ICT New Day Opening Gap (NDOG) Explained

A daily opening gap can look like an obvious destination. Yesterday’s session ended at one price, the next session began somewhere else, and a rectangle between those prices seems to explain every later reaction. The difficult part is deciding whether you have marked a genuine session discontinuity, whether that particular market actually closed, and whether the rectangle improves a trade decision made before the outcome.

The ICT New Day Opening Gap, usually shortened to NDOG, provides a way to study that transition. It becomes more useful when you separate three questions: how the reference is constructed, what price does around it, and whether an executable setup offers sufficient reward for its risk. A correct drawing can still produce a losing trade. An attractive midpoint can still sit in the middle of directionless trading.

For traders in Ghana, one additional detail matters: the New York clock changes seasonally, while Ghana remains on GMT. There is also a market distinction that many explanations miss. A futures session break and a spot forex broker’s rollover are not interchangeable. This guide uses a complete hypothetical Euro FX futures example to preserve the original session logic, then shows how to label a separate EUR/USD broker adaptation honestly.

1. What is the ICT New Day Opening Gap?

In his New Day Opening Gap lesson, ICT illustrates the difference between a session’s closing print and its reopening print after the daily futures break. A reopening above the close creates an upward gap; a reopening below it creates a downward gap. The concept uses those two prints as fixed references. It does not use the highest wick before the close or the lowest wick after the open.

ICT also shows older daily gaps extended into later sessions and presents them as supporting context for ideas developed with other price arrays. That distinction matters: the gap is one location within a broader reading of price delivery. It should help you refine an existing hypothesis rather than supply a bullish or bearish conclusion on its own.

The three prices to record

Call the preceding close C and the next opening print O. The upper boundary is the larger of C and O. The lower boundary is the smaller. The midpoint is their arithmetic average. Once recorded from the correct feed, these prices remain fixed even if subsequent candles travel far beyond them. Changing the rectangle to capture a later turning point changes the observation you are studying.

For a hypothetical close of 1.08640 and opening print of 1.08720, the upper boundary is 1.08720 and the lower boundary is 1.08640. The width is 1.08720 minus 1.08640 = 0.00080, equivalent to 8.0 EUR/USD-style pips. The midpoint is (1.08640 + 1.08720) / 2 = 1.08680. These are measurement calculations, not an entry, stop, or profit target.

ICT refers to the midpoint as consequent encroachment in his 2024 mentorship lecture on opening-gap references. Treat it as a precisely reproducible location. A line drawn at 50% does not reveal how many orders remain there. Nor does a brief touch prove that an institution bought, sold, or completed an inventory adjustment.

2. Why the futures and spot forex distinction matters

The ordinary CME FX futures schedule includes a daily 60-minute break beginning at 4 PM Chicago time, equivalent to 5 PM New York, with reopening an hour later. That creates a defined interruption in that venue’s trading session. Check the applicable contract and holiday schedule before marking a particular date. The schedule is documented in CME’s FX futures specifications.

Spot forex is a different market structure. A retail chart represents a particular broker or data provider, and available dealing hours can differ. For example, OANDA Global Markets publishes a six-minute daily forex break from 16:59 to 17:05 New York time, subject to its instrument and scheduling notes. That is not a one-hour 5 PM to 6 PM closure. This is a concrete example of why you must inspect the venue instead of importing another market’s clock.

If your EUR/USD feed quotes continuously between 5 PM and 6 PM, subtracting those two prices measures an hour’s price change. Calling every price between them untraded would be incorrect. You may choose to test those endpoints as a time reference, but label the study accordingly. If your broker actually pauses quotes, you can study the final pre-pause quote and first post-pause quote as a broker rollover discontinuity. That remains a separately defined adaptation.

The practical benefit of this precision is comparability. Two traders can disagree about a gap simply because one uses futures last trades and the other uses spot bid quotes. Before arguing about which drawing is correct, compare the instrument, contract month, quote type, session template, time zone, and opening timestamp. Matching a label is not enough if the underlying measurements differ.

