Spreads, Commissions and Swap

Swap, the cost nobody budgets for

Hold a position past the daily rollover, usually 17:00 New York time, and it is charged or credited swap. Swap reflects the interest rate difference between the two currencies in the pair, adjusted by the broker's own markup. Hold the higher-yielding currency and swap can be positive. Hold the lower-yielding one and it is negative, often by more than the rate differential alone implies because the markup applies in both directions.

One detail catches people out. Swap is normally tripled once a week, on Wednesday for most pairs, to account for the weekend value date. A position that costs 2 USD a night quietly costs 6 USD that night. Over a three-week swing trade in a negative-carry direction, financing can consume a meaningful share of a modest target, and it is charged whether the trade is winning or losing.

The costs that do not appear on the pricing page

• Slippage. Fills away from the requested price in fast markets. Not a fee, but money all the same, and it is systematically worse on stop orders than on limits.
• Spread widening. The quoted spread applies to normal conditions. Your stop can be triggered by a widening bid that never traded as a real price in either direction.
• Currency conversion. Profits realised in a currency other than your account currency are converted, usually at a rate a little worse than mid.
• Account fees. Inactivity charges, withdrawal fees and dormancy rules vary widely. Read the schedule before funding, not after.

Why frequency is the multiplier

Work an example. A 1.2-pip all-in cost on 0.20 lots is about 2.40 USD per round turn. Trade five times a day, twenty days a month, and that is 240 USD a month in costs. On a 5,000 USD account that is roughly 4.8% of the account each month that the strategy must produce before you are level. Halve the frequency and you halve the hurdle. Double it and the strategy needs to be twice as good simply to stand still.

This is the argument against scalping tiny targets on a retail cost structure, and it is arithmetic rather than opinion. A 5-pip target against a 1.2-pip cost gives away roughly a quarter of the gross result on every trade. A 60-pip target gives away 2%. The same execution quality feels completely different depending on how far you are trying to travel.

What to do about it

Record costs per trade in your journal as a separate column from the result, so you can see the drag rather than infer it. Compare your broker's all-in cost on your real pairs and sizes rather than on their advertised showcase pair. Avoid trading the first minutes after a major release unless the strategy is specifically built for it, because that is when spread and slippage are both at their worst. And when you evaluate any approach, subtract costs before deciding it works, since the arithmetic of a small account leaves very little room for a hidden 5% monthly hurdle.