ICT Central Bank Dealers Range (CBDR) Explained

A small evening range can produce a surprisingly complicated morning chart. Price leaves the range, passes a projected level, returns through it, and eventually moves in the direction you expected. Looking back, it is easy to draw a convincing explanation. Trading it in sequence requires something more useful: a fixed measurement, a clear directional hypothesis, and a way to recognise when the expected reversal has failed.

The ICT Central Bank Dealers Range, usually abbreviated CBDR, supplies the measurement part of that process. It uses a particular New York time window to frame an earlier range, then repeats that range’s height above and below its boundaries. The resulting levels help organise a study of later price delivery. They are not automatic buy or sell instructions.

This guide explains the original time window, the important difference between candle-body and wick measurements, and the practical consequences for traders in Ghana. Two hypothetical EUR/USD examples show exactly how a projection can support an entry scenario, how the risk is calculated, and how a plausible setup can still lose. The examples also show why selecting the right range matters more than adding more lines.

1. What is the ICT Central Bank Dealers Range?

ICT’s Month 8 lesson on the Central Bank Dealers Range defines the six-hour window and describes measuring either its full high-to-low span or its candle-body extremes. It then repeats the chosen height outside the range. In that teaching, a relatively compact range is preferred, with 20 to 30 pips presented as ideal and a range below 40 pips generally sought for the illustrated model.

Keep the three ingredients separate: the time interval identifies the observations, the measurement rule identifies the prices, and the directional context determines what you are investigating later. A narrow range does not tell you which currency to buy. An upward projection does not tell you that the day must finish higher. Without those distinctions, a useful planning reference turns into a collection of contradictory signals.

The phrase central bank dealers range can sound more official than the evidence supports. A chart rectangle does not identify central-bank transactions, dealer inventory, or the exact orders held by banks. In this article it means a named ICT chart construction. The institutional narrative concerns how traders interpret consolidation, liquidity around prior extremes, and later displacement. Observable price data and the explanation attached to it should remain distinguishable.

A range unit rather than a statistical calculation

ICT uses standard-deviation terminology for these repeated price-range units. Here, the calculation does not take a sample of returns, compute their mean, and estimate statistical dispersion. If the selected range is 24 pips high, one full projection step is another 24 pips. Two steps represent 48 pips beyond the relevant boundary. There is no implied normal distribution or confidence interval.

For the broader drawing-tool distinction, see our guide to ICT standard deviation projections. This article concentrates on choosing the CBDR observations and using them in a session plan. The source range is defined by time, so selecting a convenient swing after a reversal would create a different study.

2. Why CBDR matters in an ICT or SMC trading plan

A fixed earlier range gives your morning preparation something concrete to compare with later expansion. Instead of describing a rally as simply large, you can record how far it travelled relative to a measurement already available before the rally. That makes the observation reproducible. It also makes disagreements easier to resolve: two traders can compare their timestamps and prices rather than debate whether a move looked extended.

In an ICT reading, the important question is how that expansion relates to the expected direction of delivery. A trader investigating a bearish day may watch a rally above the earlier range as a possible stage for a high to form. A bullish hypothesis may involve studying a decline beneath the range. These are conditional scenarios. The movement away from the range is not itself confirmation that a reversal will follow.

ICT’s intraday profiles lesson connects a compact CBDR, a subsequent Asian range, and the timing of the move after midnight to particular London profiles. The sequence matters as much as the price distance. Do not label every later reversal a textbook profile merely because it occurred near a projected level. A scenario that requires an early rally is different from one in which the rally begins much later.

A useful plan therefore assigns each component one job. The higher-timeframe analysis supplies a tentative destination. The CBDR helps define where an opposing move could be evaluated. Existing liquidity references and price arrays narrow the area. Displacement and structure provide the entry evidence. Risk calculations decide whether the remaining opportunity is worth taking under your rules.

This framework does not establish that a single algorithm controls the forex market. The BIS discussion of FX execution algorithms describes fragmented execution and liquidity across venues. Your broker’s candles show a particular feed. They can support a consistent price-action study without revealing every institution’s decisions or proving why one exact reversal occurred.