Why a 6E level is not automatically a spot EUR/USD level

Euro FX futures and spot EUR/USD both describe dollar prices for euros, but they are different instruments. CME’s FX Product Guide lists 6E as a 125,000-euro contract with a 0.00005 outright tick. The existence of a futures-to-spot basis is also explicit in CME’s FX Link discussion. A futures price should therefore stay attached to its futures contract when you write an exact order plan.

A forex trader may observe futures behaviour as context, then identify a separate execution setup on the actual spot feed. Do not paste an exact 6E midpoint into a spot order ticket and assume equivalence. If you use a conversion method, document and test it, including how the basis changes. The worked examples below deliberately keep each market’s levels and risk calculations separate.

3. Why NDOG matters in an ICT order-flow reading

An opening discontinuity connects two observations separated by time. That makes it a useful reference for asking how later delivery treats the transition: does price approach from above, cross through, pause around the midpoint, or leave decisively? The value of the exercise is the consistency of the questions. You can record the same observations on a strong trending day and on a frustrating sideways day without changing the definition after the fact.

Within an ICT or SMC framework, relate the gap to the current draw on liquidity and to nearby price arrays. Suppose your analysis already identifies an external high above the market and a sell-side sweep followed by upward displacement. An NDOG near the retracement can describe where you are waiting for evidence. It does not replace the sweep, establish the destination, or provide proof of institutional intent.

This is different from counting every reaction to a horizontal line as confirmation. Ten touches might represent repeated two-way trading around a reference, with no useful directional opportunity after costs. A more informative record describes the approach, the displacement away, the structure affected, the retracement available, and the distance to the next opposing array. Each observation answers a decision question.

Be careful with the phrase institutional order flow. OHLC candles show prices reported by the selected feed. The narrative that a particular gap is influencing delivery is an interpretation of those observations. It is not a view of every bank’s positions or a direct measurement of unfilled orders. The BIS report on FX execution algorithms describes a fragmented market in which execution and liquidity are distributed across venues. A retail rectangle cannot expose that whole system.

4. How to identify and label an NDOG on a chart

1. Identify the exact market. Write the futures contract month or the broker’s full symbol. For a broker adaptation, add whether the chart uses bid, ask, or midpoint prices. This prevents you from combining unlike observations later.
2. Confirm the session boundary. Use the exchange or broker schedule for that date. Inspect whether the platform hides extended hours. A blank region caused by display settings is not evidence that the instrument stopped trading.
3. Find the last valid pre-break print. On a one-minute futures chart, this is normally the close of the final bar before the break. If the chart labels bars by opening time, a 16:59 candle can contain the final trading before 17:00.
4. Find the first valid reopening print. Use the open of the first actual session bar. If the first valid trade occurs slightly after the scheduled reopening, record that timestamp. Do not manufacture a price for a missing bar.
5. Calculate the boundaries and midpoint. Use the two prints, independent of direction. Preserve their original identities as close and open alongside the sorted upper and lower prices.
6. Extend the reference and log its state. Record first touch, midpoint touch, full traversal, and later revisits separately. Freeze the original numbers before studying the subsequent candles.

Do not silently substitute the settlement price for the final session trade. Settlement is a separate exchange valuation reference and may differ from the closing print used in this construction. Similarly, a daily candle displayed under a custom session template may not expose the exact close you need. Inspect intraday data around the boundary instead of assuming every platform’s daily close field means the same thing.

Dates also deserve a deliberate rule. A gap created on a Wednesday evening in New York may belong to a platform session labelled Thursday. Store both the New York opening date and the platform trading date. A label such as “Wednesday 18:00 NY, Thursday session” is more informative than an unexplained weekday. Use the same convention in screenshots, journal rows, and saved chart templates.

The first diagram isolates construction. The blank time interval represents the specified futures break; the coloured rectangle represents the price difference between the two session prints. These are different dimensions. Candles after reopening can enter the rectangle immediately, so an opening gap does not imply that its entire price interval will remain untouched for the rest of the session.

