What an exchange actually costs you, beyond the advertised fee

Comparing exchanges on their advertised trading fee is like comparing flights on the base fare. The number is real, it is just not what you pay.

The advertised rate is usually the wrong one

Almost every venue charges two different trading fees. A maker order rests on the order book and waits to be filled; a taker order fills immediately against orders already there. Makers add liquidity and are charged less, sometimes nothing, occasionally paid a rebate. Takers remove liquidity and are charged more.

Marketing quotes the maker fee, because it is the lower number. Most retail activity is taker activity, because most people place market orders and want them filled now. If you place market orders, the taker rate is the only one relevant to you, and the gap between the two is frequently a multiple rather than a few percent.

Both rates are also usually tiered by thirty-day volume, and often discounted further for holding the venue’s own token. A pricing page showing the top-tier rate describes a customer you are probably not.

The cost that never appears on your statement

The spread is the gap between the best available buy price and the best available sell price. Cross it and you have paid a cost that is never itemised anywhere.

This is why “zero commission” venues are not necessarily cheaper. A platform earning from the spread instead of a commission has simply moved the charge somewhere less visible, and a wide spread with no fee can easily cost more than a tight spread with one. Spreads also widen during volatility and on thinly traded assets, which is exactly when people tend to trade.

Size makes it worse

Slippage is the difference between the price you expected and the price you got, and it happens when your order is larger than the quantity available at the best price. The order fills progressively deeper into the book at worse prices.

On a major pair with a small order this is negligible. On a thin pair, a large order, or during a fast move, it can dwarf every fee you were carefully comparing. Slippage is also the reason a converter — including ours — gives a reference rate rather than a quote. Nobody fills you at the mid price.

Getting money in and out

Deposit and withdrawal charges are frequently the largest single cost for smaller accounts, and they are usually fixed rather than proportional. A flat withdrawal fee is a rounding error on a large balance and a meaningful percentage on a small one.

Crypto withdrawals also carry a network fee, and venues differ in whether they pass on the actual cost or apply a fixed charge that may exceed it. Fiat withdrawals vary by method, and the cheapest advertised route is often unavailable in most countries.

Putting it together

To compare two venues honestly, price a realistic round trip rather than a single side: buy the amount you would actually buy, at the fee tier you actually occupy, crossing the spread you actually see, then sell it back and withdraw. Our profit calculator handles the fee arithmetic including the break-even price, which is higher than most people assume because the exit fee is charged on the larger sale value.

Two habits reduce the total more than venue-shopping does. Using limit orders where you can converts taker fees into maker fees and avoids crossing the spread. And withdrawing less often, in larger amounts, spreads a fixed charge across more value.

One further charge is easy to miss entirely: currency conversion. Depositing in a currency the venue does not hold natively means a conversion at a rate the venue sets, and that rate frequently carries a margin over the interbank price that is not disclosed as a fee at all. On a platform quoting everything in dollars, a non-dollar funding route can add more than the trade itself costs.

Getting money in is a cost too

Funding method changes the total more than most fee comparisons acknowledge. A bank transfer is typically the cheapest route and the slowest. A card deposit is immediate and frequently carries a percentage charge several times any trading fee — and card issuers often treat crypto purchases as a cash advance, adding their own charge and interest from day one, which appears on a statement the exchange never sees.

Depositing a stablecoin from another venue avoids fiat charges but incurs a network fee on the way out of the first platform. For smaller amounts, choosing the network deliberately matters: the same transfer can cost trivially little or a great deal depending on which chain it moves over, and venues differ in which they support.

A worked example

Consider a modest round trip on a venue advertising a one-tenth of one percent maker fee. You deposit by card at one and a half percent. You buy with a market order, so you pay the taker rate rather than the advertised maker rate, and you cross a spread. You later sell, paying the taker fee again and crossing the spread again. You withdraw, paying a flat charge.

The advertised figure accounted for a small fraction of that total, and the deposit method accounted for more than all the trading fees combined. This is the ordinary case rather than a contrived one, and it is why a comparison built only on published trading fees ranks venues on the smallest component of what they cost.

Two adjustments help more than switching venue: fund by transfer rather than card where the delay is tolerable, and use limit orders so you are charged the maker rate and are not crossing the spread.

Why our fee tables are not live yet

We publish no exchange fee comparison at present, and that is deliberate rather than an omission. Fee schedules differ by tier, by jurisdiction, by payment method and by order type, and they change without notice. Publishing a table seeded with plausible figures would be a set of factual claims about real businesses that we had not verified and could not date.

When it ships, every cell will carry its source and the date it was read, and anything the operator does not publish will read “Not disclosed” rather than being estimated. That standard is set out on our Methodology page.