Home › Learn › Lesson 4

Scalp, day, swing or position — which kind of trader are you

By TradingPrimer · Published 2 Sep 2026

Almost every article on trading styles asks you to look inward. Are you patient or impulsive? Do you like fast decisions? Can you sleep with a position open? Those are real questions, but they are the second set of questions, and answering them first is how beginners end up in the style that is arithmetically hardest to win in. This lesson does it the other way round. It works out what each style costs you before you have made a single good or bad decision — and finds, among other things, that a scalper needs to win 55.6% of two-to-one trades just to break even where a longer-term trader needs 33.7%, that the cheapest style overall is not the slowest one, and that a patient investor holding a perpetual contract can end up paying more friction than a day trader.

The stop is not a percentage you pick. It goes under the last low your chart can see — and which low that is depends on the frame. Each step up the frame ladder is about sixteen times the elapsed time and so about four times the distance.

What actually separates the four trading styles?

How far away you put your stop-loss. Everything the textbooks list as a defining feature — the timeframe, the holding period, the leverage, the number of trades — turns out to be a consequence of that one distance rather than an independent choice.

Start with the conventional description, because it is accurate as far as it goes. The course this site is built from divides market participants into four:

This is one school's division. Other courses draw the lines in different places — some split "day trading" from "intraday swing", some treat position trading as investing rather than trading. The boundaries are conventions, not facts about the market.

Read that table again and notice something odd: three of its five columns are about time, and none of them is about money. Yet the two things a beginner actually has to budget are money and hours. So let us replace the table with a single number and derive the rest.

The number is the stop distance, written s, expressed as a percentage of your entry price. A stop-loss is the price at which you accept the idea was wrong and get out. If you buy at $60,000 and your stop is at $59,640, your stop distance is $360, or 0.60%.

Where does that number come from? Not from picking a percentage you like. A stop goes at the price that says your reason for being in the trade is gone — and in practice, for a long, that means just below the last low the market made before you bought. If price takes out that low, the sequence of higher lows you entered on is broken, and the trade is wrong for a reason you can point at. Lesson 12 on support and resistance and Lesson 15 on higher highs and lower lows are about finding those points; here we only need the consequence.

And the consequence is this: which low you can see depends on the chart you are looking at. The same rally, viewed on three frames, offers three different "last lows" at three different depths — a shallow one a few candles back, a deeper pullback further back, a much deeper one at the start of the move. The stop is not a percentage you chose; it is a distance the chart handed you.

Illustrative prices built for this lesson, not a market snapshot. What matters is the ratio between the rows, not the exact figures.

Why do those depths quadruple rather than, say, double? Because price moves further as time passes, and it does so at a predictable rate: over a period of length t, the typical size of a move scales with the square root of t, not with t itself. Four times the time gives you roughly twice the range; sixteen times the time gives roughly four times the range. Each step up the frame ladder — 1m to 15m, 15m to 4H — is about sixteen times the elapsed time, so the last low on that frame sits about four times further away. That is why the four styles land on a ladder that quadruples:

0.15%  →  0.60%  →  2.40%  →  9.60%

Those four numbers are a working model for this lesson, not a measurement of any particular market. Round numbers four times apart are easy to check the arithmetic against, and they sit in the right neighbourhood for the frames each style uses. Your own stops will differ. The identities below do not depend on the exact values — they depend on the ratios, and those are set by the square-root rule, not by us.

Two more pieces of vocabulary and we can do all the arithmetic in the rest of this lesson.

• R is the money you lose if the stop is hit. If you risk 1% of a $10,000 account, R = $100. Everything below is measured in R, which makes it true for any account size.
• Notional is the face value of the position — the number the exchange charges fees on. It is not the money you put up. Buying $4,167 of Bitcoin is a notional of $4,167 whether you paid for all of it or posted $500 of margin against it.

Why does a scalper need leverage when a position trader does not?

Because the position size is not chosen — it is computed. Once you fix how much you are willing to lose and how far away the stop is, the size of the position is already decided, and nobody gets a vote.

The relationship is one line:

Notional = R ÷ s

If you are willing to lose $100 and the stop is 1% away, a 1% loss on the position has to equal $100, so the position must be $10,000. Halve the stop to 0.5% and the position must double to $20,000 to keep the same $100 at risk. This is the whole of position sizing, and Lesson 42 takes it apart properly. Here we only need what it implies about the four styles.

