ICT IOFED Explained: Shallow Fair Value Gap Entries

You mark a fair value gap, place an entry at its midpoint, and watch price turn just inside the nearest edge. The market then reaches the objective you had identified, while your order remains unfilled. It is tempting to conclude that the midpoint was wrong or that you should enter every gap immediately. Neither conclusion follows from one missed trade.

The ICT Institutional Order Flow Entry Drill, usually shortened to IOFED, gives you a way to study partial entries into a fair value gap. Its practical value lies in understanding the difference between an attractive price and a price the market actually offers. That distinction matters when institutional order flow analysis points toward a nearby liquidity objective and the retracement remains shallow.

This guide develops a specific, testable implementation: a limit order just inside the near edge of a completed FVG, with context and structural confirmation established beforehand. You will learn the geometry, compare it with a midpoint order, and work through a winning long and a losing short. All numerical examples are constructed for education, not historical performance records or current trade signals.

1. What is the ICT Institutional Order Flow Entry Drill?

In ICT’s 2023 lesson on immediate rebalance and institutional order flow, around 25 minutes into the presentation, he describes an institutional order flow entry drill as a partial entry into a gap that does not completely fill before the next move. That is the central distinction to retain: partial engagement of a previously identified imbalance, rather than a requirement to revisit its entire range.

The name can make the technique sound like a direct view of institutional orders. It is not. On an ordinary forex chart, you observe prices and candle relationships, then apply an interpretive framework. The presence of a shallow return does not identify who bought, who sold, or how much inventory any participant holds.

For practical study, separate three statements. First, the FVG has valid geometry. Second, the market partially revisits it. Third, a trade placed under predefined rules makes or loses money. A valid answer to one statement does not settle the others. This separation prevents the label IOFED from becoming a substitute for an entry plan.

A partial return can have different depths

A one-pip penetration into a wide gap and a return through more than half of a narrower gap are both incomplete revisits. Calling every partial revisit identical conceals meaningful differences. Record the actual penetration and whether the midpoint was reached. In this article, the worked entries sit close to the first boundary encountered by a returning market.

That narrower choice makes the examples reproducible. It also makes the trade-off visible: a near-edge buy is higher than a midpoint buy within the same bullish gap. With the same protective stop and target, it carries more price risk and less price reward. It may fill on a shallow pullback that leaves the midpoint untouched, but that alone does not make it superior.

2. Identify the near edge, far edge, and midpoint correctly

Begin with a completed three-candle formation. In a bullish FVG, the third candle’s low is above the first candle’s high. The interval between those two prices is the gap. When price returns from above, the third candle’s low is the near edge and the first candle’s high is the far edge. The middle candle provides the intervening directional movement.

In a bearish FVG, the third candle’s high is below the first candle’s low. A return from below meets the third candle’s high first, making it the near edge. The first candle’s low is the far edge. This reverses the vertical arrangement, so a bearish near edge is the lower boundary, whereas a bullish near edge is the upper boundary.

Consider a bullish gap with first candle high 1.08674 and third candle low 1.08722. Its width is (1.08722 − 1.08674) ÷ 0.00010 = 4.8 pips. Its midpoint is (1.08674 + 1.08722) ÷ 2 = 1.08698. A later low of 1.08712 penetrates the gap by (1.08722 − 1.08712) ÷ 0.00010 = 1.0 pip, or 1.0 ÷ 4.8 × 100 = 20.8%, rounded.

The first and third candles do not overlap across this interval, but the middle candle travels through it. An FVG therefore does not prove that the interval contained no trading. It is also different from a previous-close-to-new-open jump, which uses another measurement. Preserve the original boundaries rather than redrawing them to match a later turning point.

Wait until the third candle has finished

During a live third candle, its low can fall or its high can rise. A gap that appears open halfway through the bar may disappear before the close. For the model developed here, no order is authorised until the third candle closes and the FVG is confirmed. You cannot use that candle’s earlier wick as the fill of an order placed afterward.

The site’s FVG guide covers the underlying three-candle formation. The consequent encroachment guide explains the midpoint reference. IOFED adds a question about the depth offered on a later return, rather than changing either calculation.

