Market risk researchXAUUSD average daily range: 20 years of gold data by year and month
Huyen X. Vu · Quantex Research, AlgoTrade Pte. Ltd., Singapore · ORCID 0000-0002-2209-842X · Download PDF
Abstract
How far does gold move from high to low in a normal day? On 5,138 trading days from 2005 to 2024 the mean daily XAUUSD range was 1.520% of price (median 1.279%). The year mattered most: 2008 averaged 2.878% and 2018 0.966%, almost three times less, while the month made little difference (March 1.629%, July 1.408%). Wide days were common: 20.2% of days spanned 2% or more. On 4.0% of days the range was at least twice the average of the 14 days before, with a peak of 6.78 times on 15 April 2013.
Keywords: gold; XAUUSD; daily range; volatility; volatility clustering
JEL: G14; Q02
Highlights
- The average day spans 1.520% of price. That was US$19.78 on average. The median day was 1.279% (US$16.65). Half of all days fell between 0.925% and 1.819%.
- The year matters most. The widest year was 2008 with a mean range of 2.878%; the quietest was 2018 at 0.966%. The wide year moved 2.98 times as much. In US$ the average day grew from US$5.27 in 2005 to US$33.04 in 2024, mostly because the gold price rose.
- The month matters little. March was the widest month (1.629%, 1.07 times the average day) and July the quietest (1.408%, 0.93 times).
- Big days are common. 20.2% of days had a range of 2% or more, and 6.2% had 3% or more.
- Range can jump fast. On 4.0% of days (about one in 25) the range was at least twice the average of the 14 days before. The largest jump was 6.78 times, on 15 Apr 2013.
1. Introduction
How far does gold move from high to low in a normal day, and how much does that change? We measured the daily range of XAUUSD on 5,138 trading days from 3 Jan 2005 to 31 Dec 2024 and grouped it by size, by year and by month. We also checked how often a day is much wider than the two weeks before it.
2. Data and method
Method
- Data: daily XAUUSD bars from the public dataset "xauusd-gold-price-historical-data-2004-2025" on Hugging Face (file XAU_1d_data.jsonl), the same file as notes 08 and 09. Window 3 Jan 2005 to 31 Dec 2024. 2004 is a part year and 2025 has long gaps, so both are left out.
- Days used: a day counts only if the bar before it is at most 4 calendar days earlier. 2 days were skipped for data holes (November 2005 and November 2006). 5,138 days were used, the same days as note 08.
- Daily range: high minus low of the daily bar, as a % of the previous bar's close, and in US$. The median (1.279%, US$16.65) matches note 08.
- True range: the higher of high and previous close, minus the lower of low and previous close, as a % of the previous close.
- 14-day ratio: the day's range % divided by the mean range % of the 14 days used before it. For early January 2005 the look-back uses late-2004 bars.
3. Results
How big is a normal day?
Daily range here is the day's high minus its low, divided by the previous day's close. Over 5,138 days the mean was 1.520% and the median 1.279%. The mean is higher than the median because a few very wide days pull it up: the widest day in the data, 17 Sep 2008, spanned 11.710% (US$91.10).
| Point in the spread | Daily range |
|---|---|
| 10% of days below | 0.701% |
| 25% of days below | 0.925% |
| Median day | 1.279% |
| 75% of days below | 1.819% |
| 90% of days below | 2.570% |
| 95% of days below | 3.207% |
| 99% of days below | 5.321% |
| Daily range | Days | Share of days |
|---|---|---|
| 0.5% or more | 5,036 | 98.0% |
| 1% or more | 3,566 | 69.4% |
| 1.5% or more | 1,966 | 38.3% |
| 2% or more | 1,036 | 20.2% |
| 3% or more | 317 | 6.2% |
| 4% or more | 122 | 2.4% |
| 5% or more | 63 | 1.2% |
Read the first table as: on 90% of days the range was below 2.570%. The second table counts days with at least that range. The narrowest day was 2 Jan 2006 at 0.097%, a holiday session.
