A short-term strategy may seek only a few pips of movement, making repeated costs a large share of that target. Start with the full round-trip calculation for the intended size and trading window. A small cost difference can change a thin positive result into a loss, but the effect depends on the strategy’s distribution of wins, losses and unfilled orders. Longer-term traders must include costs too.

Why scalping lives and dies on costs

In a hypothetical comparison, a two-pip round-trip cost is 40% of a five-pip target and 2% of a hundred-pip target. The ratio is twenty times larger in the first case. These are shares of a target, not expected profits: losing trades and missed executions still matter. A larger target makes the relative charge smaller without making it disappear.

Frequency multiplies a constant cost difference. Assume fifty completed trades per day for twenty days: that is 1,000 round trips. A saving on each identical trade multiplies by 1,000 too. This is hypothetical arithmetic, not a recommended frequency or evidence that fees rather than market losses explain most scalpers’ results.

A tight spread with commission versus a wider spread alone

One pricing model combines a wider spread with no separate trading commission; another combines a narrower spread with an explicit charge. Names such as Raw or ECN account do not establish either the legal counterparty or the total cost. Read the actual schedule, including minimum fees and whether a quoted commission covers one side or a round trip.

For this simplified round-trip comparison, spread cost plus commission = spread in pips × pip value + round-turn commission. Assume a constant spread at entry and exit, no slippage and one EUR/USD lot of EUR 100,000: a pip of 0.0001 is USD 10. Financing, conversion and other charges remain separate. If the result already uses actual ask/bid fill prices, the spread is embedded and must not be deducted again.

Let me use a hypothetical comparison — the figures are illustrative and meant to show the mechanism, so check your own rates with your broker.

Cost of one EUR/USD trade, 1 lot (hypothetical example)
Standard accountspread 1.2 pips × 10 USD = 12 USD, commission 0 USD → 12 USD
Raw/ECN accountspread 0.1 pips × 10 USD = 1 USD, commission 7 USD → 8 USD
Difference4 USD in favour of the Raw account

In this model the Raw account is USD 4 cheaper per identical round trip. If all fifty daily trades had that size and cost difference, the saving would be USD 200 per day or USD 4,000 over twenty days. Neither rate is an observed result. Different spreads, commissions, minimum fees and fills can reverse the ranking at any size. The article on spread versus commission shows how to compare the same transaction under both schedules.

Small differences in full round-trip cost accumulate with repeated trades. Compare spread, commission, minimum fees and execution for the same size.

Execution, latency and slippage — the scalper's second front

Measure slippage against a stated, executable reference: ask for a buy or bid for a sale at a specified time. A hypothetical adverse half-pip difference is 10% of a five-pip target; a favourable difference has the opposite effect. In MT5 Instant Execution a broker can return a new quotation when it declines the requested price. Such a requote changes the execution choice, rather than guaranteeing a lost entry.

Check the execution mode and fill policy for the actual symbol. Market Execution in MT5 permits the broker to determine the fill price without another quotation; it does not ensure acceptance or eliminate slippage. IC Markets (EU) states that its MetaTrader Raw servers are in Equinix NY4. That describes infrastructure, not measured price improvement on your account. Compare latency separately from fills and missed orders. See order execution time, the ECN versus market maker comparison and ForexMechanics on choosing a broker.

The broker rulebook — when scalping is restricted

Read the applicable agreement for automation, order constraints, holding periods, news trading and definitions of prohibited practices. A label such as ECN or market maker does not establish these permissions. Ordinary short-term trading is not automatically the same as exploiting erroneous or delayed quotations. Ask for the specific clause relevant to your intended method.

A cancellation clause is not proof that every withheld profit is justified. Its application depends on the agreement, circumstances and governing law. Preserve order history, logs and correspondence if a dispute arises. Compare the actual contracting entity and execution policy; a broker may remain your counterparty even when it hedges externally.

Typical cost and an unusually expensive fill

A single average does not reveal whether costs are evenly distributed. In this small set, the middle value looks much milder than the mean.

Scroll sideways to see the full diagram

Typical cost and an unusually expensive fillHypothetical total costs: 0.2, 0.2, 0.2, 0.2 and 3.2 pips. Total 4 pips, mean 0.8, median 0.2; the last cost represents 80% of the total.Five round-trip costs0.2 p#10.2 p#20.2 p#30.2 p#43.2 p#5Median: 0.2 pip Mean: 0.8 pipMost expensive fill: 80% of total
Entirely fictional data: five equally weighted trades with the same volume and pip value. Full costs use one common unit. This illustrates a distribution; it is not a broker measurement, a frequency estimate or a forecast.
One large cost can dominate the total despite a low median.

For your own comparison, record the median alongside the mean, the most expensive fills and the conditions in which they occurred. Distinguish round-trip cost from the displayed spread alone. Five examples explain the arithmetic but cannot establish service consistency.

How much cost a scalper really sees

Spread and commission do not cover the whole bill. Financing follows the product’s rollover cut-off, so opening and closing on the same calendar date does not itself establish zero financing. Check conversion, payment and inactivity charges separately. A spread can change during the trading window; a minimum advertised rate is not the cost of every fill.

So a scalper's real cost is not one number from an advert, but a distribution: how much you pay in a calm London session, and how much at three in the morning or a minute after an inflation print. I cover the full set of items that make up the bill separately in the piece on the real costs of trading. Without that map it is easy to pick an account that looks cheap in the brochure and turns out expensive in your specific trading window.

What to do before you open a scalping account

  1. Compute the cost per trade for your pair. Put the spread and commission into the formula and compare several accounts in one currency — the total round-turn cost. Check the minimum commission per order at the same time, because that is what spoils the maths on micro-positions.
  2. Separate functionality from live execution evidence. Demo can test order handling and automation, but read what it simulates. IG’s UK demo documentation says it omits slippage and size/price rejections. A successful demo order does not prove live fill quality.
  3. Read the terms for scalping and EAs. Look for clauses on a minimum holding time, news trading and automated trading. If you plan an Expert Advisor or news scalping, make sure the account you choose allows it.
  4. Identify protection for your product and status. Check the entity and applicable national retail CFD measure. Negative-balance protection limits covered account liabilities; it does not prevent losing the funds in that account.