Answer

The answer: compare brokers and accounts on normalized all-in cost

The foundation of comparing several brokers or account types fairly is not lining up advertised headline spreads. Normalize to the same instrument, quantity, time and spread basis, order type, holding period and account currency, then compare the all-in cost that adds average spread, commission, minimum fee, a slippage assumption, swap/funding and conversion together. Line up the monthly cost (per-trade cost times monthly trade count) and the cost ratio against an assumed gross target as well, and the difference in burden between account types becomes clear.

Normalizing matters because account types differ in the shape of their cost. A raw-type account stacks a round-turn commission onto a tight spread; a standard-type account folds cost into a wider spread with no commission, so the same trade has a different breakdown. If you want the full build-up of trading cost first, start with the pillar, the Trading Cost Calculation Guide covering spread, commission, swap and break-even, which shows which parts this article is summing. The individual components are broken out across converting spread pips into money, turning commission into all-in cost and swap and overnight financing. This article focuses on placing those parts side by side under identical conditions.

Basics

Terms and assumptions: headline vs average spread, commission modes, swap and conversion

Before comparing, align the terms and fictional assumptions used throughout. Every figure below is a fictional educational example, not real broker pricing, real fills or the contract terms of a specific account. Actual spread, commission, minimum fee, swap/funding and execution policy vary by broker, account, instrument, jurisdiction and time, so always confirm each provider’s official documentation.

• Headline spread: the tightest-instant spread shown in advertising. It can differ from the realized average spread.
• Average / typical spread: the average spread observed in the session where you actually place orders. Use this one for comparison.
• Commission: a fee charged per trade. It comes in modes (per side vs round-turn, per-lot vs fixed), so confirm the charging unit to avoid double counting.
• Minimum fee: the floor on the fee per trade. At small size the minimum fee dominates the proportional part.
• Slippage assumption: the expected gap between intended and realized price. It shifts with the session and volatility, so treat it as a sensitivity input (see calculating slippage as execution cost).
• Swap/funding: the receipt or payment for holding a position overnight. Negative swap raises cost in proportion to holding days.
• Conversion: the FX conversion when the profit/loss currency differs from the account currency. Where an account charges conversion, add it as a separate stage.

The worked examples reuse one consistent fictional dataset. The instrument is USD/JPY (1 pip = 0.01; at 1.0 lot, 1 pip = ¥1,000), the comparison quantity is 1.0 lot, and the account currency is JPY with P/L also in JPY, so conversion is 0. The monthly trade count is 40 round-turns, and the assumed monthly gross target that forms the cost-ratio denominator is ¥300,000. Every table, figure and mini-calculator default below is aligned to these assumptions. Choosing the lot size itself and required margin are out of scope here and are covered in calculating forex lot size.

Fair basis

The six fair-comparison conditions: misalign one and the ranking breaks

Before comparing, align the following six conditions. If even one differs between accounts, you compare a difference in assumptions rather than cost, and the ranking breaks. They are distinguished by number and label, not color alone.

These six conditions are the minimum set that puts a comparison on the same footing. Condition 3 (time and spread basis) and Condition 5 (holding days) move the result the most. The next section turns the columns you should record into a template.

Template

Comparison template: which columns to record

For a fair comparison, record the following columns for each account, keeping the advertised headline spread separate from the realized average spread. Dividend adjustments and rolls are usually 0 in forex but matter for instruments such as index CFDs, so keep them as columns (for the class-by-class differences, see trading costs by asset class).

To avoid confusing advertised values with realized fills, keep the capture time, account label, currency and update date as metadata. The next section folds these columns into a single equation.

Formula

The all-in cost equation: match units, then add

The normalized all-in cost (round-turn, per trade) sums the following parts in account currency. Spread and slippage convert from pips into money, commission uses its charging unit, and swap adds the holding-days portion.

Commission follows its charging mode. When it is proportional per lot with a minimum fee, apply the floor as follows.

Swap cost multiplies the receipt or payment per day per lot by holding days and quantity. Negative swap (a payment) is a positive cost; positive swap (a receipt) is a negative cost, handled by sign.