3. CBDR times in Ghana and the midnight date change

Ghana stays on GMT throughout the year. New York changes between standard time, UTC minus five, and daylight time, UTC minus four. Consequently, the same New York CBDR window appears one hour earlier on a Ghanaian clock during New York daylight time. Use the date-specific New York offset rather than assuming that a fixed GMT setting will remain correct. The Ghana and New York time-zone records document that difference.

The familiar Ghana windows of 7 to 10 AM for London and noon to 3 PM for New York are the standard-time conversions for the New York-anchored conventions shown here. If your plan follows a London-local opening event, track that clock independently. London and New York have different clock-change dates, so transition weeks deserve particular attention.

Give a cross-midnight range one explicit date label. For example, a Tuesday CBDR in New York spans Tuesday evening into Wednesday morning in Ghana during standard time. Write “Tuesday NY CBDR, used for Wednesday London” in the journal. Calling it merely Wednesday’s range can accidentally combine data from the wrong evening with the right morning.

You do not need to trade while the measurement window is forming. A Ghana-based trader preparing before London can inspect the completed prior-evening range, then add the subsequent Asian observations. The practical requirement is that the whole source interval has ended before you treat its final boundaries as known. Any projection drawn earlier should be labelled provisional and excluded from a test of completed-range rules.

4. How to identify the range on a chart

1. Choose one feed and instrument. Record the broker, full symbol, and whether the displayed prices are bid, ask, or midpoint. Do not use another broker’s extremes to improve an unsuccessful projection after the event.
2. Set the New York time reference. Identify the local equivalent of 14:00 and 20:00 for the date being studied. Confirm how your platform timestamps candles and whether its server offset changes seasonally.
3. Declare the measurement timeframe. The examples use 15-minute candles. A different timeframe can change body extremes even when the full-window wick extremes stay the same.
4. Fix the boundary convention. This guide includes bars opening from 14:00 through 19:45 New York, with the last bar ending at 20:00. It excludes the new bar opening at 20:00. That is a stated implementation convention, not a claim that every indicator handles endpoints identically.
5. Calculate body and wick boundaries separately. Preserve both sets. Never combine a body high with a wick low unless you deliberately define and test that different measurement.
6. Record the completed result. Save the prices, pip width, source date, and projection levels before evaluating the next London session. Keep the original data available for later review.

On a complete 15-minute series, the stated six-hour interval contains 24 bars. That count is a data check, not permission to invent missing candles. A broker interruption, holiday, or data outage can leave observations absent. Inspect the timestamps and label the window incomplete when necessary. Compressing the visible chart so the missing period disappears does not repair the underlying record.

A candle that crosses the boundary is another problem. If a custom timeframe contains prices from both sides of 20:00, its high, low, or close may include information outside the intended interval. Use bars that align with the boundary or reconstruct the measurement from suitably detailed data. Write the rule once and keep it unchanged across winning and losing sessions.

5. Candle bodies versus wicks: two different measurements

For each candle, the top of the body is the greater of its open and close, and the bottom is the smaller. The body-range high is the greatest body top across the selected bars. The body-range low is the smallest body bottom. You are measuring the envelope of all those bodies, not adding up their individual lengths and not taking only the first open and final close.

The wick range is simpler: take the maximum candle high and minimum candle low over the same interval. ICT’s lesson on projecting daily highs and lows discusses examining both wick and body versions. Preserve their different results. A preference for body measurements does not mean wick prices were imaginary or that bodies directly reveal traded volume.

In the synthetic 24-bar example below, the body high is 1.08874 and the body low is 1.08634. The body width is 0.00240, or 24.0 pips. The wick high is 1.08893 and the wick low is 1.08613. Their width is 0.00280, or 28.0 pips. Both describe the same time window, but they produce different projected locations.

Why the timeframe changes the body result

Imagine a 15-minute candle whose close sits near an intrahour high, followed by a reversal before the hour ends. That close contributes to the 15-minute body envelope. On an hourly chart, the same price might appear only in the upper wick because the hourly close is lower. The hourly body envelope can therefore be narrower even though both charts contain the same underlying prices.

