The short answer
We tested two common ways to trade the gold weekly open on 104 weeks of hourly COMEX gold data (October 2024 to October 2026):
- As a bias (long for the week if Monday’s 10:00 am New York close is above the weekly open, short if below): 96 trades, −$827 per ounce. It lost.
- As a retest level (wait for a move of half a daily ATR away from the open, then place a limit order back at the open, stop half an ATR beyond, target one ATR): 53 trades, 43.4% winners, +$781 per ounce, profit factor 1.69. All six variants we ran were positive.
The retest result is encouraging but fragile: 53 trades is a small sample, and most of the profit came in 2026, an exceptionally volatile year. In calendar 2025 the main version lost $46.
What the weekly open is
Gold futures trade from 18:00 New York time on Sunday to 17:00 on Friday, with a short daily break. The weekly open is the very first price of that week. Traders watch it for two reasons:
- Bias. If buyers are in control, price spends the week above its opening level; if sellers are, below. In our window, 61 of 104 weeks closed above their opening price.
- Reference. Price often returns to the weekly open during the week, and some traders expect a reaction there. It is a cousin of the midnight and daily opens used in Power of 3.
In GMT the weekly open is 22:00 in US summer and 23:00 in US winter. Our code anchored it to 18:00 New York time for every week, so the US clock change is handled. Your broker’s XAUUSD chart may open a few minutes later, at a slightly different price.
Exact rules
Version A: weekly open bias
| Rule | Detail |
|---|---|
| Decision | Close of the Monday 10:00 to 11:00 am New York hourly bar |
| Direction | Long if that close is above the weekly open, short if below |
| Exit | Close of the last bar of the week (Friday 17:00 New York) |
| Stop | None (time exit only); R measured as one daily ATR(14) |
Version B: weekly open retest
| Rule | Detail |
|---|---|
| Set-up window | From Monday 03:00 New York time to the end of Wednesday |
| Direction | The first move of 0.5 x daily ATR(14) away from the weekly open sets the side: above means look to buy, below means look to sell |
| Entry | Limit order at the weekly open, valid until the end of Thursday |
| Stop | 0.5 x daily ATR beyond the weekly open (about $47 on average) |
| Target | 1.0 x daily ATR from the weekly open (2R) |
| Exit | End of the week if neither is hit |
| Trades | At most one per week |
Daily ATR comes from the days before the week starts, so no future data is used.
How we tested it
Data. COMEX gold futures (GC=F) 1-hour bars from 30 October 2024 to 30 October 2026 (104 trading weeks), plus daily bars since January 2015 for a daily version of the bias rule (Monday’s close versus the weekly open, held Tuesday to Friday). Execution. Market orders at the next bar’s open; limit orders at the level, or at the open if price gaps beyond it. On a limit-fill bar the stop counts if reached and the target does not. Cost. $0.30 per ounce per round trip. Results in US dollars per ounce, 1 ounce per trade.
The results
Version A: bias
| Test | Weeks | Win rate | Avg per week | Profit factor | Net | Max drawdown | Split |
|---|---|---|---|---|---|---|---|
| Hourly, Oct 2024 – Oct 2026 | 96 | 49.0% | −$8.61 | 0.80 | −$826.70 | $1,295.90 | Year 1 −$1,058; Year 2 +$232 |
| Daily, 2015 – 2026 | 554 | 48.6% | −$2.25 | 0.87 | −$1,247.10 | $2,168.70 | 2015–20 +$568; 2021–26 −$1,815 |
Version B: retest, with variants
| Variant (trigger / stop / target, in daily ATR) | Trades | Win rate | Avg per trade | Profit factor | Net | Max drawdown | Year 1 | Year 2 |
|---|---|---|---|---|---|---|---|---|
| 0.5 / 0.5 / 1.0 (main) | 53 | 43.4% | +$14.74 | 1.69 | +$781.20 | $355.80 | +$89.20 | +$692.00 |
| 0.3 / 0.5 / 1.0 | 68 | 44.1% | +$15.83 | 1.72 | +$1,076.50 | $277.80 | +$152.10 | +$924.40 |
| 1.0 / 0.5 / 1.0 | 18 | 55.6% | +$24.56 | 2.71 | +$442.00 | $118.30 | +$126.70 | +$315.30 |
| 0.5 / 0.5 / 0.5 | 53 | 56.6% | +$11.51 | 1.69 | +$610.20 | $297.50 | +$189.40 | +$420.80 |
| 0.5 / 0.5 / 1.5 | 53 | 41.5% | +$19.39 | 1.90 | +$1,027.60 | $356.70 | +$167.40 | +$860.30 |
| 0.5 / 1.0 / 1.0 | 53 | 47.2% | +$7.54 | 1.24 | +$399.80 | $552.70 | −$213.50 | +$613.30 |
By calendar year, the main version made $62 in the last two months of 2024, lost $46 in 2025 and made $765 in 2026.
