Scalping vs swing trading crypto: how time frame changes fees, stops and leverage
Scalping and swing trading are two ways of holding a trade, separated mainly by time. A scalper aims to catch small moves over seconds to minutes; a swing trader aims to catch larger moves over days to weeks. That difference changes how much of your target the fees eat, how wide your stop needs to be, and how leverage interacts with liquidation. This guide works through those numbers using the simulator fees (0.05% taker, 0.02% maker on perpetual futures).
Time frames and what each style looks like
There is no strict definition, but the usual working ranges look like this:
| Scalping | Swing trading | |
|---|---|---|
| Typical holding time | Seconds to minutes | Days to a few weeks |
| Charts often used | 1-minute to 15-minute | 4-hour, daily, weekly |
| Target per trade | A small move, often well under 1% | A larger move, often several percent |
| Stop distance | Tight, a fraction of a percent | Wider, often a few percent |
| Trades per week | Many | A handful or fewer |
| Time at the screen | High, during the session | Lower, with periodic checks |
The percentages are illustrative, not rules. What really separates the styles is the ratio between the move you are trying to catch and the fixed cost of trading it.
How fees change the scalping maths
On the simulator's perpetual futures a market or stop order that fills immediately pays the taker fee of 0.05% of notional, and a resting limit order that gets filled pays the maker fee of 0.02%. You pay a fee on both the open and the close, so a round trip costs between 0.04% and 0.10% of notional depending on how each side fills. See maker vs taker fees for the mechanics.
Take an illustrative scalp: $5,000 notional, a target of +0.3% and a stop of -0.2%. Before fees that is $15 to win against $10 to lose. Now add a taker fill on both sides (about $2.50 each way, $5.00 round trip):
Net win = 15.00 - 5.00 = $10.00 Net loss = 10.00 + 5.00 = $15.00 Break-even win rate = 15 / (10 + 15) = 60%
Without fees the same trade breaks even at a 40% win rate (10 / (15 + 10)). The round-trip fee of 0.1% is one third of the 0.3% target, and it turns a trade that looked like 1.5 to 1 in your favour into 1 to 1.5 against you. The table repeats the sum for different fill types on the same trade (fees are approximate, because the exit notional moves slightly with price):
| Fills (open / close) | Round-trip fee | Net win | Net loss | Break-even win rate |
|---|---|---|---|---|
| No fees (theory) | $0.00 | $15.00 | $10.00 | 40% |
| Maker / maker | $2.00 | $13.00 | $12.00 | 48% |
| Maker / taker | $3.50 | $11.50 | $13.50 | 54% |
| Taker / taker | $5.00 | $10.00 | $15.00 | 60% |
On the simulator only a filled resting limit order (including a reduce-only limit used to close) pays the maker rate. Market, stop-market and triggered stop-limit orders, and attached stop-loss, take-profit and trailing-stop closes, pay the taker rate. So a stopped-out scalp pays taker on the exit, and the maker / maker row assumes both sides were resting limit orders.
The same fees on a swing trade
Now take an illustrative swing trade at the same $5,000 notional, with a target of +6% and a stop of -4%. Before fees that is $300 to win against $200 to lose. Taker fills on both sides cost about the same $5.00:
Net win = 300.00 - 5.00 = $295.00 Net loss = 200.00 + 5.00 = $205.00 Break-even win rate = 205 / (295 + 205) = 41%
The fee is now under 2% of the target, and the break-even win rate only moves from 40% to 41%. The dollar cost is identical; what changed is how far price has to travel to pay for it.
On a real exchange, a perpetual position held across funding times also pays or receives funding, which matters more the longer you hold. Funding is not simulated on CryptoPaperTrading, so remember to account for it separately when you think about longer holds.
Stop sizes and position size
If you fix the amount you are willing to lose per trade, the stop distance decides the position size. Say you risk $50 (0.5% of the $10,000 virtual balance) on each trade:
- Scalp, 0.2% stop: size = $50 / 0.002 = $25,000 notional. A taker round trip on that is about $25, so a full stop-out costs roughly $75, not $50.