For complete data with aligned window boundaries, resampling should preserve the overall wick high and low. It need not preserve the body high and low. This is why the phrase “use the bodies” is incomplete as an operational instruction. Add the timeframe. If you change from five-minute to fifteen-minute bars, treat it as a new measurement variant and compare its results separately.

Do not choose bodies on days when they produce an attractive projection and switch to wicks on days when the wick version fits better. That gives the method information from the future. A practical protocol can use bodies as the primary map and keep wicks as a sensitivity check. If the maps disagree materially, record the disagreement instead of selecting whichever one later appears correct.

6. How to calculate CBDR projections and apply the range filter

Let H be the chosen high, L the chosen low, and D = H – L. The first upper projection is H + D. The second is H + 2D. The first lower projection is L – D, and the second is L – 2D. These formulas start at the relevant outer boundary. Starting at the midpoint would produce different prices and should not be silently substituted.

The body and wick upper-one levels differ by 1.09173 – 1.09114 = 0.00059, or 5.9 pips. At upper two they differ by 9.9 pips. That is a meaningful separation for an intraday trade with a small stop. Adding a wider source unit moves each successive projection farther away. A minor-looking measurement choice can therefore alter the proposed trade area substantially.

The traditional compact-range preference should be treated as a condition of the model being studied, not a timeless volatility law. A 39.9-pip range and a 40.1-pip range are very similar market observations even if a strict rule places them in different groups. For the example protocol here, both recorded widths must be below 40 pips, with 20 to 30 pips noted as the preferred category. Requiring both is this article’s conservative study choice.

If the body width is 27.4 pips but the wick width is 43.1 pips, that protocol rejects the session. Do not delete the wick extreme to make the case qualify. A different study could investigate body-only eligibility, but its results should remain separate. The same applies to a very narrow interval whose width is only a few times the current spread: mathematical projections exist, yet their practical usefulness may be limited.

A range filter does not identify a daily maximum. Price can move through one, two, or several repeated units without reversing. These distances describe a map, not a barrier. The worked short below uses an upper projection as an area to investigate a possible high; its profit target comes from a separate destination analysis. Calling the projected high itself a short trade’s target would confuse the setup location with the exit.

7. How to turn a projected area into a trade plan

Begin with the daily or four-hour context and state a tentative destination. In a bearish scenario, identify what lower liquidity or price array you expect price to investigate. Also identify what would contradict that expectation. A price projection cannot make a bearish thesis valid when the broader market is consistently delivering higher and the proposed downside objective is no longer relevant.

Next, compare the projected area with information that existed before the approach: an old high, an old low, a higher-timeframe FVG, or an order-block reference. Choose one observation area rather than treating every extension as another place to fade price. Confluence is useful when different observations answer different questions. It is less useful when several labels simply rename the same candle movement.

Define the trigger before the reaction

The teaching protocol used here waits for a raid of a premarked swing, displacement through an opposing short-term swing, and a completed three-candle FVG. It requires a candle close beyond the swing for consistency. That close requirement belongs to these examples; it is not a claim that every ICT entry model uses an identical rule. The third candle must finish before an order can rely on its final high or low.

The projection supplies location, while the sequence supplies evidence of a change in short-term delivery. If price arrives at the selected level and simply keeps moving, there is no entry under this protocol. If displacement occurs but no acceptable retracement follows, there is also no entry. Write both outcomes in the journal, because they are part of the method’s actual opportunity rate.

Separate the stop, target, and cancellation decisions

The stop should relate to the structural event that invalidates the setup. It need not sit precisely one pip outside a projected line. The target should relate to a pre-existing destination and be assessed from the actual entry price. A visually impressive projected area can still leave insufficient room for an entry after confirmation.

For the examples, an unfilled order expires at the end of the specified London study window or earlier if the raid extreme is broken again against the proposed position. There is one eligible attempt per sequence. A second reversal would require a newly documented setup rather than an automatic re-entry. These rules keep the model from accumulating unrecorded exceptions every time price fails to behave as expected.