Substituting the fictional Scenario A (average spread 0.2 pips, ¥500/lot round-turn commission, ¥300 minimum fee, 0.2 pips round-turn slippage, −¥130 daily swap) holding 1.0 lot for one day gives:

So the spread-plus-slippage part of ¥400, the commission part of ¥500 and the swap part of ¥130 add up to ¥1,030 as the per-trade all-in cost. Reading that as break-even and cost ratio connects to break-even pips, cost ratio and trading friction. The next section applies the same equation to three fictional accounts.

Scenarios

Scenario A/B/C all-in comparison

Rather than real broker names, we compare three fictional accounts A/B/C with different cost structures. All are normalized to USD/JPY, 1.0 lot and ¥1,000 per pip.

At 0 days held (intraday, no overnight rollover), the order is A ¥900 < C ¥1,150 < B ¥1,300. Looking only here, Raw-type A wins, but that is a single slice that ignores holding period. In the next section, extending the hold changes the order.

Holding days

Rankings flip by holding period

Once holding-day cost is reflected, negative swap accumulates in proportion to days. Account A, with the largest daily swap, becomes less favorable the longer it is held. Here are the per-trade all-in costs at intraday, one day and ten days.

The intraday order A -> C -> B reverses completely to B -> C -> A at a ten-day hold. At ten days, A is 900 + 130×10 = ¥2,200, B is 1,300 + 35×10 = ¥1,650, and C is 1,150 + 60×10 = ¥1,750. The crossovers land roughly at 4–5 days for A vs B and 3–4 days for A vs C. This is a textbook case that “a narrow-spread account is always cheapest” is false. Triple-day schedules and holding treatment can be more complex still, so verify the actual daily swap in each account’s official table (for the calculation detail, see the swap and holding-days article).

Sensitivity

Spread sensitivity: how total cost moves when realized spread drifts

Average spread drifts with session and market conditions. So we multiply each account’s assumed spread by a factor (1.0/1.5/2.0/2.5/3.0/4.0) and line up how the round-turn spread + slippage + commission part (0 days held) moves as sensitivity. This is neither a forecast of the current value nor a cheapest designation; it is a fictional assumption for seeing the swing when spread drifts. Narrow-spread A has a gentle slope; wide-spread B reacts steeply.

What this sensitivity shows is that a narrow-spread account is robust: total cost rises little even when spread drifts. But as the previous section showed, extending the hold changes the order via swap. In other words, do not decide the cheapest from a single slice. The figures by multiple are A: ¥900/1,000/1,100/1,200/1,300/1,500; B: ¥1,300/1,800/2,300/2,800/3,300/4,300; C: ¥1,150/1,500/1,850/2,200/2,550/3,250.

Minimum fee

Minimum fee and the small/large crossover

A commission-type account may set a minimum fee per trade. At small size the minimum fee dominates and the per-lot burden jumps. We compare Raw-type A (rate ¥500/lot, ¥300 minimum, spread + slippage 0.4 pips = ¥400/lot) against no-commission Standard-type B (spread + slippage 1.3 pips = ¥1,300/lot) on round-turn total cost by lot size (0 days held).

At 0.1 lot, A’s effective commission stays at the ¥300 floor, which is ¥3,000 per lot-equivalent, or six times the proportional part (¥500/lot). As a result, below about 0.33 lot the no-commission B has the lower total cost. In other words, if you mostly trade small, the presence of a minimum fee matters; if you mostly trade large, the level of the proportional fee matters. Comparing at your own average lot is the practical approach.

Monthly & ratio

Monthly cost and cost ratio: view by frequency and target

Beyond a single all-in, line up the monthly cost (per-trade cost times monthly trade count) and the cost ratio against an assumed gross target. Here we compute at intraday (0 days), 1.0 lot, 40 round-turns per month and an assumed monthly gross target of ¥300,000.

Multiplying by frequency widens the gap. For short-term, small, high-turnover trading, A’s ¥36,000 per month is the lightest and its 12.0% cost ratio the smallest. But this assumes 0 days held; extend the hold and the order changes as in Table 3. The cost ratio is a gauge of how far cost bites into an assumed gross target, and the smaller the target, the more the ratio jumps. How turnover bites is covered in more detail in comparing trading costs by style (scalping, day trading, swing).

