The XAUUSD Trading Day: How Gold Moves Through Tokyo, London and New York

Gold trades almost 23 hours a day, and that is exactly the trap. Because the chart never stops, it is tempting to treat 03:00 and 16:00 as the same market. They are not. XAUUSD runs through three distinct personalities every day — a quiet Asian build, a London decision, and a New York main event — and a surprising share of losing gold trades are not bad ideas. They are reasonable ideas executed in the wrong hours.

This post is a map of the typical gold day: what each session actually does, the one number I check before every entry, and a three-check routine that takes less time than a coffee. Nothing here requires an indicator — a notepad is enough. At the end I'll show how I keep the whole backdrop on one panel.

A housekeeping note first: times below are MT5 server time as used by the common GMT+2/+3 brokers. Check your broker's offset once, write the three session windows on a sticky note, and never think about it again.

Tokyo: the quiet build (roughly 01:00–09:00)

The Asian hours rarely decide the gold day. Participation is thin, moves are slow drifts rather than impulses, and the occasional physical-demand headline aside, nothing structural tends to happen. What Asia reliably produces is structure for later: a range. Stops accumulate above the Asian high and below the Asian low, and the European entrance usually resolves them.

The job here is preparation, not trading:

• Mark the Asian high and low.
• Write down ADR(20) — more on this number in a minute.
• Note today's high-impact US releases and their times.

The classic Tokyo mistake is buying a "breakout" of yesterday's range in a market with nobody in it. Most of those drifts are quietly retraced when London arrives.

London: the first decision (the loud part is roughly 09:00–12:00)

When European liquidity arrives, the first order of business is very often the Asian range: one side gets swept, stops are collected, and the day's real auction begins. For gold specifically, the London morning prints one of the day's extremes surprisingly often — which means the first hours are less about riding a trend and more about watching which side of the overnight range gets defended.

A sweep of the Asian low that snaps straight back inside the range is exactly the setup I described in the London open post — the sweep, tap, reaction sequence works on gold the same way it works on FX pairs.

Two London-specific warnings. First, the initial push gets reversed by New York often enough that "London direction" is a hypothesis, not a conclusion. Second, between the London morning and the US data there is usually a lull. A dead chart in the early European afternoon is normal, not a signal.

New York: the main event (from 15:30)

For XAUUSD, New York is the center of gravity. COMEX futures volume dwarfs the rest of the day, the big US releases land at 15:30 and 17:00 server time (08:30 and 10:00 in New York), and the London–New York overlap gives you the fastest, widest hours on the clock. Trend days do most of their travel here. Range days pivot here, usually on the data.

If you only have ninety minutes a day for gold, this overlap is the window to watch — with one caveat: a high-impact release will happily trade through any technical level on the chart. Know the release times before you care about the levels.

After roughly 20:00 the day is effectively done. Ranges compress, moves stop following through, and spreads widen noticeably into the 23:59 rollover — worth remembering before you leave a tight stop resting overnight.

ADR: the fuel gauge

ADR(20) is the average daily range of the last twenty trading days — the average distance from high to low. For gold it is a single number (lately somewhere in the 110–150 point area depending on the regime), and it answers the question that kills most late entries: how much day is left?

The arithmetic is one division. Take today's range so far and express it as a percentage of ADR(20):

• Under roughly 40% going into New York: the day still has fuel. Continuation setups have room to pay.
• Around 90–130%: the tank is mostly spent. Chasing a breakout here means buying the top of the day's fuel, and the odds of mean reversion against you rise sharply.
• Beyond roughly 130%: only fresh news keeps extending a day like this. Without it, expect the range to stop growing.

An average is not an appointment — gold does not owe you 100% of ADR every day, and on a big news day it can do 200%. But as a filter it is brutally effective at one thing: stopping you from paying breakout prices late in the afternoon on a day that has already travelled its distance.

The screenshot above shows the idea in panel form: 155 points travelled against a 114-point ADR(20) — 135%, bar red. Whatever the pattern on M5 looks like, that is a late day.

The three checks

The whole structure compresses into a small routine:

1. Before London (five minutes): Asian high and low marked, ADR(20) written down, release times noted.
2. At the London open: watch the sweep. Which side of the Asian range was taken — and did price accept beyond it, or snap back inside? That single observation sets the bias for the next several hours.
3. Before New York: check ADR% and the calendar. Under half the tank with a clean bias — continuation is on the table. Tank nearly empty, or a red release inside the next hour — stand down or size down.

None of this predicts anything. It just stops the two most common gold mistakes: trading Tokyo as if it were New York, and buying the 130%-of-ADR breakout.

What not to do

• Don't take breakout trades during Tokyo. Thin-liquidity breaks are the ones London un-does first.
• Don't treat ADR as a target. It measures how much is left, it doesn't promise how far price goes.
• Don't hold a scalp through a high-impact release because it is "almost at target". The release does not know about your target.

Putting the backdrop on one panel

Everything above fits on a sticky note and a morning glance at the calendar, and I ran it that way for a long time. The tedious part is keeping it current through the day: which session is live and how long until the next one, today's range against ADR right now, whether a high-impact release sits inside the next hour, and whether the dollar is pressing gold or supporting it while all of that happens.

That is the part I automated. Gold Pulse (free) keeps the whole backdrop on the chart: the session row with a countdown to the next open, the day's range drawn as an ADR percentage bar, a live countdown to the next high-impact release, and a one-line macro verdict — USD and risk sentiment combined into HEADWIND or TAILWIND for gold. The idea is the same as this post: read the backdrop first, then argue with the chart.

The Pro version adds the alert layer — macro verdict flips, T-minus warnings before releases, session opens and level breaks — so the backdrop can watch itself while you do something else. The free version carries the full panel; start there.

If your broker's session times or calendar feed don't line up with what you see here, post a comment on the product page — I read everything and usually reply within a day.