CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. 76% of retail investor accounts lose money when trading CFDs with this provider.

The true cost of trading gold: a 90-day spread study

An illustrative 90-day XAUUSD spread dataset, sampled once per second, shows how the cost surface moves by session — including the hours where spread costs can triple.

Educational note: figures, benchmarks and scenarios in this article are illustrative examples. They are not IUX production data, execution reporting or a representation of client outcomes.

6 min read

What 7.8M spread samples show

Spread is most of what a CFD trader pays on gold. This explainer uses an illustrative dataset — XAUUSD spreads sampled once per second for 90 days (7.8M samples) — broken down by session, to show how time of day changes the cost of a trade.

11.2¢

Median Raw-account spread, London/NY overlap

3.1×

Spread multiple in the 22:00–23:00 UTC dead zone

7.8M

One-second samples in the study

Median XAUUSD spread by session (Raw account)

US cents per oz

London / NY overlap (12:00–16:00 UTC)11.2
London morning13.4
NY afternoon14.8
Asia session19.6
Rollover dead zone (22:00–23:00 UTC)34.7
Illustrative 90-day dataset sampled at 1 Hz. Not IUX production spread reporting — live spreads vary by account, instrument and market conditions.

What traders should do with this

  • Size-sensitive strategies (scalping, high frequency of entries) should concentrate activity in the London/NY overlap — the cost per round trip is ~40% lower than Asia hours.
  • Avoid resting market orders through the 22:00 UTC rollover: the combination of 3× spread and thin book produces the worst fills of the day.
  • Swing traders holding for days can mostly ignore session spread — one 20-cent difference on a 40-dollar move is noise.
Fixed 1% risk per tradeSame signals, no stop discipline
Simulated 120-trade scalping equity: identical signals executed in the overlap window vs the dead zone. · Illustrative data — past performance is not a reliable indicator of future results.

Educational content only. Figures, benchmarks and charts are illustrative examples, not IUX production data or a representation of trading outcomes. Nothing here is investment advice.

More from the desk

Research · Jul 2026

Execution quality report: measuring the milliseconds

How execution quality should be measured: an illustrative six-month dataset of 4.1M orders shows why medians, percentiles and news-window behavior — not averages — tell the real story.

Research · May 2026

Why most retail stops are in the same place (and what that costs)

An illustrative dataset of 1.2M stop-loss placements shows the pattern: 61% cluster within 5 pips of round numbers and recent swing points — visible, crowded, and cheap to run.

Important information

Access to this website is restricted. IUX does not solicit, market, or provide services to persons located in, resident in, or citizens of Australia, Canada, the United Kingdom, the United States of America, or Malaysia. By selecting "I Confirm" below, you expressly represent and warrant that you are not a citizen or resident of, and are not accessing this website from, any of the jurisdictions listed above. If you do not meet this requirement, you must refrain from accessing the website.