What is the spread on gold with XM Ultra Low?

XM advertises gold spreads from roughly 0.20-0.25 USD per ounce on Ultra Low accounts, with no commission. Real quotes breathe with liquidity: tightest during the London-New York overlap, widest around the daily rollover at 00:00 server time. Always verify on your own live feed before sizing a scalp.

What is the best time of day to trade gold?

The 14:00-22:00 Bangkok time window (London morning into New York morning) carries the deepest liquidity and the cleanest follow-through. The Asia session is fine for scalping inside established ranges, but breakouts usually need London or New York volume behind them to hold.

What lot size should I use for gold?

Anchor it to your stop, not your conviction: one 0.01 lot position moves about 1 USD per 0.01 of price, so a 45-dollar stop on 0.01 lot risks roughly 45 USD. Keeping risk near 1 percent of equity means about 0.01 lot per 5,000 USD with that stop distance.

How do you manage drawdown?

With three hard rules: risk no more than 1 percent per position, stop trading for the day after three consecutive stops, and cut size in half after any week that ends down more than 3 percent. Drawdown management happens before the trade, not during it.

Where do your gold signals come from?

From my own price-structure work on the XAUUSD feed from CandleAPI by iCafeFX — session ranges, round-number shelves and compression bands — cross-checked on the MT5 chart before publishing. No signal is copied from another provider, and no level is ever invented for content.

How I Traded Gold This Week on XM Ultra Low — Process Journal (Aug 2026)

Gold trading week on XM — dark desk with golden charts

By Aj.Bom Kittitat — XM VIP Partner for 13+ years, iCafeFX · Published 15 Aug 2026

I've been trading XAUUSD on XM for 13 years — through the 2015 flash moves, the 2020 explosion, and the slow grind of the last two years. This journal is a process journal, not a trophy wall. I don't post P&L screenshots, and I won't invent results here. Every price in this post comes from one source I trust and control: the XAUUSD feed from CandleAPI by iCafeFX, whose last H1 close before writing sat at 4,376.19 (high 4,379.99, low 4,372.10, timestamped 2026-08-14 19:00 UTC, 15-minute delayed). What you get instead of a fantasy win rate is how the week actually looked through my eyes, and how each idea becomes an order on XM Ultra Low.

What the week actually was: compression

The single most important fact of the week is boring: gold went sideways. The last daily candles compressed into an H1 band roughly eight dollars wide, 4,372 to 4,380, with ATR14 on the hourly around 14.5 — which means the average hourly travel barely exceeded the band itself. Compression weeks punish both the buyer of breakouts that never come and the seller of highs that never break. They reward patience and punish improvisation. So the week, for me, decomposed into three standard setups — the shelf, the scalp, and the flat. Here is each one, exactly as it is structured.

Setup 1 — The shelf buy at 4,350

Every month this year has left the same fingerprint on the chart: dips into the 4,350 round number get absorbed. Not reversed violently — absorbed, candle after candle, like a sponge. That shelf is the highest-quality demand zone on the current chart, and my weekly plan (published as the weekly signal card) is to buy the pullback into 4,350–4,360 with a stop under the weekly structure at 4,315 and targets at 4,420 and 4,480.

How that becomes an order: a limit order inside the zone, not a market buy when the chart looks scary. Size first — with a ~45 dollar stop, 0.01 lot per 5,000 USD of equity keeps risk near one percent. Then patience: if the week never dips to the shelf, there is no trade, full stop. The biggest lesson of thirteen years is that the trades you don't take pay for the ones you do.

Invalidation is mechanical: two H4 closes below 4,315 and the scenario is dead — no averaging, no widening the stop because "gold always comes back." That sentence has ended more trading careers than any broker ever did.

Setup 2 — The compression scalp inside the band

While waiting for the shelf, the 4,372–4,380 band itself offers the only intraday trade worth taking in a compression week: fade the edges, tiny targets, only when both London and New York are active. The intraday card documents the current grid — buy the 4,372–4,374 shelf, out at 4,381/4,385, dead if the envelope 4,360–4,395 breaks.

This is where XM Ultra Low earns its keep, because a scalp's edge is mostly a cost question. At an advertised spread of roughly 0.20–0.25 USD per ounce with no commission, a 0.05-lot scalp pays about 1.00–1.25 USD per round turn in spread — small enough that a five-dollar target survives it. On a wide-spread account the same trade can start one dollar underwater and the math quietly dies. Verify the live quote before you send it: spreads breathe with liquidity, and the daily rollover (00:00 server time) widens them dramatically for a few minutes. I simply don't trade the rollover.

Scalp rules from this week, unchanged in a decade: one attempt per edge, absolute stop (the 4,368 from the card), full exit when the band breaks either way. Scalps are high-frequency decisions with low-frequency patience.

Setup 3 — The Friday flat

Friday is where process beats intelligence. US data lands Friday afternoons Bangkok time, the weekend gap sits two days away, and the swaps you'll pay or receive depend on your account type — check the Specification window in MT5, because swap values are revised weekly and I refuse to quote a number that may be stale by the time you read it. My standing rule: no new gold positions after 22:00 ICT on Friday, and anything not worth holding through a two-day gap gets closed. The market pays you nothing for heroically holding a scalp through Sunday's opening print.

What XM Ultra Low changes, honestly

Three things, after thirteen years on this broker. First, cost: the gold spread profile suits scalps and swing entries alike — advertised from about 0.20–0.25 USD per ounce, no commission, but always confirm on your live account, because your entity and account type decide your real number. Second, execution: limit orders at my shelf levels fill where I place them or not at all, which is all a structure trader needs; I don't chase milliseconds. Third, the ecosystem: MT5 on desktop and mobile, and a support line in Thai — as a Thai trader I no longer accept brokers who make me file disputes in a second language.

What it doesn't change: discipline. No account type fixes over-sizing. The 1 percent rule comes from me, not from the broker.

Lessons from the week

Compression is information. An eight-dollar band after a trending month is the market loading its next move. Prepare levels, don't predict direction.
Round numbers are load-bearing. 4,350 has absorbed dips all month. Trade with the structure that repeats, not the story that excites.
Cost is part of the setup. A five-dollar scalp target and a 0.25-dollar spread coexist; a five-dollar target and a one-dollar spread do not. Know your number before the trade.

If this process resonates, the weekly card and the intraday grid publish on this site, and the live levels post in the Telegram channel at session opens. Open an XM account through the link below if you want the same execution profile I describe here — and risk only what you can afford to lose. That is not a disclaimer sentence to me; it is the first rule of the next thirteen years.