Average daily range by year
| Year | Days | Mean | Median | 90% of days below | Mean US$ | Median US$ | Average close | Widest day | Date |
|---|---|---|---|---|---|---|---|---|---|
| 2005 | 251 | 1.172% | 1.045% | 1.868% | US$5.27 | US$4.60 | US$444.01 | 3.277% | 20 Dec 2005 |
| 2006 | 250 | 2.177% | 1.929% | 3.430% | US$13.21 | US$11.75 | US$603.79 | 7.216% | 13 Jun 2006 |
| 2007 | 257 | 1.554% | 1.417% | 2.485% | US$10.96 | US$9.60 | US$696.01 | 4.810% | 12 Nov 2007 |
| 2008 | 258 | 2.878% | 2.464% | 5.063% | US$24.69 | US$21.40 | US$872.09 | 11.710% | 17 Sep 2008 |
| 2009 | 257 | 1.914% | 1.673% | 3.177% | US$18.49 | US$16.52 | US$973.03 | 7.169% | 18 Mar 2009 |
| 2010 | 256 | 1.471% | 1.386% | 2.271% | US$17.94 | US$16.37 | US$1,226.11 | 4.615% | 4 Feb 2010 |
| 2011 | 256 | 1.836% | 1.523% | 3.237% | US$29.70 | US$23.52 | US$1,573.10 | 7.952% | 26 Sep 2011 |
| 2012 | 257 | 1.350% | 1.215% | 2.192% | US$22.44 | US$20.02 | US$1,668.75 | 5.863% | 29 Feb 2012 |
| 2013 | 258 | 1.817% | 1.559% | 2.951% | US$25.20 | US$21.87 | US$1,409.60 | 10.691% | 15 Apr 2013 |
| 2014 | 258 | 1.315% | 1.136% | 2.137% | US$16.55 | US$14.30 | US$1,266.03 | 6.781% | 1 Dec 2014 |
| 2015 | 258 | 1.359% | 1.226% | 2.171% | US$15.78 | US$14.56 | US$1,159.34 | 5.588% | 20 Jul 2015 |
| 2016 | 258 | 1.466% | 1.267% | 2.322% | US$18.27 | US$15.90 | US$1,249.43 | 8.592% | 24 Jun 2016 |
| 2017 | 257 | 0.989% | 0.943% | 1.514% | US$12.42 | US$11.80 | US$1,258.46 | 2.044% | 15 Mar 2017 |
| 2018 | 258 | 0.966% | 0.879% | 1.516% | US$12.27 | US$11.21 | US$1,268.56 | 2.938% | 11 Oct 2018 |
| 2019 | 258 | 1.097% | 0.945% | 1.909% | US$15.45 | US$13.27 | US$1,393.58 | 3.650% | 13 Aug 2019 |
| 2020 | 259 | 1.719% | 1.384% | 2.796% | US$30.36 | US$24.94 | US$1,771.94 | 7.309% | 16 Mar 2020 |
| 2021 | 258 | 1.316% | 1.167% | 2.077% | US$23.68 | US$20.95 | US$1,798.97 | 4.662% | 8 Jan 2021 |
| 2022 | 258 | 1.418% | 1.273% | 2.213% | US$25.58 | US$22.98 | US$1,801.60 | 5.058% | 24 Feb 2022 |
| 2023 | 257 | 1.218% | 1.077% | 1.830% | US$23.69 | US$20.74 | US$1,942.97 | 6.017% | 4 Dec 2023 |
| 2024 | 259 | 1.378% | 1.242% | 2.076% | US$33.04 | US$30.63 | US$2,389.53 | 4.251% | 7 Jun 2024 |
Mean and median of the daily range in each calendar year. "90% of days below" = the 90th percentile. US$ values follow the gold price, so compare years with the % columns. Average close = mean daily close of the year.
Average daily range by month
| Month | Days | Mean | Median | vs average | Days 2% or more |
|---|---|---|---|---|---|
| January | 424 | 1.496% | 1.334% | 0.98x | 18.9% |
| February | 402 | 1.497% | 1.259% | 0.98x | 19.4% |
| March | 438 | 1.629% | 1.367% | 1.07x | 25.1% |
| April | 416 | 1.494% | 1.264% | 0.98x | 19.0% |
| May | 439 | 1.463% | 1.284% | 0.96x | 20.3% |
| June | 428 | 1.486% | 1.264% | 0.98x | 17.3% |
| July | 440 | 1.408% | 1.238% | 0.93x | 17.5% |
| August | 445 | 1.478% | 1.249% | 0.97x | 18.9% |
| September | 425 | 1.612% | 1.297% | 1.06x | 22.1% |
| October | 441 | 1.614% | 1.311% | 1.06x | 20.6% |
| November | 420 | 1.571% | 1.277% | 1.03x | 21.0% |
| December | 420 | 1.492% | 1.252% | 0.98x | 21.9% |
All 20 years together. "vs average" = mean range of the month divided by the mean of all days (1.520%). The spread between months (0.93 to 1.07 times) is small next to the spread between years.