5. NDOG times in Ghana, including daylight saving

Ghana uses GMT throughout the year. New York uses UTC minus five during standard time and UTC minus four during daylight time. The conversion therefore changes even though your Ghanaian clock does not. Use a chart zone that follows America/New_York automatically, then translate the session events to GMT. The Ghana time-zone record and New York time-zone record show the underlying clock difference.

The familiar 7 to 10 AM London window and noon to 3 PM New York window in Ghana are the standard-time conversions for those New York-anchored conventions. If your plan instead follows a London-local opening event, track that clock separately. London and New York do not always change clocks on the same dates. The session convention you actually trade should be written next to the time window.

There is no requirement to sit awake until the gap forms. A Ghana-based trader preparing for London can inspect the completed session transition in the morning, mark it, and record what happened overnight before considering an entry. That may fit work or study commitments better than treating every new rectangle as an evening trading opportunity. The examples here use winter GMT times explicitly so their event order is reproducible.

6. NDOG versus NWOG, FVG, and the midnight open

These references can overlap without becoming identical. An FVG is constructed from candle relationships; an NDOG is constructed from session endpoints. The midnight open is a single price, not a two-boundary gap. The weekend interval belongs in a separate NWOG study, even when an indicator uses the same drawing style for both.

Use each reference for a distinct job. In a trade plan, the NDOG may identify the area to monitor, an FVG may define the proposed retracement entry, and an external swing may define the destination. If you count all three as independent proof of a profitable trade, you overstate what the chart has shown. Several labels can describe substantially the same price event.

Consider a midpoint inside an FVG only 0.00010 away from its centre. That alignment makes your map more specific, but it does not automatically triple the evidence. A useful journal compares setups with and without the alignment under the same entry rules. Otherwise, the most decorated chart can feel strongest simply because it contains the most vocabulary.

7. How to trade around an NDOG without making it the signal

Start with a destination and a falsifiable scenario

Before selecting a gap, identify the price objective your analysis is investigating. It might be an external high, an external low, or another pre-existing array. State what would weaken the idea. For example, if you expect upward delivery after a sell-side raid, repeated downward expansion through the raid extreme is evidence against the scenario. A nearby NDOG cannot make that contrary evidence disappear.

Next, define the observation area. You can study a boundary, the midpoint, or an overlapping FVG, but decide which one matters before the reaction. Choosing the upper edge after a midpoint failure, then choosing the lower edge after another failure, turns one planned setup into unlimited attempts. In this guide’s example protocol, there is one eligible entry attempt per specified sequence.

Require observable displacement and completed structure

For teaching consistency, the examples use a candle close beyond a previously identified short-term swing, followed by a completed three-candle FVG. This close requirement is an explicit rule of the example, not a claim that every ICT model has an identical trigger. The swing must exist before the displacement candle breaks it. The third FVG candle must finish before an order relying on its final high or low becomes eligible.

A wick into the gap is therefore an observation, not the whole entry. Ask whether the subsequent movement actually changes the short-term delivery you were monitoring. Also check whether the resulting retracement is still available at an acceptable price. A textbook sequence that leaves immediately without returning is a missed trade, not permission to move the entry nearer the target after the fact.

Put the stop beyond the event that invalidates the setup

A stop placed one tick outside the NDOG simply because the rectangle ends there may ignore the actual structural risk. In the bullish example, the key failure point is the raid low, so the stop sits beneath it with a stated buffer. The gap midpoint defines the chosen entry location; it does not automatically define invalidation. These are separate decisions with separate purposes.

Calculate reward from the actual planned entry to a preselected destination. Include realistic execution assumptions and reject plans whose remaining distance is too small. Decide what cancels an unfilled order, such as a fresh structural failure, the end of the study window, or a scheduled event the plan excludes. Once cancelled, do not keep the old order active while mentally switching to a new scenario.