Take a $10,000 account and a 1% risk rule, so R = $100:

Read the right-hand column as the answer to a question nobody asks out loud. The scalper is not using 6.67× leverage because they are reckless; they are using it because a $66,667 position will not fit inside a $10,000 account and there is no other way to risk exactly $100 behind a 0.15% stop. Meanwhile the position trader is holding a tenth of their account in notional. Leverage is not merely unnecessary there — it is arithmetically absent. The course's observation that "position traders almost never use leverage" is not a statement about their temperament. It falls out of the sum.

This reframing matters because it reverses the order of two decisions that beginners usually make backwards. The dangerous sequence is: choose leverage first, then find a stop that fits. The safe sequence is: choose the risk and the stop, then accept whatever leverage the arithmetic hands you. In the second sequence, leverage is a readout. In the first, it is a lever, and it is the lever that empties accounts. Lesson 8 works through what happens when the liquidation price arrives before the stop does.

What does each style hand to the exchange on every trade?

An amount that is fixed as a share of your risk, not of your account — and it is the fee rate divided by the stop distance. That single fraction is the reason the four styles are not equally easy.

Fees are charged on notional. So the fee you pay is f × notional, where f is your round-trip friction — the spread you cross plus a taker fee on the way in and another on the way out. Substituting the notional identity from the previous section:

fee = f × (R ÷ s)  ⇒  fee measured in R  =  c = f ÷ s

Stop and stare at that for a second, because it says something that is not obvious. Your account size has vanished. Your risk percentage has vanished. The cost of trading, expressed as a share of the money you are risking, depends on exactly two things: the venue's fee schedule and how far away you put your stop. A $500 account and a $5m account pay the same c.

Use 0.05% per side, which was a typical retail taker tier on the major venues as of August 2026 — check your own, the tiers differ and yours is the one that matters. That gives f = 0.10% for a round trip, ignoring spread for the moment:

The break-even column comes from a small identity worth keeping. If a winner pays k R and you pay c in fees whether you win or lose, then a winner nets k − c and a loser costs 1 + c, and setting the two sides equal gives:

break-even win rate = (1 + c) ÷ (k + 1)

With no fees and a 2R target that is the familiar 1÷3. The fee does not add a little to it — it multiplies it by (1 + c). For the scalper, (1 + 0.667) means the bar they have to clear is 66.7% higher than the frictionless one. That is the whole gap between 33.3% and 55.6%, and it exists before we say a single word about whether their analysis is any good.

Here is the same fact in money rather than percentages. Take an ordinary trader — 45% win rate, 2R targets — which is a perfectly respectable, unglamorous edge. Gross, they make 0.45×2 − 0.55×1 = +0.35R per trade. Now subtract c:

The same person, with the same skill, is profitable in three styles and losing money in the fourth. And the size of the gap is easy to under-feel when it is written as a decimal, so here it is as a question instead: how good would the scalper have to be to end up where the swing trader ends up? Set 3p − 1 − 0.667 equal to the swing trader's +0.308R and solve: p = 65.8%. The scalper must win nearly two trades in three to match somebody winning fewer than one in two.

None of this says scalping cannot work. It says scalping is a strategy for people who have already solved the fee problem — by earning a maker rebate, by trading a venue tier ordinary retail accounts do not get, or by having an edge large enough to pay 0.667R a trade and still have something left. It is a bad first style precisely because it demands the largest edge from the person with the least.

Which style is really the cheapest once you count funding?

Swing — and the answer is genuinely surprising, because the style that looks cheapest on the fee table, position trading, is the one most exposed to the cost we have not counted yet.

Everything above is per-trade cost. But on a perpetual futures contract there is a second meter running, and it charges by the hour rather than by the trade. The funding rate is a periodic payment between longs and shorts that keeps the contract tethered to spot. The baseline on major venues is 0.01% every eight hours — three settlements a day, so 0.03% of notional per day, as of August 2026. It can be much higher, and it can be negative, in which case you are paid; the baseline is what we can reason about.

Funding is charged on notional too, so it goes through the same conversion:

funding per day, in R = 0.03% ÷ s

And now the finding that makes this section worth reading. Ask how many days of funding it takes to equal one full round trip of fees. Set 0.03% × D ÷ s equal to 0.10% ÷ s — and watch the stop distance cancel from both sides:

D = 0.10% ÷ 0.03% = 3.33 days

Every three and a third days you hold a perpetual, at baseline funding, you pay again what it cost you to open and close it. That number does not care about your stop, your style, your account or your leverage. A scalper and a pension fund holding the same contract accrue the same number of round-trips-per-week. It is one of the very few things in trading that is genuinely universal, and almost nobody who chooses a style ever computes it.