Mini calculator

Mini calculator: three-scenario comparison worksheet

Enter three account conditions (average spread, value per pip, commission mode/rate, minimum fee, slippage, daily swap, holding days and quantity), and it estimates each condition’s per-trade all-in, break-even units, monthly cost and cost ratio for educational purposes, sorted lower under these inputs. It shows no recommendation, cheapest label or broker name. Inputs are computed only in your browser and are never transmitted or stored. First check the equation with the static example below (Table 6), then swap the numbers in the tool beneath it.

Record

Recording comparison data: don’t confuse advertised values with realized fills

To make a fair comparison reproducible, record realized fills, not advertised headline spreads, in fixed columns. At minimum, keeping the capture time, account label, currency, instrument, quantity, headline spread, average spread, commission, minimum fee, slippage, swap, conversion and update date lets you recompute all-in cost later on the same basis. Measure several times in the same session and use the average of realized measurements for average spread.

Lining accounts up at the same capture time and currency also reveals the gap between headline and average (pronounced for A and B). Running this ledger continuously to track changes in cost conditions and swap is Premium-level territory. The thinking behind change auditing is covered in the trading cost audit and ledger article.

Pitfalls

Common mistakes

• Comparing on headline spread alone: judging by the advertised tightest value and overlooking average spread, commission and swap. Normalize to all-in cost.
• Ignoring holding days: deciding the cheapest from the short-term ranking and missing that negative swap flips it long-term. Line up intraday, 1-day and 10-day.
• Mixing up the commission unit: confusing per side with round-turn, or per-lot with fixed, causing double counting or undercounting. Match the units.
• Ignoring the minimum fee at small size: seeing only the proportional amount at 0.1 lot and missing that the minimum fee makes it expensive. Check the crossover.
• Not aligning the session: measuring one account at dawn and another in Tokyo hours so the spread basis conflicts. Measure at the same time.
• Deducting positive swap as fixed: subtracting a receipt swap as a permanent discount up front and underestimating cost. Plan conservatively.
• Ignoring non-cost factors: deciding on cost alone without regulation, client-money protection, execution policy, product range and support. Check separately.

All of these are avoided by the principle of normalizing with the six conditions and viewing several slices across holding period and lot size. Next we summarize this as a practical checklist.

Checklist

Practical checklist

A checklist to run before and after comparing account or broker trading costs.

• Did you align the six conditions (instrument, quantity, time and spread basis, order type, holding days, account currency)?
• Did you use the average spread for your session rather than the headline spread?
• Did you unify the commission unit (per side / round-turn, per-lot / fixed) and avoid double counting?
• Did you confirm whether a minimum fee applies and judge whether it bites at your average lot?
• Did you add the slippage assumption as round-turn pips and view several values as sensitivity?
• Did you add the holding-days portion of swap/funding and check the ranking at intraday, 1-day and 10-day?
• Where the P/L and account currencies differ, did you add conversion as a separate stage?
• Did you line up monthly cost (times frequency) and the cost ratio against an assumed gross target?
• Did you record the capture time, account label, currency and update date, keeping advertised values separate from realized fills?
• Did you check non-cost factors (regulation, client-money protection, execution policy, product range, support) separately?

Once you have checked a single condition’s all-in cost with your own numbers, the next step is your real inputs. When you want to flex several conditions at once, that is the point to consider Pro and Premium features. You can browse the other articles from the English article library.

Related

Related reading and read next

References

Sources and further reading

• SG Group Trade Cost Calculator (use it free) — check one-way / round-trip cost, break-even and all-in cost.
• SG Group Trade Cost Calculator plans — features and scope of Free, Pro and Premium.
• Trading costs by asset class — unit and holding-fee differences across FX, XAUUSD, index and crypto CFDs.
• Stress-test cost assumptions in a backtest — reflecting execution assumptions in historical testing.
• For precise definitions of spread, commission, swap/funding, execution policy, minimum fee and contract size, confirm the primary materials of each exchange, regulator or provider.