8. Complete worked example: a bullish 6E NDOG retracement

This is a constructed Euro FX futures sequence, not a historical trade or a claim about current prices. Assume a liquid contract month, ordinary trading hours, no contract roll between observations, and winter Ghana time. Every quoted 6E price uses the 0.00005 outright tick. The chart shows last-traded prices; order fills and the costs below are explicit assumptions.

The preceding evening’s final pre-break print is 1.08640 at the end of the session before 10 PM Ghana time. The next opening print is 1.08720 at 11 PM. Mark the resulting NDOG from 1.08640 to 1.08720, with consequent encroachment at 1.08680. An already visible external high at 1.08860 is the bullish destination under investigation. The proposed exit will be one tick beneath it at 1.08855.

During the following morning, an existing intraday low at 1.08635 is raided to 1.08625 in the five-minute candle ending at 7:40 AM GMT. Price then recovers. A short-term swing high at 1.08695, already formed before the raid, is broken by the 7:50 candle’s close at 1.08710. The example’s bullish shift condition is now satisfied. The sweep itself did not authorize a purchase.

The 7:45 candle has a high of 1.08665. The 7:55 candle closes with its low at 1.08695, leaving a completed bullish FVG between 1.08665 and 1.08695 around the 7:50 displacement. Its midpoint is (1.08665 + 1.08695) / 2 = 1.08680, exactly the older NDOG midpoint. Only after 7:55 is the buy-limit order eligible under this protocol.

Assume a retracement fills the buy at 1.08680 during the 8:00 candle. The stop is 1.08610, three ticks below the raid low of 1.08625. The target remains 1.08855. Neither the stop nor the target is moved to improve the ratio. In the illustrative path, price subsequently trades through the target region during the 8:35 candle before visiting the stop, and the exit is assumed filled at the specified target.

The dollar conversion follows directly from the 125,000-euro contract: 0.00005 x 125,000 = $6.25 per tick. The $12.25 cost allowance is an illustrative assumption, not a broker quote or a maximum possible loss. Here it is additional to the idealised entry and exit prices. Do not add the same spread or slippage again if your recorded fills already include it.

Position size is constrained by whole contracts. For a hypothetical $20,000 account using a self-imposed 0.5% risk budget, the budget is $100. Dividing $100 by the estimated $99.75 loss per contract gives approximately 1.0025, so the plan permits one standard 6E contract after rounding down. An $80 budget would permit zero such contracts under these assumptions. Margin availability does not change that risk-budget arithmetic.

The favourable outcome is only one possible path. If the same filled position reached 1.08610 first, its planned gross loss would be $87.50 and its estimated loss with the stated allowance $99.75. If price never retraced to 1.08680, there would be no entry and no realised R result. A chart touching the limit price does not alone prove that a real resting order received a fill in the exchange queue.

9. A losing spot EUR/USD adaptation and its calculation

Now change markets explicitly. Assume a hypothetical broker pauses its EUR/USD quotes around rollover and reports a final bid of 1.09207 before the interruption and a first returning bid of 1.09183. The broker’s recorded boundary times are 16:58:59 and 17:05 New York on the preceding evening. This is a broker rollover-gap study, not an assertion that spot forex shared the futures one-hour closure.

The local quote gap is 1.09207 minus 1.09183 = 0.00024, or 2.4 pips. Its midpoint is 1.09195. During the next winter London morning, a bearish hypothesis targets an already identified low at 1.09029. Price first raids a local high to 1.09237. A five-minute close below the premarked swing at 1.09169 occurs at 8:05 AM GMT, and a bearish FVG from 1.09183 to 1.09199 is completed at 8:10.

The proposed sell is the FVG midpoint at 1.09191, which lies within the local rollover gap but differs from its midpoint. Assume a bid-side entry fill at 8:15. The stop is 1.09253, above the raid high, and the target is 1.09036, before the lower objective. This distinction prevents a misleading claim that every successful-looking overlap must be an exact midpoint match.