3. Why shallow entries matter in institutional order flow analysis

ICT analysis commonly organises a trade around a potential destination, such as an earlier high or low associated with inferred stop orders, and price arrays that could support delivery toward it. In that framework, the FVG is an intermediate reference. The objective gives the retracement a purpose; the rectangle by itself does not tell you which direction deserves attention.

Suppose a bullish displacement has broken an internal high after a lower-timeframe correction into a previously marked bullish area. A shallow return into the resulting gap is consistent with continued upward delivery. It remains only one possible interpretation. Price can also pause briefly before returning more deeply or reversing outright.

The entry decision exchanges price improvement for participation

A deeper limit buys at a lower price or sells at a higher price. That is favourable if it fills under otherwise unchanged conditions. The cost of demanding that improvement is that some moves leave without offering it. A near-edge order accepts a less favourable price to participate in more shallow returns, assuming the market actually reaches the executable limit.

This is not a reason to celebrate every extra fill. Some additional fills will be losses that a deeper order avoided. Other shallow entries may spend longer in drawdown because they were filled earlier in a retracement that continued. The correct question is whether the complete set of outcomes improves under a consistent risk budget.

Do not infer a universal algorithm from a clean reaction

The BIS report on FX execution algorithms describes a fragmented market in which execution practices can reduce the visibility of trading activity. That supports a modest practical conclusion: one broker’s candles do not show the full market’s institutional order book. A precise-looking response is an observation to test, not proof that all participants follow one hidden instruction.

Your chart work becomes more useful when the language remains accountable. Write that price entered 20.8% of the gap and moved toward a premarked high. Do not replace those observations with a claim that a particular bank defended the level. The first description can be reproduced from the data; the second requires evidence the chart does not provide.

4. How to identify an IOFED candidate before the return

The difficult part is selecting candidates before you know how deeply they will retrace. If you inspect only charts where the midpoint was missed and the target was reached, a near-edge method will appear excellent by construction. Start with an eligibility process that can include missed orders, full fills, and losing trades.

1. Mark context and a destination. Identify the higher timeframe area being used and the potential liquidity objective before selecting the entry gap. Record why the direction is plausible.
2. Observe a directional change or continuation signal. For this study model, require a completed displacement close through a previously marked internal swing. A wick through the swing alone does not pass that rule.
3. Confirm the FVG. Wait for the third candle’s close, then record the first and third candle boundaries from one consistent feed.
4. Choose the order rule. Use a fixed offset or another written method to place the limit inside the near edge. Do not decide the offset from a turning point that has already happened.
5. Calculate stop distance, reward, and costs. Reject the setup if the resulting ratio does not meet the model’s minimum.
6. Set expiry and cancellation conditions. A limit order needs a defined lifetime. Save the no-trade result if price never offers the planned entry.

The market structure shift guide explains the structural component. Here it occurs during the departure that creates the FVG. Waiting for another structure shift after the return would be a different, more delayed entry model. Both can be studied, but their fill records should not be blended.

Use one execution timeframe initially. A five-minute gap and a nested one-minute gap can have different boundaries and different confirmation times. Choosing whichever one retrospectively fits the best entry creates an inconsistent sample. If you later introduce a nested-gap rule, state exactly when and how the smaller formation becomes eligible.

5. A complete trade plan for the near-edge version

The following settings belong to this article’s educational model. They are not presented as mandatory ICT rules or proven optimal parameters. Use five-minute candles, a first return to a newly confirmed FVG, and a limit order 0.3 pip inside its near edge. The offset is below the upper boundary for a buy and above the lower boundary for a sell.

In the examples, the stop sits beyond the displacement candle’s relevant extreme, with a 1.3-pip buffer. For a long, that means below its low; for a short, above its high. This is a local invalidation choice. It can stop the trade before the wider session thesis fails, which is acceptable if that is the risk definition you chose in advance.

The target is placed slightly before a premarked liquidity reference. Require at least 2R after the stated cost allowance. Expire an unfilled order after two additional five-minute candles, or at the session cutoff if earlier. Cancel it if the target is reached before entry. Use one attempt for the selected setup, one protective stop, and one full-position target.