What the results say, in plain English
- The weekly open is not a reliable bias. Being above it on Monday morning said little about where gold would be on Friday. The daily version since 2015 made money in the quiet years to 2020 and lost after, so even that is unstable.
- As a level, it behaved better. Gold often moved away from the weekly open early in the week and came back to it, and buying or selling that return made money in every variant we tried.
- Robust to settings, not to time. All six variants were positive overall, which is a good sign that the result is not one lucky parameter. But all six made most of their money in year 2, and the 2025 calendar year was roughly flat.
- Small sample. 53 trades is about one a fortnight. Ten unlucky trades would change the picture.
- Costs barely matter here. With trades worth tens of dollars, the result at $0.60 per ounce was $765, almost unchanged.
When the weekly open fails
- Trend weeks. In strong one-way weeks gold leaves the weekly open on Monday and never returns: no trade, or a stop-out if it returns only briefly.
- News-driven Mondays. Weekend headlines can gap gold away from Friday’s close, so the weekly open sits far from the action.
- Low-volatility periods. When the daily ATR shrinks, the half-ATR stop is tight and normal noise can hit it.
- Broker differences. If your platform’s week opens later or at a different price, your level is different from ours.
How to use the weekly open in practice
- Draw it every Sunday, alongside the previous week’s high and low and the daily pivots from our pivot points guide.
- Wait for price to leave it. The retest idea needs a move away first.
- Look for confluence. A weekly open that sits near a support or resistance zone is a stronger level.
- Respect the week’s timeline. Late-week retests had less time to reach the target before Friday’s close.
Risk sizing
The main retest stop averaged about $47.47 per ounce. Risking $100:
- $100 / ($47.47 + $0.30) is about 2.1 ounces (0.02 lots).
- The worst losing streak was 7 trades, about $700 at that size, and the maximum drawdown per ounce ($356) would have been about $750.
Use the position size calculator with each week’s ATR, since the stop changes week to week.
Test it yourself
Use your broker’s weekly opening price, not ours, and record every week for at least six months: when price first moved away, how far, and whether it returned. Then code version B and run it with your costs. If you get a similar result in 2025 as well as 2026, the idea is worth more of your attention. Every other test in this series is on the strategies tested hub.
Practise it with other traders
The weekly open is a good level to study as a group: everyone can draw the same line on Sunday and compare notes on Friday. In the XAUUSD Global community, members share weekly levels and review how price reacted. The first month is free, then $19.99 a month.
Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.
Frequently asked questions
What is the weekly open in gold?
It is the first traded price of the week. For COMEX gold futures the week opens at 18:00 New York time on Sunday, which is 22:00 GMT in US summer and 23:00 GMT in winter. Your broker’s XAUUSD chart may open a few minutes later and at a slightly different price.
Is price above the weekly open bullish for gold?
Not reliably. In our test, going long for the rest of the week when Monday’s 10:00 am New York close was above the weekly open (short when below) lost $827 per ounce over 96 weeks. A daily version since 2015 lost $1,247 over 554 weeks.
Does trading a retest of the weekly open work on gold?
In our two-year hourly test, yes, so far: waiting for gold to move half a daily ATR away from the weekly open and then placing a limit order back at the open made $781 per ounce over 53 trades, profit factor 1.69. Every variant we tried was positive, but most of the profit came in 2026 and the sample is small.
What stop should I use when trading the weekly open?
Our retest version used half of the daily ATR(14) beyond the weekly open, which averaged about $47 per ounce over the test. A full ATR stop was still positive but much weaker.
Can I use the weekly open with daily pivots and previous-week levels?
Yes. The weekly open is one of several reference levels; traders often look for it to line up with the previous week’s high or low, a pivot level or a round number.
Why did the weekly open retest do so well in 2026?
Gold’s daily range in 2026 was very large, so moves away from the weekly open and back were large too, and an ATR-sized target was reached more often. In the calmer calendar year 2025, the main version roughly broke even.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.