What Is the New Week Opening Gap? The Two Clocks
The new week opening gap: Friday's last price against the week's first. Why the 6:00 PM mark is an hour late on spot forex, and 1,190 weekends measured.
The new week opening gap, or NWOG, is the distance between the last price traded before the weekend and the first price traded when the new week opens, with no prints in between. It is two horizontal lines on a chart: Friday's closing price and the new week's opening price.
Which two lines you get is a choice of clock. Most published instructions tell you to mark the new week at 6:00 PM New York time. On spot forex that is an hour after the week has already opened, and on our own EUR/USD data the two instructions disagree about whether the gap is up or down on 242 of 1,190 weekends.
What is the new week opening gap?
It is the gap left by the weekend: Friday's last traded price, the new week's first traded price, and nothing traded between them.
Candela Charts gives a short definition of the version most pages teach, calling it "the price gap between the closing price on Friday at 4:59 PM EST and the opening price on Sunday at 6:00 PM EST". ICT Traders states the part that matters most once the gap exists, which is what the two prices become: "If Sunday opens below Friday's close, Friday's closing price becomes the NWOG high and Sunday's opening price becomes the NWOG low", and the mirror for a gap the other way.
Inner Circle Trader's tutorial explains why the space is empty: nothing trades over the weekend, so the band between the Friday close and the Sunday open holds no liquidity at all. The page calls that band a void, and puts the word in quote marks itself.
That emptiness is the difference between this gap and a fair value gap, and ICT Traders puts it in a line: "an NWOG is a genuine gap in trading activity". A fair value gap is a shape inside continuing trade. This one is a real break in it. How far to push the distinction is contested rather than settled: the same Inner Circle Trader page calls the NWOG "a real fair value gap", so treat the family resemblance as an open question.
What time is Friday's close and the new week's open on the thing you trade?
On spot forex, Friday 16:59 New York time and Sunday 17:00 New York time. On an index it is several hours earlier on Friday and around 18:00 on Sunday. Those are not the same clock, and most published instructions do not say which instrument theirs belongs to.
We measured it on our own data, 1-minute candles aggregated to hour bars. Spot EUR/USD from 2004-01-01 to 2026-08-31, which is 1,190 weekend breaks, and GER40, our German 40 index, from 2013-01-02 to 2026-08-31, which is 729.
| Spot EUR/USD, 1,190 weekends | GER40, 729 weekends | |
|---|---|---|
| Last print before the weekend | Friday 16:59 ET on 1,176 | Friday 16:14 ET on 259, 16:00 on 172, 15:59 on 169 |
| First print of the new week | Sunday 17:00 ET on 1,084, 17:01 to 17:04 on 86 more | Sunday 18:02 on 173, 18:03 on 148, and on 124 weekends not until Monday 02:00 |
| Length of the break, median | 48.02 hours (min 22.00, max 72.02 over holidays) | 50.05 hours (min 20.93, max 130.68) |
So the 6:00 PM instruction fits the index, not spot forex. ICT Killzone gives forex its own clock, saying "the Friday close occurs at approximately 5:00 PM ET and the Sunday open at approximately 5:00 PM ET" for "a clean 48-hour gap", and our median of 48.02 hours says that is the right shape.
Inner Circle Trader prints both clocks on one page without saying they differ. Its instruction tells you to mark 4:59 PM and 6:00 PM, while further down the same page it names the exchange clock behind the second figure: a "weekly close at 17:00 ET Friday and reopen at 18:00 ET Sunday". That is a CME Globex schedule, not a spot forex one. ChartLabs carries spot forex, indices, commodities and crypto, so the second clock is the one our index data follows and the first is the one our forex data follows.
The sources do not even agree on which day the new week opens. TradingFinder tells you twice to mark "the opening price on Monday" and twice to use the Sunday open, on the same page.
FXNX states the problem in terms of the broker rather than the instrument: "If your broker doesn't use the New York close", then "your gaps will not align", and so "Always use a broker with a true NY 5:00 PM EST close". Check your feed, and check which instrument you are on.
Does the 5 PM or the 6 PM mark change the gap?
Yes, and by more than the rounding you would expect. On the same 1,190 EUR/USD weekends the two instructions differ by a mean of 7.34 pips.
They also disagree about direction. The 18:00 ET mark gives the bigger of the two gaps on 46.3% of weekends, and the two marks put the gap on opposite sides of Friday's close on 242 weekends, which is 20.3%.
On GER40 the same test gives a mean difference of 0.06 index points and a sign disagreement on 3 weekends, 0.4%. The clock only matters where it is wrong for the instrument. Mark an index at 18:00 and you have lost almost nothing. Mark spot forex at 18:00 and one weekend in five has you drawing the gap the wrong way round.