Now put both meters on the same table. Holding periods are the middle of each style's stated range: five minutes, four hours, five days, ninety days.

Three things fall out of that table, and each of them contradicts something people say confidently.

First, friction is U-shaped, not a slope. It is worst at the fast end because fees are per-trade, worst at the slow end because funding is per-day, and lowest in the middle. Swing trading is not a compromise between two better options; on this arithmetic it is the cheapest place on the curve.

Second, the patient trader on the wrong instrument pays more than the impatient one. A 90-day perpetual position costs 0.292R against a day trader's 0.175R. Ninety days at 0.03% is 27 round trips' worth of fees on a position that was opened and closed once. The trader believes they are being patient and low-cost; the meter disagrees.

Third, and this is the actionable part: for a long hold, the instrument matters more than the style. Same analysis, same entry, same stop, same 90 days — 0.292R on a perpetual against 0.010R on spot. A factor of 28, decided by which tab you opened. At 45% and 2R, that is the difference between netting +0.058R and +0.340R per trade: funding quietly eats 83% of the position trader's edge. Spot has its own costs — your capital is fully committed rather than partly, and shorting is harder — but if you intend to hold for months, the funding meter is the first thing to check, not the last.

How many hours a day does each style need from you?

Fewer than you think for the slow styles, and more than any employed person has for the fast ones — and the honest way to see it is to count decision points rather than hours.

A timeframe is a decision generator. Every time a candle closes on the chart you trade, you are being invited to have an opinion. Crypto runs 24 hours a day, seven days a week, so those invitations do not stop at 4pm the way they do in equities. Count them for a full year:

The last column is the point. Three hours a day is three hours a day whichever chart you open — but on any intraday frame it buys you 12.5% of the decisions, and on the daily chart it buys you all of them, because there is only one close per day and you can choose to be there for it. The 1-minute chart closes 525,600 times a year against the weekly chart's 52 — a ratio of 10,080 to 1. Your availability is not a fixed handicap; it is a handicap in proportion to how many decisions your chosen frame manufactures.

This is also where one traditional argument for scalping needs scoping. The classic case — taught in equities and foreign exchange, and repeated in crypto courses including the one this site learned from — is that closing every position intraday removes the risk of holding overnight. In a market that shuts at a bell and reopens with a gap, that is a serious advantage. Crypto never shuts. There is no overnight session and therefore no opening gap; a violent move can arrive at 3am on a Sunday exactly as easily as at 3pm on a Tuesday. What the scalper actually avoids is holding through hours they cannot watch — which is a real benefit, but a much smaller and differently-shaped one than the version imported from the stock market.

And the swing trader's answer to the same problem is not to be awake more. It is the 2.40% stop. A wide stop is what lets a position survive the hours you are not there; a 0.15% stop cannot survive them, which is why the fast styles require presence and the slow ones require patience. Once again the two features are the same feature seen from different ends.

What happens if you pick one style and trade another?

You pay the fast style's costs while chasing the slow style's targets, and the bill multiplies by exactly the ratio between the two stop distances.

This is the single most common way a chosen style goes wrong, and it does not look like a mistake while you are making it. You analyse on the four-hour chart, you place a swing trade with a 2.40% stop and a target 4.80% away, and then — because the position is open and you are anxious — you watch it on the 15-minute chart. Over a five-day hold, that chart will close 480 times. Each close is an invitation. Sooner or later you accept one, and you exit on a 0.60%-scale wobble.

What did that cost? Go back to c = f ÷ s. You planned around s = 2.40% but you acted on s = 0.60%, so your friction is multiplied by the ratio of the two:

friction multiplier = s_planned ÷ s_{acted on} = 2.40 ÷ 0.60 = 4×

Your c goes from 0.042R to 0.167R and your break-even from 34.7% to 38.9%. But the fee is the smaller half of the damage. The target was 4.80% away and your effective leash is now 0.60% — one eighth of the distance. The trade you designed cannot reach its target inside the behaviour you are actually exhibiting. You have not made the swing strategy riskier; you have replaced it with a worse day-trading strategy while keeping the swing strategy's paperwork.

The mirror-image error is worth naming too, because it is just as common and this site has already priced it. In Lesson 21 we worked out what happens when a trader borrows a stop from a higher frame than the one they entered on: the stop widens by the square root of the frame ratio — 4.90× going from 1H to daily — and a 2.00R trade collapses to 0.41R. Two errors, opposite directions, one cause. Acting on a shorter frame than you planned multiplies your costs; borrowing structure from a longer frame than you planned destroys your reward. In both cases the damage comes from mixing two frames, not from either frame being wrong.