8. Complete worked example: a bearish EUR/USD setup

This is a constructed sequence, not a historical result or a statement about current EUR/USD prices. Assume a broker bid chart, 15-minute CBDR measurement, five-minute execution observations, and winter Ghana time. The prior evening’s source window runs from 7 PM to 1 AM GMT. Its body boundaries are 1.08634 and 1.08874, with a 24.0-pip unit; its wick width is 28.0 pips. Both meet this example’s eligibility rule.

The primary map uses bodies, selected before London. Its upper-one projection is 1.08874 + 0.00240 = 1.09114. A pre-existing 15-minute bearish FVG spans 1.09097 to 1.09139, placing that projection inside a broader observation area. Prior highs near 1.09109 provide the buy-side reference to monitor. The wick upper-one level is higher at 1.09173; it remains documented as a different map.

The bearish hypothesis has an independently identified downside destination: a 15-minute bullish FVG from 1.08847 to 1.08881. The old CBDR body high, 1.08874, sits inside it. The proposed exit is 1.08877, within that destination and before reaching the old body boundary from above. That target is fixed before the trade. The setup does not require the entire evening range to be traversed.

During London, price raids the old highs and reaches 1.09127 in the candle closing at 8:00 AM GMT. This is 1.3 pips above the body upper-one projection and inside the premarked bearish FVG. It does not reach the wick upper-one projection. A trader who required that wick level would have a different eligibility result. Our body-based protocol now waits for its trigger rather than pretending both versions agreed.

A previously formed short-term swing low is 1.09053. The five-minute candle ending at 8:10 closes at 1.09039, below that low. The 8:05 candle’s low is 1.09085, and the 8:15 candle finishes with a high of 1.09061. Those outer candles establish a bearish FVG from 1.09061 to 1.09085 around the displacement. Its midpoint is (1.09061 + 1.09085) / 2 = 1.09073.

Only after 8:15 does the sell-limit order become eligible. Assume the bid retracement fills it at 1.09073 in the 8:20 candle. The stop is 1.09143, which is 1.6 pips above the raid high. The target remains 1.08877. In the illustrated path, the ask is assumed to reach the target during the 8:55 candle before reaching the stop. The diagram’s bid low alone would not prove that exit.

For a hypothetical $2,000 USD account with a self-imposed 0.5% risk budget, the cash budget is $10. Assume a standard EUR/USD lot is 100,000 euros, worth $10 per pip in a USD account, and a 0.01-lot sizing step. The maximum size from the estimated 7.8-pip loss is $10 / (7.8 x $10) = 0.1282 lots. Rounding down permits 0.12 lots.

At 0.12 lots, pip value is $1.20. Estimated loss is 7.8 x $1.20 = $9.36, and the hypothetical net reward is 18.8 x $1.20 = $22.56. The ratio remains $22.56 / $9.36 = 2.41R. These are teaching assumptions. Broker contract size, commission, actual spread, and execution can change the money result. The cost allowance is not a limit on possible slippage.

A stop-first path would produce the planned loss instead. A missed retracement would produce no trade. If the bid and ask record does not establish the order of entry and exits, the correct journal classification is uncertain, not automatically profitable. MetaTrader’s execution explanation distinguishes bid-side selling from ask-side buying, which matters when assessing a short exit from a bid chart.

9. A bullish projection that produces a losing trade

Consider a separate hypothetical winter session. The completed 15-minute body range runs from 1.09126 to 1.09346. Its unit is 1.09346 – 1.09126 = 0.00220, or 22.0 pips. The wick range is 1.09109 to 1.09367, a 25.8-pip span. Both qualify under the same filter. The body lower-one projection is 1.09126 – 0.00220 = 1.08906.

A bullish thesis is being investigated near a pre-existing 15-minute bullish FVG from 1.08879 to 1.08919. The projection falls inside it. In the five-minute bar closing at 7:35 AM GMT, price raids a premarked low at 1.08911 and reaches 1.08891. It then recovers, closes above the previously identified swing high at 1.08963 at 7:45, and completes a bullish FVG from 1.08937 to 1.08957 at 7:50.