How big is a new week opening gap, normally?
Small. On EUR/USD, Friday's last print to the week's first print: median 6.1 pips, mean 12.1, p90 29.6, max 192.9, with 568 gaps up, 613 down and 9 flat.
On GER40 the same measurement gives a median of 22.2 index points, p90 96.4 and a maximum of 747.3.
ICT Killzone publishes thresholds, and our figures let you check them. It says "If the gap is at least 10 NQ points (5 pips for EUR/USD), a tradeable NWOG exists." NQ is an index futures contract, a market we do not carry, so the EUR/USD half is the one to read here, and that 5-pip floor is cleared on 670 of 1,190 weekends, 56.3%. It calls a meaningful forex NWOG 10 to 30 pips, and that band holds 25.0% of weekends. Only 3.9% of weekends gap more than 50 pips.
So on a little over half of weekends there is a gap this source would call tradeable, and on a quarter of them a gap it would call meaningful. The rest of the time there is barely a gap to mark.
Does the new week opening gap get filled?
On our data, usually, and usually fast. Of the 1,181 EUR/USD weekends with a gap of at least 0.1 pip, price traded back to Friday's closing price inside the same week on 1,124 of them, 95.2%, and touched the midpoint between the two prices on 1,164, 98.6%.
Median time to the full fill: 1.0 hour. The 25th percentile is 0.0 hours, meaning the fill happened inside the opening hour. The 90th percentile is 16 hours. By Monday 17:00 ET, 91.3% had filled; by Tuesday 17:00 ET, 93.1%.
That is a description of what price did, on one instrument, over one window, with one definition of "filled". It is not a strategy result and it is not a claim that trading the gap works. We did not test an entry, a stop or a target, and we are not going to publish one.
Two things follow from it anyway. First, a level whose median fill is the first hour of the week is not a level that stays relevant for weeks. Second, we are not correcting anyone else's fill figure with ours, because a different measurement from a method we cannot see is not comparable. Ours is EUR/USD, 1,181 weekends from January 2004 to August 2026, price trading back to Friday's closing price inside the same week.
ICT Killzone gives the one deadline anybody publishes: "If the NWOG has not filled by Tuesday's close", it calls what is left a failed NWOG. On our data that describes the remaining 6.9%.
Do bigger gaps fill less often?
Yes. The fill rate falls steadily with gap size, and the time to fill climbs with it.
| Gap size, EUR/USD | Weekends | Filled in the same week | Median hours to fill |
|---|---|---|---|
| Under 5 pips | 511 | 98.4% | 0 |
| 5 to 10 pips | 258 | 97.3% | 1 |
| 10 to 30 pips | 297 | 92.6% | 3 |
| 30 to 50 pips | 68 | 83.8% | 15 |
| Above 50 pips | 47 | 80.9% | 18 |
GER40 behaves the same way and more sharply at the top: of 185 weekends gapping more than 50 index points, 73.5% filled inside the week, with a median of 10 hours.
What is the new day opening gap, and can you mark one on spot forex?
The new day opening gap, or NDOG, is the same idea applied to a daily boundary instead of a weekly one. On spot forex, measured as it is published, there is no gap there to mark.
WritoFinance gives the definition everyone else only gestures at, calling it "the price gap between the closing price at 5:00 PM and the opening price at 6:00 PM (New York time)", and explains the mechanism as a daily halt: this happens "because Foreign Exchange Market halts trading for one hour daily from Monday through Thursday".
We checked that halt on our own feed and it is not there. The 17:00 to 18:00 ET hour has minute bars on 4,725 weekdays, a mean of 58.7 of its 60 minutes, and the only weekdays missing it are the 1,187 Fridays, where the weekend break is supposed to be.
Measured exactly as published, over the 4,676 Monday to Thursday windows with a price at both edges: the 5:00 PM price against the 6:00 PM price is a median of 2.3 pips apart, mean 4.1, p90 9.5. Every one of those windows has minute bars inside it, a mean of 58.8 of 60 minutes, and price traded through both edges of the supposed void in 95.7% of them.
There is a real daily rollover on our forex feed, and it is tiny: across 4,709 weekday rollovers the median step is 0.20 pips, and not one of them has a break in the clock.
So on a feed that does not stop at 17:00 ET, the NDOG instruction does not mark a void. It marks one hour of ordinary trading. This object belongs on instruments with a real session close, which on our data means the index side rather than spot forex, and you should check your own feed for an actual break before drawing one.
How do you mark a new week opening gap on a chart?