Which is why "pick a style" is not soft advice about self-knowledge. It is a request to commit to a single value of s and then to look at the chart that value belongs to, so that the number you planned with and the number you act on are the same number.

So how do you choose, in four steps?

Work backwards from the two things you cannot change by wanting to: what your venue charges you, and how many hours you can actually sit down. The style is whatever is left standing.

Step 1 — measure your real round-trip friction. Not the headline fee: the spread on the pairs you trade plus your taker fee twice over. On a major pair with a 0.01% spread and a 0.05% taker tier, f = 0.11%. On a large-cap altcoin with a 0.05% spread, f = 0.15%. On something thin, it can be 1.5% of spread alone. Lesson 9 shows how to read that off an order book in about a minute.

Step 2 — decide what share of your risk you are willing to hand over. Call it your fee budget, c_max. Ten percent is a defensible line: it means one trade in ten is being paid to the exchange. Then rearrange c = f ÷ s:

minimum stop = f ÷ c_max

With f = 0.11% and a 10% budget, your minimum stop is 1.10%. That single line has just eliminated scalping (0.15%) and day trading (0.60%) from your list, on arithmetic, before personality entered the room. Loosen the budget to 25% and the minimum stop falls to 0.44%, which puts day trading back on the table — at the price of a quarter of your risk going to the venue on every trade. That is the trade-off, stated in a form you can argue with.

Step 3 — check the hours against the frame. Convert your minimum stop back into a frame with the quadrupling ladder, then look up how many closes that frame produces and how many of them your daily window contains. If the answer is 12.5%, you are choosing to be absent for seven of every eight decisions your own method generates. That is survivable on a 4-hour chart with a wide stop and it is not survivable on a 5-minute chart with a tight one.

Step 4 — if the hold is longer than three days, check the instrument before the strategy. Above 3.33 days on a perpetual you are into repeat round trips, and by 90 days you are paying twenty-seven of them. Ask whether the trade needs leverage at all. If the honest answer is no — and at a 9.60% stop the arithmetic in section 2 says you are holding a tenth of your account in notional — then spot removes the entire funding meter for free.

Only now is it worth asking whether you like fast decisions. Temperament decides between the options the arithmetic leaves you; it does not get to overrule the arithmetic. A style you cannot afford has a negative expectancy no matter how much it suits you, and a style you can afford but cannot sit still for has a negative expectancy too. You need both boxes ticked, and the money box is the one you can check today.

When is this advice wrong?

In four identifiable situations, and it is worth knowing which one you are in before you take the table above as a verdict on your style.

When your fee schedule is not the one we assumed. Everything here is a function of f, and f is not a law of nature. If you trade as a maker at 0.02% per side rather than a taker at 0.05%, f falls from 0.10% to 0.04% and the scalper's c drops from 0.667R to 0.267R — break-even from 55.6% to 42.2%. That is 13.4 percentage points of win rate bought purely by using resting limit orders and a better tier. For a fast trader, the fee schedule is a larger lever than the strategy, and any argument about scalping that does not name a fee tier is an argument about nothing.

When the spread is the real cost and we only counted fees. Our headline table used f = 0.10%, which is fees alone. Add the spread and the fast end deteriorates quickly. On BTC/USDT with a 0.01% spread the scalper's break-even goes to 57.8%; on a large-cap altcoin at 0.05% it is 66.7%; on a thin small-cap where round-trip friction runs to 1.6%, c becomes 10.67R — larger than the 2R target, which means the trade loses money at a 100% win rate. The general condition is worth memorising: a style is arithmetically impossible whenever friction ≥ k × s. Scalping in crypto is therefore a BTC and ETH activity, not a matter of nerve.

When funding is not at baseline. The 3.33-day identity uses 0.01% per eight hours. Funding is a market price: it spikes when one side is crowded and it goes negative when the other is, in which case a long is being paid to hold. Venues also shorten the settlement interval when the rate hits its cap, which compresses a day's worth of funding into a few hours. So treat 3.33 days as a baseline, check the live rate before a long hold, and never assume the cost is symmetric between longs and shorts.

When you cannot execute the style you can afford. The arithmetic tells you which styles are survivable; it cannot tell you which one you will follow at 2am after two losses. A cheap style you abandon has worse expectancy than an expensive style you execute. That is the honest limit of everything above, and it is why Lesson 1 puts process before instruments. Use the cost table to shorten the list, not to pick the winner.