How often a day beats its 14-day average range
Many EAs and traders size stops and targets on a recent average range, often the last 14 days (ADR14). We compared each day's range with the mean range of the 14 trading days before it.
| Day's range | Days | Share of days | About one day in |
|---|---|---|---|
| 0.5x or more | 4,813 | 93.7% | – |
| 1x or more | 2,113 | 41.1% | – |
| 1.5x or more | 651 | 12.7% | 1 in 8 |
| 2x or more | 204 | 4.0% | 1 in 25 |
| 2.5x or more | 72 | 1.4% | 1 in 71 |
| 3x or more | 35 | 0.7% | 1 in 147 |
Ratio = day's range divided by the mean range of the previous 14 days in the data. The median ratio was 0.909: a normal day is a little smaller than the recent average, because the average includes the wide days. The largest ratio was 6.78 on 15 Apr 2013: a range of 10.691% after a 14-day average of 1.576%.
The 14-day average itself moved a lot. Its lowest reading was 0.659% (20 Sep 2023) and its highest 5.734% (28 Oct 2008). On 80% of days it was between 0.957% and 2.230%.
Range and true range
Average true range (ATR) also counts a gap from the previous close. For gold daily bars the difference was small: mean true range 1.532% against a mean range of 1.520%, 1.008 times. The previous close was outside the day's high-low on 7.0% of days.
4. Discussion
What this means for an EA
- A stop or grid step set to one average day is hit often. 38.3% of days had a range of 1.5% or more, and the 14-day average range was reached or beaten on 41.1% of days.
- A fixed US$ stop does not stay the same size. US$5.27 was an average day in 2005; in 2024 it was US$33.04. Set stops as a % of price or as a multiple of a recent range.
- Test the quiet years and the wide years. A setting tuned on 2018 (0.966% a day) meets days 2.98 times wider in a year like 2008.
5. Limitations
Limits
- One data source. Not checked against a second feed. Other brokers and futures data give slightly different highs and lows.
- Broker day. The daily bar follows the broker's server day, not a fixed UTC day. Note 10 uses a 22:00 to 22:00 UTC day built from hourly bars, so its day ranges differ a little.
- Spread is not included. Bar prices do not include the spread or any slippage.
- The past only. The data ends in 2024 for this note. The record day on 30 January 2026 is not in it. Every number here comes from our own script.
Appendix
Related notes
- The 20 largest single-day moves and how often a day closes 2%, 3% or 5% away: XAUUSD largest daily moves: 20 years of gold data.
- Which hours and sessions make the daily range: XAUUSD average range by hour and session (Asia, London, New York).
- The direction side of the same months, which months gold rose or fell from 2005 to 2024: Gold seasonality by month: 20 years of XAUUSD returns.
FAQ
What is the average daily range of gold (XAUUSD)?
From 3 Jan 2005 to 31 Dec 2024 the mean daily high-low range of XAUUSD was 1.520% of the previous close, or US$19.78. The median day was 1.279% (US$16.65). Half of all days fell between 0.925% and 1.819%.
How has gold's daily range changed over the years?
In US$ it grew from a mean of US$5.27 a day in 2005 to US$33.04 in 2024, mostly because the gold price rose. As a % of price it moved between 0.966% in 2018 and 2.878% in 2008.
Which month has the widest gold daily range?
March, with a mean range of 1.629% (1.07 times the average day). July was the quietest at 1.408% (0.93 times). The gap between months is much smaller than the gap between years.
How often is a gold day twice its average range?
On 4.0% of days (204 of 5,138, about one day in 25) the range was at least 2 times the average range of the 14 days before. 3 times or more happened on 0.7% of days, about one in 147.
References
- XAUUSD daily data: Hugging Face, ZombitX64/xauusd-gold-price-historical-data-2004-2025 (file XAU_1d_data.jsonl, revision ff6469dc, downloaded 18 Oct 2026). The dataset card names the Kaggle dataset "novandraanugrah/xauusd-gold-price-historical-data-2004-2024" as its source.
Declarations
- Author: Huyen X. Vu, Quantex Research, AlgoTrade Pte. Ltd., Singapore (ORCID 0000-0002-2209-842X).
- Data availability: Daily XAUUSD prices from a public dataset (Hugging Face, ZombitX64/xauusd-gold-price-historical-data-2004-2025), as described in the Method section. Our calculation files are available on request.
- Use of AI tools: Data processing, analysis code and drafting were assisted by AI tools (Claude, Anthropic). The author checked the results and is responsible for the content.
- Conflict of interest: The author is the founder of AlgoTrade Pte. Ltd., which develops and runs systematic trading models.
- Funding: No external funding.
- Peer review: Not yet peer reviewed. Comments are welcome at info@quantex.com.sg.
- How to cite: Vu, H. X. (2026). XAUUSD average daily range: 20 years of gold data by year and month. Quantex Research, AlgoTrade Pte. Ltd. https://quantex.com.sg/research/2026-10-18-xauusd-average-daily-range/