Set the chart to the clock you intend to work in, find the last bar before the weekend and the first bar after it on your own feed, and mark both prices plus the midpoint between them.
WritoFinance gives the most practical recipe for the midpoint, which is to draw a Fibonacci tool with 0, 0.5 and 1.0 from the gap's low to its high, so the 0.5 line lands in the middle. Candela Charts explains why that middle line gets the attention. It calls the 50% level the consequent encroachment, "the 50% retracement level of the NWOG", and "one of the most reactive price levels". On our data price touched it on 98.6% of gapped weekends.
In ChartLabs the pieces for this are: one timezone preference, set once, that charts, trade times and every analytics breakdown display in, which is the step that decides what "Friday 16:59" reads as on your screen; drawing tools for a horizontal line, a horizontal ray and a Fibonacci retracement; and one settings dialog covering colours, line styles and text for every drawing, with your own defaults saved to reuse. Right-click a drawing and Apply defaults resets it to your saved look, and switching on Remember last used starts new drawings from the style you used last. Drawings and sub-panel indicators save automatically as you work, so the lines are still there when you come back to the week.
One thing the timezone setting does not do is move the bars. It changes the times shown, not how bars are grouped: daily bars cut at 00:00 UTC, which is where ChartLabs cuts them. So set the chart to New York time to read Friday 16:59 off the screen, not to shift where our days begin.
Where the published instructions leave you
The instructions agree on the shape: two prices, nothing traded between them, and a midpoint worth marking. They differ on the clock, and the clock is almost always published without an instrument attached to it. That is why the same article can be right for an index chart and an hour out on a forex one.
So take the shape from them and the times from your own feed. Candela Charts has the definition and the two marks in the fewest words. ICT Killzone is the one with a forex clock and published size thresholds. FXNX is the one that explains why your broker's gaps may not line up with the article's, and it also notes that "there is no 100% guarantee" a gap fills.
How do you check this on your own data?
Replay the week after the gap and write down four things per weekend.
- Set your chart to New York time, or to your own zone with daylight saving applied. Not to a fixed GMT offset.
- Find the last bar before the weekend and the first bar after it on your own feed, rather than at a time somebody published. Mark both prices and the midpoint.
- Record the gap size in pips or points.
- Replay forward bar by bar and record whether price returned to Friday's closing price, how many hours that took, and whether the midpoint was touched before the far edge.
- Keep the weekends with no real gap in your sample. On our data that is a large minority, and dropping them is how a fill rate gets inflated.
ChartLabs is built for step 4. You replay historical charts bar by bar and take practice trades as if the market were live, so the gap is marked before you can see what happened next. Session setup lets you pick a date range by hand, or draw random dates by length and period if you do not want to choose the week yourself, and replay runs at a speed you set. Every trade you log captures a snapshot of the chart as it was at entry, so you can look back at a gap trade and see exactly what you saw. The data is sourced from Dukascopy, cleaned to remove anomalies and built into 1-minute candles, which is the same data behind every number above.
New to running a test like this? Our beginner's guide to backtesting a strategy covers the logging, and forex backtesting covers the session and spread details that matter on EUR/USD.
The short version
- The NWOG is Friday's last traded price against the new week's first, with nothing traded in between.
- The two marks are a clock choice. On spot forex the week's first print is Sunday 17:00 ET on 1,084 of 1,190 weekends, not 18:00.
- Marking at 18:00 instead of 17:00 changes the gap by a mean 7.34 pips and flips its direction on 20.3% of weekends. On GER40 the same choice costs almost nothing.
- The median gap is 6.1 pips on EUR/USD. Just over half of weekends clear the one published 5-pip floor.
- 95.2% of gapped weekends saw price trade back to Friday's close inside the same week, at a median of one hour. That is a description, not a strategy.
- Bigger gaps filled less often and more slowly, from 98.4% under 5 pips down to 80.9% above 50.
- The new day opening gap does not exist on our spot forex feed. All 4,676 windows we measured have price trading inside them.
To check a gap on your own chart, with the marks drawn before you can see the outcome, start a free trial of ChartLabs and replay the week bar by bar.
Related Topics
Related
Backtesting a Trading Strategy: A Beginner's Guide
What backtesting is, why it matters, and how to run one step by step: defining your rules, choosing clean historical data, stepping through the chart trade by trade, logging results, and reading the metrics that tell you whether your strategy has a real edge.
Forex Backtesting: Test Your Strategy on Real Data
A practical guide to backtesting forex strategies: why session dynamics, spread variation, and pair-specific behavior change your results, what clean historical data actually means, and a five-step process for testing a strategy before you risk real capital.