Common mistakes

Frequently asked questions

Which trading style is best for beginners? The one whose costs your account can absorb and whose chart you can actually be present for — which for most people with a job lands on swing trading, and the reasoning is arithmetic rather than taste. Swing has the lowest total friction on our table at 0.104R per trade against the scalper's 0.667R, and its 2.40% stop is wide enough to survive the hours you are asleep. Just as importantly, a 4-hour chart closes 2,190 times a year against the 1-minute chart's 525,600, so a beginner's inevitable early mistakes are made a few times a week instead of a few times an hour. That matters more than it sounds: mistakes at a low rate are learnable, and mistakes at a high rate are just a drawdown.

Is scalping profitable for retail traders? It can be, but it demands the largest edge from the person with the smallest one, which is why it is a poor first choice. At a standard 0.05% taker tier and a 0.15% stop, fees alone come to 0.667R per trade, so a 2R setup needs a 55.6% win rate to break even against the 33.3% a costless trader would need. Add a typical BTC spread and it is 57.8%; on a large-cap altcoin it is 66.7%; on a thin coin with 1.6% round-trip friction the trade cannot be profitable at any win rate at all. The people who make scalping work have usually solved the fee problem first — maker rebates, a volume tier, the deepest pairs only — rather than solved the chart-reading problem harder.

How long should I hold a crypto position? That depends on the instrument as much as the analysis. On spot there is no holding cost, so the honest answer is: as long as the reason you entered is still true. On a perpetual there is a meter running, and at baseline funding of 0.01% every eight hours, every 3.33 days of holding costs you the equivalent of one full round trip of fees — regardless of your stop, style or account size. A 90-day hold is 27 round trips. So a long-horizon idea belongs on spot unless you have a specific reason to need leverage, and a perpetual position that has drifted past a week without a plan is quietly being taxed.

Does swing trading need less time than day trading? Much less, and the reason is countable rather than vague. Day trading on a 15-minute chart generates 35,040 candle closes a year, of which a three-hour daily window lets you see 12.5%. Swing trading on the 4-hour chart generates 2,190, and on the daily chart just 365 — and one well-chosen daily check covers every daily close. So the same three hours buys you a small fraction of an intraday method and complete coverage of a daily one. What swing trading demands instead is tolerance: holding through moves you have decided in advance not to react to, which many people find harder than watching a screen.

Do I have to use leverage to day trade or scalp? At a fixed risk percentage, effectively yes, and that is the clearest illustration of why leverage is arithmetic rather than attitude. Risking 1% of a $10,000 account behind a 0.15% stop requires a $66,667 position, which is 6.67× the account; behind a 0.60% stop it requires $16,667, or 1.67×. The same rule behind a 9.60% position-trading stop requires $1,042 — a tenth of the account, no leverage anywhere. If you refuse leverage while keeping a tight stop, you are not reducing risk; you are silently cutting your risk per trade to a fraction of a percent, which is a legitimate choice but a different one from the plan you wrote down.

How these numbers were produced. Three identities do all the work and each is stated in the text where it is used. First, notional = R ÷ s, which follows directly from the definition of a stop: a loss of s percent on the position must equal R. Second, cost measured in R is c = f ÷ s, because both fees and funding are charged on notional; this is why account size and risk percentage cancel out of every cost figure on this page. Third, break-even win rate is (1 + c) ÷ (k + 1), from setting p(k − c) = (1 − p)(1 + c). The stop ladder 0.15 / 0.60 / 2.40 / 9.60% is an illustrative model chosen to quadruple, following the square-root-of-time scaling of price ranges; it is not a measurement of any market and your own stops will differ, but the ratios between the four styles are what the conclusions rest on. Taker fees are assumed at 0.05% per side and funding at 0.01% per eight hours, levels typical of major venues' standard retail tiers as of August 2026 — verify against your own venue's published schedule, since every figure here scales directly with them. Spread figures are the illustrative band from our liquidity lesson, not a live measurement. Candle-close counts assume a 24/7 market and are exact. Holding periods are the midpoints of each style's conventional range. No win rate, success rate or profitability statistic for any trading style is quoted anywhere on this page, because we have not measured one and the figures circulated for these styles cannot be reproduced without knowing their sample. The four-way division of participants, the timeframes and the observation that leverage falls as the holding period rises follow the slide course this site learned them from. Sources: our own arithmetic, stated inline. Published 2 Sep 2026.