Forge Strategy Lab

Article

How far does gold move in a day?

Over the last 12 months, the average daily range for XAU/USD was $112.08 (1,121 pips). The median day was smaller at $92.10 (921 pips). The gap between the two shows that a relatively small number of unusually wide days pulled the average higher, making the median a useful guide to the middle of the distribution.

These figures come from our full daily range data for gold, built from one-minute XAU/USD prices covering 2,623 complete trading days from 15 August 2016 to 23 October 2026.

The quiet-to-wild spectrum

Not every day finishes close to the median. Across the most recent 12 months, covering 24 October 2025 to 23 October 2026 and 258 complete trading days, the ranges were distributed as follows:

  • One in 10 days had a range below $53.96 (540 pips)
  • One in four days had a range below $69.85 (698 pips)
  • The median daily range was $92.10 (921 pips)
  • One in four days had a range above $124.52 (1,245 pips)
  • One in 10 days had a range above $177.33 (1,773 pips)

This is a wide distribution. A quiet day can cover only a fraction of the distance seen during a particularly active one. These bands provide context for judging whether a target requires an ordinary move or an unusually large one, although the full daily range cannot be known until the session is complete.

Recent conditions can differ substantially from the longer-term picture. The average daily range over the latest 20 trading days, from 28 September to 23 October 2026, was $73.46 (735 pips). The latest complete trading day, Friday 23 October 2026, recorded a range of $58.44 (584 pips).

I would not reduce a take-profit automatically just because the recent 20-day average has fallen. Most of my automated strategies use targets defined and tested in advance. I use the recent range as context, particularly when deciding whether a target still looks realistic after gold has already completed a substantial move before the entry appears.

Does the day of the week matter?

The weekday averages over the latest 12 months were:

  • Monday: $107.16
  • Tuesday: $107.48
  • Wednesday: $111.96
  • Thursday: $117.40
  • Friday: $116.58

Thursday recorded the widest average range and Monday the narrowest. However, all five weekdays remained within a relatively narrow band compared with the much larger variation between individual trading days.

The weekday alone is therefore weak evidence for changing a target. It is more useful as an additional piece of context alongside current volatility, how far gold has already moved and the time at which an entry occurs.

What the range means for profit targets

A profit target requires price to travel a certain distance from the entry, while the daily range measures the full distance between the day’s high and low. These are not directly interchangeable because an entry rarely occurs exactly at the daily high or low.

A 500-pip target does not necessarily require a daily range of only 500 pips. Gold may already have moved considerably before the trade is opened, or it may initially travel against the position before reaching the target. Entry timing and direction therefore matter as much as the headline daily range.

The daily-range distribution is still useful. If a strategy regularly requires price to travel a distance from entry that is large relative to a normal day, the target may only be reached during more active conditions. That does not make the target wrong, but it should be tested across quiet, typical and unusually volatile periods.

Over the latest 12 months, the average daily range was equivalent to 2.48% of gold’s opening price. Expressing the range as a percentage helps keep the figures in context as the underlying gold price changes over time.

Position size magnifies every movement. Where a broker defines one standard lot of gold as 100 ounces, a $1 movement is worth $100 on one lot. Using Forge’s convention that one gold pip is $0.10, one pip is worth $10 on one standard lot. Traders should still check their broker’s contract specification because lot and pip values can vary.

Most of my automated strategies use a fixed stop and take-profit because those rules can be tested consistently across years of data. I only use a trailing stop when it forms part of the strategy that was originally tested. The day’s range provides context, but I would not change a live strategy’s exit rules simply because one session appears quieter or busier than usual.

When does the high or low tend to form?

Daily-range information becomes more useful when combined with the time at which gold has historically reached its extremes. Our data on when gold makes its high and low of the day shows that those extremes can form much earlier than many UK traders expect.

For this particular dataset, Forge assigns every open hour to one of three non-overlapping analysis blocks:

  • Overnight and Asian analysis block: 23:00-07:00 UK time
  • London-morning analysis block: 07:00-13:00 UK time
  • US and New York analysis block: 13:00-22:00 UK time

These blocks are used to prevent the same hour from being counted twice. They should not be interpreted as the complete official trading hours of the Asian, London or New York markets, which overlap.

These analysis blocks also differ from the windows I use personally. I mark the Asian high and low from 01:00 to 05:15 UK time. My main London trading session runs from 08:00 to 11:00, while my New York session runs from 15:00 to 18:00. These are the practical windows I use when developing and testing strategies; the session percentages below belong to Forge’s separate, non-overlapping analysis blocks.

Using those analysis blocks, the US and New York period produced the day’s high on 49% of days and its low on 44% of days during the latest 12 months. The overnight and Asian block produced the high on 42% of days and the low on 46%. The London-morning block produced 9% of highs and 10% of lows.

At an hourly level, 23:00 UK was the most common hour for both extremes, first recording the daily high on 11% of days and the daily low on 14%. The 01:00 hour first recorded the high on 10% of days, while 15:00 first recorded the low on 11%.

Across the latest 12 months, the 23:00-07:00 analysis block contained at least one of the day’s two extremes on 82% of trading days. This does not mean 82% of days completed their entire meaningful move overnight. It means either the high, the low or both were first established during that block.

For a practical example of how the range and timing figures played out across a real week, see our gold week in numbers for 19 October to 23 October 2026. That week recorded an average daily range of $62.45 (625 pips), and price broke the measured Asian range on all five trading days.

I do not place an order simply because 23:00 is historically the most common hour for a daily extreme. When my trading begins, I want to know where the overnight price action sits in relation to the Asian high and low I record from 01:00 to 05:15. I then watch how price reacts to, sweeps or breaks those levels during my 08:00-11:00 London window and 15:00-18:00 New York window. The levels provide structure and context rather than an automatic entry signal.

Putting it together for trade planning

The median daily range provides a baseline, while the quiet-to-wild bands show how widely individual sessions can vary. Weekday averages add some context, and the hourly data shows when the day’s extremes have historically been established most often.

None of these figures predicts the next session by itself. Their value is in helping traders test whether an entry window, stop and target remain realistic across different volatility conditions.

A target larger than the distance gold normally travels after a strategy’s entry is not impossible, but it will naturally be reached less often. That trade-off should appear in the backtest rather than being decided from one recent day or a single average.

For my automated strategies, I follow the exit rules that were tested rather than changing them during the trade. If a strategy is designed to trail after price reaches a defined level, I follow that rule. If it uses a fixed target, I leave the target in place. The better approach is to test an exhaustion or range filter beforehand rather than improvise an exit because the day already looks stretched.

About Glen Hall

Glen Hall trades gold (XAU/USD) and is a founding member of Forge Strategy Lab. He develops and tests systematic gold strategies, using backtesting and forward testing to understand what works, what doesn't, and how strategies behave in changing market conditions.

This page is education, not financial advice. The figures describe how gold has moved in the past; they do not predict how it will move next. Trading involves risk and you can lose more than you expect. Never risk money you cannot afford to lose.

Turn a pattern into a tested strategy.

Build a gold strategy around what you have just read, without code, and backtest it on the same years of prices. Free to start, no card required.

Start free