Keep execution filters separate from chart geometry

For this study, the observed spread must be at or below 0.8 pip when the order is placed and remain within that condition until it fills. Exclude new orders during the ten minutes before and fifteen minutes after a scheduled high-impact announcement affecting either currency. Cancel pending orders as that exclusion period begins. These are research settings, not assurances of orderly execution outside those times.

If a trade remains open when the session ends, exit at the available executable price and record that result. Do not extend the holding period until the target is eventually reached. If you plan to hold through later announcements, that needs a separate management rule and sample; it is not part of these examples.

A gap too narrow for the chosen offset or too costly relative to its stop distance is not eligible. Neither is a trade with an attractive rectangle but no credible room to its objective. Skipping it is a completed decision. There is no obligation to convert every recognised IOFED pattern into an order.

6. Complete bullish EUR/USD example with specific prices

Imagine a winter London session observed from Ghana, with all times in GMT and all candle times below identifying completed five-minute closes. Before 8:00 AM, the preparation notes identify a bullish higher timeframe area and a potential buy-side liquidity objective at 1.08996. At 8:00 AM, a move below an earlier low of 1.08608 reaches 1.08589, then returns above the old low.

An internal high at 1.08662 is marked before the departure. The candle closing at 8:10 AM has a high of 1.08674 and closes at 1.08657. It has wicked above the internal high but has not closed above it. The next candle, closing at 8:15 AM, expands upward and closes at 1.08748, with a low of 1.08644.

The third candle closes at 8:20 AM with a low of 1.08722. Only now is the bullish FVG from 1.08674 to 1.08722 confirmed. Its midpoint is 1.08698. Assume the session and execution filters pass and the target remains untouched.

Place the order after the information exists

The buy limit is 1.08722 − 0.00003 = 1.08719, which is 0.3 pip inside the near edge. The protective stop is 1.08644 − 0.00013 = 1.08631, 1.3 pips below the displacement candle’s low. The target is 1.08983, 1.3 pips before the premarked objective at 1.08996.

The next candle, closing at 8:25 AM, reaches a low of 1.08712. It penetrates the FVG by one pip and passes the planned entry in this simplified reference-price sequence. The midpoint at 1.08698 remains untouched. The later continuation reaches the target during the candle closing at 8:50 AM.

The 0.8-pip allowance represents an illustrative combined estimate for spread, commissions converted to pips, and execution friction. It is not a current broker quote. Estimated loss is 9.6 pips and estimated target profit is 25.6 pips. Conditions can make the realised result worse, particularly if an exit slips.

Position size follows the risk budget

For a hypothetical US dollar account of US$5,000, a selected risk budget of 0.3% is US$15. With a standard EUR/USD contract of 100,000 euros, a pip is US$10 per standard lot in that account currency. The calculated size is US$15 ÷ (9.6 × US$10) = 0.15625 standard lot.

If the broker supports 0.01-lot increments, round down to 0.15 lot. Estimated loss becomes 9.6 × US$1.50 = US$14.40, and estimated target profit becomes 25.6 × US$1.50 = US$38.40. The ratio is unchanged by that proportional sizing. An account denominated in another currency needs the relevant pip-value conversion.

The earlier sweep low at 1.08589 is below the chosen stop. This plan deliberately invalidates the entry at a nearer reference associated with the displacement, rather than holding until the entire session extreme breaks. A trader who requires a stop beyond the sweep is evaluating a different risk structure and must recalculate the trade before placing it.

7. Compare the IOFED entry with a midpoint order and a late chase

Use the same bullish setup to compare alternatives, holding the stop at 1.08631 and target at 1.08983. These are separate hypothetical choices, not three orders to place together. The purpose is to isolate the effect of changing the entry price before adding other differences in management.

A midpoint buy would sit at 1.08698. Its price risk is (1.08698 − 1.08631) ÷ 0.00010 = 6.7 pips, and its price reward is (1.08983 − 1.08698) ÷ 0.00010 = 28.5 pips. The gross ratio is 28.5 ÷ 6.7 = 4.25R. With the same 0.8-pip cost allowance, it becomes 27.7 ÷ 7.5 = 3.69R.

