Scalping is the strategy most often described and least often described accurately. The popular version is a fast trader clicking rapidly; the mechanical version is an order-placement policy that tries to be paid the spread rather than pay it, repeated many times at small size.

What follows is how the mechanism is constructed, the conditions it depends on, and the cost structure that determines whether any of it survives contact with an exchange's fee schedule.

Being Paid the Spread Instead of Paying It

Every order is one of two kinds. A limit order rests in the book at a price the trader names and waits for someone else to cross to it; a market order crosses immediately to whatever is already resting. The first adds liquidity to the book and is called a maker; the second removes it and is called a taker.

That distinction is the whole basis of the strategy, because most exchanges price the two differently. A maker fill is cheaper than a taker fill, sometimes materially, and on some venues a maker is paid a rebate rather than charged at all. A strategy that only ever rests orders therefore starts each trade from a better cost position than one that crosses.

So the mechanism is: rest a buy just below the current mid-price, wait for the market to come down to it, and on a fill place a resting sell above. Both legs are maker orders. The trade earns the difference between the two resting prices, minus whatever the venue charges.

Size stays small and deliberately so. The strategy is not trying to be right about direction; it is trying to be present in the book often enough that ordinary two-sided movement fills both legs. Frequency substitutes for magnitude.

Calm, Liquid, and Two-Sided

The conditions this depends on are specific. The book has to be liquid enough that a resting order actually gets filled rather than sitting untouched, and the spread has to be stable enough that the two legs can be priced relative to each other in advance.

Price also has to move in both directions. A market oscillating inside a range crosses a resting buy and later a resting sell, filling both legs of many trades. That is the environment the strategy was built for.

A one-directional market is the opposite. If price falls steadily, resting buys fill and resting sells do not, and the strategy accumulates inventory it cannot exit at its target. The positions are not wrong in any analytical sense; they are simply unfinished, and they stay unfinished for as long as the direction persists.

This is why the same strategy behaves so differently across sessions without changing at all. Its edge is conditional on the market being two-sided, and nothing inside the strategy can make that true.

The Fee Floor Is the Whole Constraint

A round-trip means two fills, so it pays fees twice. On a retail crypto venue that combined cost commonly lands somewhere under one percent of notional, and the exact figure depends on the venue, the order type and the account's volume tier. The number matters less than its consequence.

The consequence is a floor. Any target that would produce a gross gain smaller than the round-trip cost produces a net loss, however often it fills. A strategy taking many tiny round-trips is unusually exposed to this, because the smaller the intended move, the larger the fee is as a proportion of it.

The disciplined construction sets each sell target at the entry price plus the desired net gain plus the full round-trip cost, so a fill is arithmetically net-positive before it happens. That single rule is what separates a scalping strategy from a machine that trades constantly and loses slowly.

It also explains why maker-only execution is not a stylistic preference here. Paying the taker fee on both legs can move the floor above the size of move the strategy is trying to capture, at which point the strategy has no edge left to express.

Where It Breaks Down

The clearest failure is the trending market described above: resting buys fill, resting sells do not, and inventory builds. A timeout or a stop bounds how long a position can stay unfinished, which converts an indefinite hold into a small realised loss. That is a deliberate trade — many small certain losses in exchange for not carrying an unbounded one.

A subtler failure is adverse selection. A resting order is, by construction, filled by someone who wanted to trade at that price right then. When the market is about to move, the orders that get filled first are the ones on the wrong side of it. A maker is compensated for that risk by the spread and the fee difference; when the spread narrows or volatility rises, the compensation stops covering the risk.

The third is competition. Resting at a good price only works if the order is near the front of the queue, and on a liquid pair the front of the queue is contested by participants with faster infrastructure. A strategy resting behind them fills later and on worse terms.

Scalping is often presented as a speed contest. Mechanically it is a cost contest: a policy for being paid the spread, constrained by a fee floor that decides in advance which trades are worth taking at all.

Note: This explains how a process works. It is not legal advice, it is not specific to any debt, and it is not a substitute for a licensed attorney in your state. Rules and time limits vary by state and change over time — check the cited sources.