3 / 6 · 8 min
Inside the day: scalping and day trading
Both styles close before the day ends and therefore never pay for time. But their demands differ, and the main difference is not speed — it is what you pay the exchange.
What these two styles share
Both the scalper and the day trader close their position the same day. Two shared properties follow. First, they never pay funding — settlement happens three times a day and they are already out. Second, they carry no overnight risk: news at four in the morning finds them out of the market. They pay for this in other ways: number of trades, attention, and having to close by the clock rather than by the idea.
How they actually differ
| Scalping | Day trading | |
|---|---|---|
| Position lives | seconds and minutes | hours |
| Trades per day | tens | 1–5 |
| What you pay | maker fees | taker fees, and that is fine |
| What is critical | execution and spread | choosing the moment |
| Instrument | only the most liquid | almost any liquid one |
The main arithmetic: taker versus maker
This distinction matters more than all the others and is barely discussed. A market order pays the taker fee — around 0.05% on major venues, and you give up the spread as well. A limit order that sat in the book and got filled pays the maker fee — about 0.02%, and instead of giving up the spread you collect it. A round trip at market costs roughly 0.1% plus spread; a round trip on limits costs about 0.04%. Two and a half times, and on a short horizon that decides everything.
The same entry, two ways of executing it
A coin with a typical five-minute travel of 0.5% — the median across 167 liquid coins on 09.08.2026. Scalping at market: costs 0.1%, headroom 5. Scalping with limit orders: costs 0.04%, headroom 12.5. At thirty trades a day the first way eats 3% of the account daily, the second 1.2%. Over a month the difference runs into tens of percent, and it comes not from markup but from which button you enter with.
A limit entry has its own price: it does not always fill, and part of the moves leave without you. That is the scalper's real choice — not where to enter but whether you are willing to miss trades in order to pay half as much. Whoever enters at market thirty times a day pays the exchange more than the markup earns, and no amount of entry accuracy covers it.
How many instruments actually qualify
A measurement across 167 liquid coins shows on how many of them a five-minute horizon makes sense at 0.1% costs. With headroom of at least three — 100 coins, that is 60%. With headroom of five — 84, exactly half. With headroom of ten — 43, a quarter of the list. On an hourly horizon the picture differs: 106 of 167 travel at least one percent. That is the price of a short horizon: the shorter it is, the fewer instruments where it works at all.
Turning an intraday trade into an overnight one
The most expensive mistake of both styles, and it always looks the same: the position is down, the day is nearly over, let it sit and I will close in the morning. At that moment the trade changes style but not size — and the size was computed for a horizon of hours, not days. By morning funding and overnight news you were not watching have been added to the loss. The tell is simple: if the decision to hold appeared AFTER entry, it is not a change of style but unwillingness to take a loss.
Day trading: freedom from seconds, obligation to close
Day trading requires neither maker fees nor constant attention: five trades a day at 0.1% cost half a percent, tolerable against an hourly travel of one or two percent. What it does require is exit discipline. A scalper's position closes itself — the market either gave or took within minutes. A day trader has the whole day to talk himself into waiting a bit longer, and that is where the style breaks most often. The practical rule is simple: the closing time is set BEFORE entry, alongside the stop.
Scalping where the spread is wide
A coin turning over a hundred thousand dollars a day can show excellent five-minute travel and still be unusable. The spread there reaches half a percent, and a thousand-dollar order moves price by itself. Headroom computed from fees turns out two or three times smaller in reality, and you will not see it in the statement: slippage is not a separate line, it simply makes your fill worse. Before scalping, look in the order book, not only at the chart.
The screener shows both turnover and NATR across windows. Filter to coins turning over more than ten million and sort by 5-minute NATR — the top of that list is where a short horizon has headroom. Then open the order book on a couple of them: the width of the spread will tell you more about suitability than the beauty of the chart.
About volatility and NATRCompute your month in fees
Take your exchange statement for the last month and write down two sums: how much you paid in fees and what your net result was. Divide the first by the absolute value of the second. If fees are more than a third of the result, your horizon is shorter than your entry method allows — and the cheapest cure is switching to limit orders, not hunting for a new strategy. While you are there, check what share of your entries were at market: for most people it is higher than they remember.
Why is scalping at market almost always unprofitable?
Because of arithmetic, not entry quality. A round trip at market costs about 0.1% plus spread, and at thirty trades a day that is roughly 3% of the account every day. Beating that requires a very high win rate — noticeably higher than any markup delivers on five-minute noise. A limit entry cuts the round trip to about 0.04% and the bar immediately becomes reachable. That is why real scalpers talk about execution rather than signals.
An intraday position is down as the day ends. Close it or keep it overnight?
Close it, and the question is about style rather than the market. You sized the position for a horizon of hours: the stop and the acceptable loss were chosen for that. Keeping it overnight gives you the same size on a different horizon, where funding, night volatility and the impossibility of intervening are added. If you genuinely have a reason to hold longer, the correct move is to close and reopen at swing size with a recomputed stop. It looks like a pointless extra action, but it is exactly what separates a changed decision from an unwillingness to take a loss.