconds.
- A locked quote sidesteps it for a straightforward purchase: Banxa holds its price for roughly 3 minutes, so the number approved is the number paid.
- Mechanics only: this is a how-it-works guide, not trading advice.
The screen said £100. You pressed buy, the spinner turned for a second, and the receipt came back at £100.50. Fifty pence a unit, gone somewhere between the click and the fill. No error message, no warning, and no refund button.
The gap has a name: slippage, the difference between the price you expected and the price your order actually executed at. It is probably the most common surprise in crypto trading, and the machinery behind it turns out to be fairly ordinary once you walk through it. This piece does one worked example, the three situations where slippage bites hardest, and the one purchase model that sidesteps it. It describes mechanics and offers no trading advice.
Where the 50p actually went
Exchange prices come off an order book, a running list of offers to buy and sell at stated prices. The ticker shows the best offer at this instant, and it does not show how much is for sale at that price. The second part is what costs you money.
Slow the trade down: the book offers 5 units at £100 and the next 5 at £101, and you place a market order for 10, which means fill me now at whatever the book holds. The first 5 fill at £100 and empty that level, the remaining 5 fill at £101, and the total bill is £1,005 for 10 units. That is an average of £100.50, or 0.5 per cent of slippage against the £100 on the screen.
Nothing malfunctioned and nobody skimmed you: the book ran out of coins at £100 and your order kept going. For scale, plenty of large exchanges advertise a headline trading fee of around 0.1 per cent, so this one order lost five times the advertised fee to slippage, and slippage appears on no fee schedule.
Thin books punish size
How much you lose comes down to the mismatch between order size and book depth. Ten units into a book holding thousands of units barely registers. Ten units into a book holding twelve is expensive.
Depth is what liquidity means in practice: Bitcoin trades around the clock across hundreds of venues, so its books refill constantly and an ordinary buy hardly dents them. A token outside the top few hundred might have one shallow market, and the awkward part is that a thin market looks completely normal until you actually trade into it. The quoted price carries no information about what size that price survives.
Size works against you in percentage terms as well: slippage grows as orders grow, so a £50 buy and a £50,000 buy of the same token behave quite differently.
On a DEX, your own trade moves the price
A decentralised exchange has no order book at all. Uniswap, live on Ethereum since November 2018, prices every trade by formula from pooled funds: the more of the pool your trade takes, the worse your rate gets, smoothly and automatically. Shallow pools move at a touch.
Because the final rate only settles when your transaction lands on chain, a DEX asks you to set a slippage tolerance, the worst rate you will accept. Set it too tight and the swap fails whenever the price drifts past your limit before landing, and each failed attempt still burns gas. Set it loose and you leave room for front-running bots that sandwich your trade and take value from both sides. That practice goes by the name MEV, and a beginner needs no more than the name. On a busy day, a swap into a thin pool can cost gas twice before anything fills, which is irritating and also normal.
Fast markets outrun the quote
The third situation needs no thin book, only speed. Every quoted price is a snapshot, and if the market reprices between your click and your fill, you are filled at the new price. Crypto supplies plenty of such moments: when US spot bitcoin ETFs were approved on 10 January 2024 and brokerage money suddenly had a route in, prices moved minute by minute for hours. When the Terra system collapsed in May 2022 and roughly $40 billion evaporated inside a week, sellers were hitting books that emptied faster than they refilled. The same word covers a calm Tuesday and a panic, but the numbers involved are very different.
These markets run 24/7, with no closing bell and no pause button, so quotes age in seconds.
What traders do about it
People who trade size tend to repeat three habits. They split big orders into smaller clips, ten £500 buys rather than one £5,000, so each clip sits inside the depth the book is showing. They use limit orders, which name a price and wait, accepting that waiting sometimes means never filling. And before anything large they look at the book or the pool itself rather than the ticker, because depth is what decides the damage. None of this abolishes slippage, it just makes the cost visible before the trade instead of after it.
The purchase model with no book in it
Everything above assumes an order book or a pool. A first crypto purchase usually involves neither, because buying through an on-ramp works on a quoted price: the service shows a rate and you take it or leave it.
Banxa has run this fiat-to-crypto plumbing since 2014, across 100-plus payment methods and 100-plus countries, and its quote locks for roughly 3 minutes in the markets it serves. Approve inside that window, whether you are buying bitcoin or solana, and the number approved is the number paid. There is no book to walk through, no pool to move, and no tolerance slider to guess at.
Once you graduate to exchanges or DEX swaps, slippage becomes a live cost that you size trades around. Until then it is optional, and the locked quote is the boring end of crypto pricing. Boring has its advantages.
Frequently Asked Questions
No. Nobody collects it. A fee goes to the exchange or the network, while slippage is the market repricing your order as it fills. It can even run in your favour: if the book moves your way between quote and fill you pay less than the screen said, which traders call positive slippage. Rare, but real.
Your slippage tolerance was tighter than the price move. The network still processed the attempt, the price check failed, the swap reverted, and the gas for that computation was already spent. Irritating by design: the tolerance existed to stop you taking a worse rate than you agreed.
No single number suits every pool, so this guide will not hand one out. The trade-off is fixed: tighter means more failed transactions in a moving market, looser means more room for the price, and for sandwich bots, to move against you. Major DEXes ship defaults of around 0.5 to 1 per cent on liquid pairs, and shallow pools often need more before anything fills at all.
There is no order book in the transaction, so the fill cannot walk away from the quote. Banxa quotes a price and locks it for roughly 3 minutes in the markets it serves. Approve within the window and the number approved is the number paid. If the window lapses, expect a fresh quote at the current rate rather than a fill at a stale one.
Usually it is just thin-market maths doing its thing. But the same neighbourhood holds real traps, including tokens whose contracts quietly block selling. If a small trade in an obscure token moves its price by double digits, read that as information about the pool, not just a cost of entry.
On price, yes: a limit order cannot fill at a worse price than the one you named. What you give up is certainty of execution. A fast market can gap straight past your level and never come back, so you avoided slippage and missed the move too. That trade-off is the whole story of order types.