Assume an ask-side buy fill at the FVG midpoint, 1.08947, during the 7:55 bar. The stop is 1.08877, beneath the raid low, and the target is 1.09122, just below the old body-range low at 1.09126. The projected area is the possible formation location for the low; the planned exit is the separate return objective above the entry.

Risk: 1.08947 – 1.08877 = 0.00070 = 7.0 pips. Reward: 1.09122 – 1.08947 = 0.00175 = 17.5 pips. Gross R:R: 7.0:17.5 = 1:2.50. With the same illustrative extra cost allowance of 0.8 pip, net reward is 16.7 pips and estimated loss is 7.8 pips. The adjusted ratio is 16.7 / 7.8 = 2.14R, rounded.

In this losing sequence, renewed selling takes the bid to the stop during the 8:05 candle. Assume the stop fills at 1.08877. The gross loss is 7.0 pips, or an estimated 7.8 pips with the stated allowance. At the same illustrative 0.12-lot size, the estimated loss is $9.36. The hypothetical target would have produced $20.04 after that allowance, but it was not reached while the trade remained open.

The correct response is to record a failed setup under the original rules. Moving the projection down to another multiple, choosing a different timeframe, or changing to the wick range after the loss would not repair that trade. If you want to test those variants, create separate studies before their outcomes. A method remains measurable only when a loss is allowed to count as a loss.

10. Common errors and a practical review process

Do not confuse related session measurements

The CBDR uses a multi-hour high-to-low or body-envelope measurement. The New Day Opening Gap uses two session-boundary prints in its futures construction. They are different objects. A broker rollover interruption inside a CBDR window deserves a data note; it does not turn the entire six-hour range into a gap.

Likewise, the subsequent Asian range and the New York midnight open add different time references. Keep their labels and construction rules separate. If an indicator combines several windows into one larger range or uses half of that combined height as its unit, it is not implementing the exact CBDR formulas shown in this guide.

Avoid precision that the data cannot support

A projection at 1.09114 is arithmetically precise. The prediction that a high must form at exactly that price is a separate and much stronger claim. In the short example, the raid extends above the line and the stop is farther away again. The trade is assessed through its structural sequence, not by demanding that the line identify the turning point to a fraction of a pip.

Do not enlarge a tolerance until every reversal counts as a hit. Define the surrounding observation area from pre-existing structure or write a numerical tolerance before replaying the session. Count approaches that continue through the area without reversing. A map that identifies many possible locations can appear accurate while providing little help in selecting executable trades.

Test the location filter separately from the entry model

Create one row per eligible source window. Record the New York date, Ghana timestamps, feed, timeframe, body and wick widths, selected measurement, fixed projection levels, and any excluded data. Then record the next-session hypothesis, premarked observation area, trigger completion time, order eligibility, fill assumptions, cancellation, and exit sequence. This separates the measurement from the trading decision.

Compare the same entry protocol with and without the CBDR-location requirement. Keep the displacement condition, FVG rule, risk sizing, target selection, session window, and cost model constant. Otherwise, a better result could come from another changed rule. Evaluate net R, frequency, losing streaks, and drawdown alongside the proportion of sessions that never offered an entry.

Maintain separate groups for body-based and wick-based maps. If you choose one after reviewing early data, assess that choice on later unseen sessions. Do not report the best-performing version from many trials as though it were the only version tested. A smaller but honestly defined result teaches more than a strong-looking statistic assembled by repeatedly changing the method.

Finally, preserve the time at which each fact became available. The final CBDR width is known only after its window closes. A three-candle FVG is known only after its third candle completes. A day’s high is only confirmed as the day’s high after the day ends. During trading, describe it as a candidate high or low. This language helps prevent hindsight from entering the rules unnoticed.

11. Frequently asked questions about ICT CBDR

A CBDR study is most useful when its time window, measurement method, and later entry requirements remain consistent. Develop the surrounding skills with our guides to fair value gaps, order blocks, the London Kill Zone, market structure shifts, and the Judas Swing. Use the completed range to frame an observation, then let the subsequent evidence and the calculated risk determine whether an entry qualifies.