- Swing, 4% stop: size = $50 / 0.04 = $1,250 notional. A taker round trip is about $1.25, so a stop-out costs roughly $51.25.
This is the hidden cost of tight stops: fees scale with notional, not with risk. For the scalp, the fees add 50% to the planned loss; for the swing trade, 2.5%. If you size scalps this way, include the round-trip fee in the risk budget. The position size calculator does the size step, and the stop-loss guide covers where to place the stop.
Leverage implications for each style
Leverage only sets how much margin backs a position; it does not change the dollar risk of a stop. What it does change is where liquidation sits. On the simulator, margin is isolated by default and maintenance margin is a flat 0.5%, so a fresh isolated long is liquidated after these falls from entry:
| Leverage | Fall to liquidation (long) |
|---|---|
| 5x | 19.60% |
| 10x | 9.55% |
| 20x | 4.52% |
| 25x | 3.52% |
Those figures come from the long formula E x (1 - 1/L) / 0.995, and you can check any entry with the liquidation price calculator.
- Scalping: a 0.2% stop sits far inside even the 25x liquidation distance, so the stop normally fires long before liquidation. The real danger for a scalper is the large notional from the previous section: more leverage makes that notional cheap in margin terms, which makes it easy to oversize.
- Swing trading: a 4% stop is beyond the 3.52% liquidation distance at 25x, so the position would be liquidated before the stop could trigger. The simulator rejects a stop-loss placed beyond the liquidation price, and the fix is lower leverage. At 20x the stop fits with only about half a percentage point to spare; at 10x or below there is a comfortable gap.
Maximum leverage also varies by coin: 25x for BTC and ETH, 20x for coins such as SOL and XRP, 10x for others such as SUI and PEPE. Anything above 10x is a Pro feature. See leverage and margin for more.
Which style to practise first
This is a question about learning, not about which style makes money; neither is better in general.
- Decision speed. Swing trades give you time to plan the entry, stop and target before you act. Scalping asks for the same decisions in seconds, which is harder while you are still learning the order ticket.
- Fee sensitivity. Scalping results depend heavily on how your orders fill. Practising it teaches fee awareness quickly, but a small mistake in order type shows up directly in the results.
- Sample size. Scalping produces many trades quickly, so you can review a large sample sooner. Swing trading takes weeks to build the same number of trades.
Whichever you start with, keep a trading journal and record the fee on every trade next to the gross result, so you can see how much of each style's outcome goes to costs.
Practising both on the simulator
CryptoPaperTrading gives you $10,000 in virtual USD and prices from Binance public market data for USDT-margined perpetuals. For scalping practice, try entering with a limit order to see the maker fee in action, and compare closing with a reduce-only limit order (maker) against an attached stop-loss or take-profit (taker). For swing practice, use the TradingView charts on higher time frames and attach a stop-loss that sits inside the liquidation price. Both can be practised side by side on the futures simulator without real money.
This guide is for education only and is not investment advice. Prices and percentages in the examples are illustrative round numbers.
Frequently asked questions
Why do fees matter more for scalping?
Because the fee is a fixed share of notional while the target is small. With a 0.3% target, a 0.1% taker round trip is one third of the gross gain. With a 6% target, the same fee is under 2% of it.
Does using maker orders help scalping?
It lowers the cost. On the simulator a resting limit fill pays 0.02% instead of 0.05%; in the worked example, maker fills on both sides cut the break-even win rate from 60% to 48%. Attached stop-loss and take-profit closes pay the taker rate.
Can I use 25x leverage for swing trading?
You can set it on BTC and ETH, but at 25x a long is liquidated after about a 3.52% fall, so a wider swing stop may sit beyond liquidation. The simulator rejects a stop-loss placed beyond the liquidation price; lowering leverage fixes it.
Is funding included when I hold a swing trade on the simulator?
No. Funding payments are not simulated. On a real exchange a perpetual position held across funding times pays or receives funding, so factor it in separately.
Is this investment advice?
No. CryptoPaperTrading is a simulator using virtual money for education, and the figures in this guide are illustrative examples only.
Practise this: Practise in the crypto trading simulator with $10,000 in virtual USD.