Risk: 1.09253 – 1.09191 = 0.00062 = 6.2 pips. Reward: 1.09191 – 1.09036 = 0.00155 = 15.5 pips. Gross R:R: 6.2:15.5 = 1:2.50. With an illustrative additional round-trip allowance of 0.9 pip, adjusted reward is 14.6 pips and adjusted risk is 7.1 pips. The adjusted ratio is 14.6 / 7.1 = 2.06R, rounded.

In this constructed losing path, the ask reaches the stop during the 8:25 candle, and assume the stop fills at 1.09253. The ideal price loss is 6.2 pips, or an estimated 7.1 pips including the stated allowance. A subsequent fall toward 1.09036 does not convert the stopped position into a winner. There is no second entry in this example because the protocol permitted one attempt.

For an illustrative 0.10 standard lot, assuming a 100,000-euro standard lot and a USD account, pip value is $1. The estimated loss is therefore $7.10 and the hypothetical net reward would have been $14.60. The planned ratio remains $14.60 / $7.10 = 2.06R. The actual assumed outcome is the loss. These numbers illustrate sizing and execution bookkeeping, not a recommendation to trade a particular account size.

For a short position, the entry and exit operate on different quote sides. MetaTrader’s trade execution explanation distinguishes selling at bid from buying at ask. A bid-only chart can therefore look clear of a stop even when the ask reaches it. Preserve spread and quote-side information when judging a rollover adaptation, especially because a narrow measured interval can be small relative to execution costs.

10. How to track old gaps and test whether they help

Use a small, consistent record rather than extending every historical gap indefinitely across the visible chart. A practical research template can show the latest five eligible daily gaps and archive older ones. Five is a manageable study choice here, not a proven optimum or a rule that makes the sixth gap irrelevant. If an older reference belongs to an active scenario, record why you retained it before the next reaction.

Give each gap separate state fields: untouched, boundary visited, midpoint visited, opposite boundary reached, and revisited after traversal. Define “filled” before using it in statistics. For example, a full traversal might require a last trade at the far boundary, while a spot study might require a particular quote side. Mixing these definitions creates a misleading fill percentage.

Do not erase the original reference after a fill. Retain the measurement in the journal and record later behaviour as a new event. At the same time, do not describe every later bounce as proof that the gap remains powerful. Repeated crossings, wide spreads, or a major change in the broader scenario may leave it with little practical use for the next decision.

A research protocol that can survive an unfavourable result

Freeze one instrument, one session definition, one entry rule, one cancellation rule, and one cost model. Save the gap at its formation time. Then advance the chart chronologically. Record whether the setup became eligible, whether the entry filled under your chosen fill assumptions, which exit occurred first, and what the result was after costs. Include days when nothing qualified. A record containing only attractive retracements cannot estimate opportunity frequency.

Separate measurement events from trade outcomes. “Reached midpoint within a session” is not the same statistic as “produced a profitable confirmed entry.” Price can touch the midpoint only after invalidating the setup, or after the trading window has ended. It can also reach both stop and target inside one bar. If lower-resolution data cannot establish their order, mark the trade ambiguous or apply a conservative predeclared rule.

To test added value, compare the same displacement-and-FVG entry protocol with and without the NDOG-location filter. Keep the other requirements fixed. Examine the change in trade count, average net R, largest drawdown, and the distribution of losses. If the filter reduces activity but does not improve the results you care about, that is useful evidence. There is no obligation to keep a reference simply because its explanation sounds sophisticated.

Finally, separate development data from later evaluation data. A rule chosen because it looked excellent on one month deserves testing on subsequent unseen sessions. Keep futures contract rolls, holidays, broker outages, and feed changes visible in the record. These can change the measurement process itself. Correcting a data problem is different from deleting an inconvenient losing trade.

11. Frequently asked questions about ICT NDOG

The NDOG becomes a practical part of a trading plan when its session definition, chart observations, and order decisions are kept precise. Build the surrounding framework with our guides to fair value gaps, order blocks, the London Kill Zone, market structure shifts, and the Judas Swing. Use those concepts to state a testable scenario, then let the actual sequence determine whether an entry qualifies.