CWN journal

What a trade actually costs you (it's not the 0.1% on the fee page)

CRCryptoWorldNews·Editorial Team··7 min read

Cost breakdown of a $10,000 perpetual round trip — $153 total, with the 0.1% headline fee accounting for only $8 (5%) of the real cost. The other 95% is spread crossings, slippage, and funding.

You buy $10k of a midcap alt on Binance. The trade fills, the screen says you paid $10 in fees, you move on. A week later the position is flat and you exit. Fees again — $10. You think you're down $20 on the round trip.

You're down closer to $150.

The 0.1% taker fee is the only number most retail ever sees, and it's the smallest piece of what a trade actually costs. The bigger costs hide in the spread you crossed to get filled, the slippage on a size your order book couldn't absorb, and — if you held a perp — the funding rate that quietly bled into your position every eight hours. None of these show up as a line item on Binance. All of them are real.

Here's what the four layers look like, with numbers from the actual exchange fee pages, and why they compound into something most active traders never count properly.

1. The spread

Every order book has a bid and an ask, and the gap between them is the spread. When you market-buy, you don't pay the mid-price — you pay the ask. When you market-sell, you don't get the mid-price — you get the bid. The spread is the immediate cost of being impatient.

On BTC/USDT spot, the spread is usually $0.10–$0.50, meaning you lose maybe 0.001–0.002% to it. Negligible. This is why people who only ever trade BTC and ETH genuinely don't believe spreads are a real cost.

Move down the cap curve and it's a different story. On a midcap alt with $5M of daily volume, the spread is regularly 20–50 basis points (0.2–0.5%). On a lowcap with $500k daily volume, 100bps is normal, 200bps when liquidity thins. You haven't placed a trade yet and you're already down half a percent, twice — once getting in, once getting out.

This is also why "I'll just use a market order" is more expensive than people realise. A limit order placed at the mid sits there waiting. A market order eats the ask. On a $10k order in a thin alt, that's $20–50 torched instantly.

2. The headline fee

This is the part everyone knows. Binance spot taker is 0.1%, dropped to 0.075% with the BNB discount. Binance futures taker is 0.04% (0.036% with BNB). Bybit perps are 0.055% taker, 0.02% maker. Kraken retail sits around 0.16–0.26% depending on volume tier. Coinbase Advanced is around 0.4–0.6% for low-volume users — and the regular Coinbase app, the one most newcomers use, is genuinely abusive: a 1% explicit fee on top of a built-in spread that frequently runs another 0.5–1.5%, all-in 2%+ on a round trip before anything else.

Two things people miss.

Maker vs taker is not a small difference. On Binance perps, posting a limit order that sits and gets filled costs roughly nothing — the maker fee is 0.02%, dropping to negative at higher tiers (i.e. they pay you). Crossing the spread costs 0.04%. Half your trades could be limits if you cared. Most retail makes them all markets and pays double for the convenience.

It's per side. A round trip — buy then sell — pays the fee twice. Five round trips a week, fifty weeks, that's 500 fee events a year. A 0.04% taker becomes 20% of capital turned over annually. That's the headline number. We haven't even started.

3. Slippage

Slippage is what happens when your order is bigger than the top of the book. You start eating the ask at, say, $1.0001, but by the time your fill is done you've also eaten $1.0010 and $1.0025. Your average fill price is worse than the price you saw on the screen.

On BTC you can put $1M through and barely move the price. On a midcap with $5M daily volume, a $50k order regularly slips 30–100bps. On a lowcap, a $10k order can slip 200bps. The book just isn't deep enough to absorb you cleanly.

Two practical points.

The book looks deeper than it is. A lot of resting liquidity on alts is from market makers who pull their orders the instant a real bid hits. The "$200k stack at the ask" you saw five seconds ago was a ghost; by the time your market order hits, half of it has cancelled. Effective depth is usually 30–50% of displayed depth on midcaps. Makarov & Schoar (2020) documented this for crypto specifically — most apparent liquidity is reflexive, not real.

Slippage is correlated with your urgency. If you're chasing a breakout, you're often hitting the ask at exactly the moment everyone else is too — depth thins, slippage doubles, you get the worst fill of your week. If you're slowly accumulating with limits, your slippage is approximately zero. Most retail does the former.

4. Funding (perps only)

If you're trading perpetual futures, there's a fourth cost that doesn't exist on spot: the funding rate. Every eight hours, longs pay shorts (or vice versa) a percentage of position size, set by the funding mechanism to keep the perp tethered to spot.

Typical majors funding sits around ±0.01% per 8h. That's ±0.03% per day, ±11% per year if it stayed flat (it never does — it oscillates). On alts, especially trending ones, funding regularly hits 0.1% per 8h, sometimes 0.3%+ during euphoria. At 0.1% per 8h, holding a long perp costs you 0.3% per day in funding alone. Hold it a week and you've paid 2.1% just to be in the trade, separate from price movement.

This is the cost most aggressively underestimated by retail. People hold a leveraged long for a month assuming they're "in profit" without ever checking that funding has eaten 4–6% of notional. The position is up 10% on price and up 4% after funding, and they wonder why their P&L line looks wrong.

Spot doesn't have this. If you're spot-only, feel smug.

What it adds up to

Stack a realistic round trip on a midcap alt — $10k notional, Binance perps, held three days, leverage off the table for now:

Total: $153 on a $10k round trip. That's 1.53% of your capital. The headline fee — the only number you actually saw — was $8 of it. About 5% of the total cost.

Do that ten times a month for a year. You've turned over $1.2M of notional and paid roughly $18k in real costs. If your edge per trade is 1%, you've spent more on costs than your edge generates. This is how active traders who think they're profitable are actually flat, and how flat traders are actually down 15% on the year.

The Barber & Odean (2000) paper found that the most active retail equity traders underperformed buy-and-hold by 6.5% annually, and the entire gap was explained by trading costs, not bad stock picking. Crypto's costs are higher and the trade frequency is higher. Do the math.

What this means for you

Limit when you can. Maker fills on Binance perps cost roughly nothing. Taker fills cost 0.04% each side. Across a year of trading, that's the difference between paying for the rest of this list and paying for nothing.

Fewer, bigger trades. Cost is per-trade, not per-dollar. Ten $10k trades cost more than one $100k trade — same notional, ten times the spread-crossings.

Avoid alt perps unless you've actually checked funding. A 0.1%-per-8h rate eats a 10× lev position 1% per day. The trade has to win that much just to break even.

Reconcile your fee statement against your fills. Most exchanges only show the taker fee in your trade history. The spread and slippage live silently in your fill prices. The funding lives in a separate statement most people never look at. If you've never added them up, you don't actually know what your trading is costing.

The dashboard surfaces realised funding and OI per coin, which is the externally-observable half of this. The other half — your spread crossings and slippage — is in your fill history. It's worth a Sunday afternoon adding it up. Most traders find a number they wish they hadn't.

References

End

Keep reading

Next read

"Measure It, or Lose It"

Last week I argued your crypto portfolio isn't really diversified. Now the uncomfortable follow-up — here is how to prove it to yourself, and what to do about it.

Next read

Why Your Crypto Portfolio Isn't Actually Diversified

You hold ten coins. You feel prudent. In the next drawdown, you will discover you were holding one position in ten costumes.