Gold Volatility by Hour: What Each Session Actually Delivers

The shape of the gold day

Spot gold runs nearly non stop from the Sunday evening reopen until late Friday in New York, interrupted only by a short maintenance halt near 5 pm New York. Inside that long window, activity arrives in waves set by which desks are staffed.

Asian hours grind sideways. London arrives and the first hour pushes through whatever edge the Asian range left exposed. The New York morning overlaps London and swallows the scheduled data. Then liquidity leaves the building and the final hours belong to position squaring and thin, untrustworthy moves.

Volatility follows people, not the clock by itself. Public holidays in either financial centre flatten the profile, and the final week of December barely resembles a normal week at all.

Asian hours: small ranges, honest levels

From the Sunday evening reopen through the Tokyo and Sydney hours, gold typically travels a fraction of what it will cover later. Single digit dollar ranges are common, and the extremes of that period are often set early and then respected.

Two things follow. A stop sized for the New York morning is far too wide here, and a target sized for New York will rarely be hit before Europe arrives. More useful, the Asian high and low become the two most obvious pools of resting orders on the chart, which is why London so often removes one of them.

These hours are for marking, not for chasing. Patient range work is possible, but the spread eats a larger share of every dollar captured.

The London open expansion

The hour around the London open is where gold normally stops being quiet. Institutional and physical flow returns, the European currencies start moving against the dollar, and one side of the Asian range gets tested.

The usual sequence is a push through the Asian high or low, a fast few dollars, then either continuation into the New York morning or a snap back through the whole range within the next hour. That second path is the classic sweep, which is why entering on the break alone is a poor habit. When we measured daily range breaks on gold, roughly one in two had failed within a week and the following leg tended to travel the other way, and the intraday version has the same character.

This hour gives the most movement for the least warning, from a level you could have drawn hours earlier.

The overlap and the data window

London and New York are both open for several hours, and that overlap holds the deepest liquidity gold ever sees. The single biggest hour, though, is usually the one containing the half past eight New York releases: inflation prints, the monthly jobs report, and the afternoon rate decisions, which land later but behave the same way.

Two features matter more than the raw size. The move is often immediate and then partly given back inside the same hour, because the knee jerk reaction and the considered reaction are different trades. And the spread widens exactly while the candle is longest, so the fill you assumed in testing is not the fill you get.

One release hour can produce more range than the previous twelve hours combined. Risk has to be sized for that, not for the average hour.

London close and the late hours

Around the London close, European desks square up and the character of the day often changes. A move that ran all morning stalls, and a drift back toward the middle of the daily range is common. That is less a trend reversal than the removal of the flow which was driving it.

After that the late New York hours thin out. Ranges shrink, levels hold less reliably, and one decent order pushes price further than the same order would have managed at lunchtime. Stops parked just beyond an obvious level are at their most exposed here, and the move that collects them is often gone by the reopen.

For most traders the working day effectively ends at the London close. What follows is management, not hunting.

Build your own hourly range table

Any published profile is an average of a feed you do not trade, over a period you did not pick. Building your own takes an afternoon and gives numbers that match your actual chart.

Pull H1 candles for several months from the feed you trade, group them by hour, and record for each hour the median high minus low, the share of days that hour set the daily extreme, and the median absolute move from open to close. Use the median, because a handful of release days will drag any average upward and flatter the quiet hours.

• Use three months at minimum, and keep release days in their own column rather than deleting them.
• Stamp every hour in New York time so the scheduled releases stay aligned.
• Rebuild it twice a year, since clock changes shift the two centres relative to each other.
• Compare two brokers if you can: the shape will agree, the absolute numbers will not.

Stepping back through a month of H1 candles on the live gold chart is slower but builds a better memory of the same shape.

What the profile changes about stops and sizing

Once you know the median range of the hour you are about to enter, several decisions stop being guesswork. A stop that sits inside the normal noise of the New York morning is not protection, it is a donation. The same distance during Asian hours can be wider than the entire session range, so you hold a position for hours and achieve nothing.

Targets work in reverse. If the current hour usually covers 5 dollars, a 20 dollar target requires the move to survive into the next session, which means you are now carrying a trade through a liquidity change you did not plan for.

Because sizing follows the stop, the hourly table quietly sets your position size too. Express the stop as a multiple of the median range of the hour you are trading, and hold that multiple steady rather than the dollar distance.

Where an hourly profile misleads you

This describes an average day, and nobody trades the average day. The limits deserve stating plainly.

A scheduled release relocates the volatility to wherever the release is, so a normally dead hour becomes the loudest one on a rate decision afternoon. Holidays, month end and the turn of the year break the pattern outright. Clock changes in Europe and the United States fall on different weekends, so for a couple of weeks each year the overlap sits an hour away from where your table claims. And the profile says nothing about direction: it estimates how far gold may travel next, never which way.

Traders who start reading direction into the clock end up with a rule such as always fading the London close, and that works until the month it does not.

A sane routine for using the clock

Before the session, mark the Asian high and low, note any scheduled release and the hour it lands in, and write down the median range for the hours you plan to be at the screen. Five minutes of work frames everything after it.

During the session, let the London open hour show its hand instead of entering on the first break of the Asian range. Treat the data window as either a planned trade with release sized risk or a stand aside. At the London close, stop looking for entries and start managing.

For the wider question of which window suits which style, our guide to the best time to trade gold covers the session choice itself.