That planned ratio is higher than the IOFED version’s 2.67R after costs. However, the illustrated pullback stops at 1.08712, above the midpoint. The midpoint order therefore remains unfilled during its permitted lifetime. Its result is no trade and zero realised P&L, not a 3.69R winner. The attractive ratio describes a conditional outcome that never became available.

Now consider an impulsive market buy at 1.08791 after the shallow return has already resolved upward. Keeping the original stop and target gives (1.08791 − 1.08631) ÷ 0.00010 = 16.0 pips of price risk and (1.08983 − 1.08791) ÷ 0.00010 = 19.2 pips of reward. Gross reward-to-risk falls to 19.2 ÷ 16.0 = 1.20R. After costs, it is 18.4 ÷ 16.8 = 1.10R.

The late entry fails the model’s 2R minimum and is rejected. Moving the target farther away solely to restore the desired ratio would change the thesis. Tightening the stop without a valid new reference would change the risk definition. A missed order does not justify either adjustment.

This path favours the near-edge entry, but one path cannot select the better method. On another day, the near edge could fill and lose while the midpoint never fills. On a deeper successful return, both could fill, with the midpoint offering the better ratio. Your comparison needs all those cases, not just the one that is easiest to illustrate.

8. Bearish worked example: a shallow fill can still lose

Consider a separate hypothetical winter New York session in EUR/USD. An earlier move has swept a prior high of 1.09454 to 1.09467. The working context is bearish, and a sell-side liquidity reference at 1.09107 is marked as a possible destination. These conditions frame the example but do not guarantee its outcome.

The first candle in the selected formation, closing at 12:25 PM Ghana GMT, has a low of 1.09386. The middle candle closes at 12:30 PM at 1.09302, below the previously marked internal low of 1.09372. Its high is 1.09396. The third candle closes at 12:35 PM with a high of 1.09334.

The confirmed bearish FVG spans 1.09334 to 1.09386. Its width is (1.09386 − 1.09334) ÷ 0.00010 = 5.2 pips and its midpoint is 1.09360. A return from below encounters 1.09334 first. After the third candle closes and the filters pass, place the hypothetical sell limit at 1.09337, 0.3 pip inside that near edge.

The stop is 1.09396 + 0.00013 = 1.09409, and the target is 1.09121, 1.4 pips before the lower liquidity reference. Price risk is (1.09409 − 1.09337) ÷ 0.00010 = 7.2 pips. Price reward is (1.09337 − 1.09121) ÷ 0.00010 = 21.6 pips. The planned gross ratio is 21.6 ÷ 7.2 = 3R, or 1:3.

With the same 0.8-pip round-trip allowance, estimated loss is 7.2 + 0.8 = 8.0 pips, while estimated target profit is 21.6 − 0.8 = 20.8 pips. The adjusted ratio is 20.8 ÷ 8.0 = 2.60R. It passes the minimum planned-ratio rule.

The candle closing at 12:40 PM reaches a high of 1.09349, filling the sell limit in the simplified reference-price sequence. The return initially remains below the midpoint. But the next candle rises to 1.09417 and closes at 1.09412. It has passed the stop at 1.09409, so the example records an estimated loss of 8.0 pips including the allowance.

The first shallow touch was real within the constructed path, and the planned ratio was acceptable. Neither fact prevented a deeper reversal. Do not erase this candidate from the sample because it later ceased to look like a successful IOFED. At the time the order was placed, it satisfied the stated rules.

9. Ghana session timing and practical order execution

Ghana stays on GMT throughout the year. New York uses UTC−5 during standard time and UTC−4 during daylight time. If your session windows are anchored to the New York clock, convert them for the date being studied. Do not let a chart platform’s server timezone silently determine when your model becomes active.

The familiar Ghana windows of 7 to 10 AM for London and noon to 3 PM for New York are winter conversions of those anchors. The clock rules are documented in the Ghana time zone reference and New York clock-change reference. A rule based on London’s local opening is a different convention, especially around weeks when London and New York change clocks on different dates.

A chart touch and an executable fill are different

A forex buy opens at the ask and closes at the bid; a sell opens at the bid and closes at the ask. If your chart displays bids, its low touching a buy limit does not by itself show that the ask reached the order. This matters when the planned entry sits only a fraction of a pip inside a gap.

The examples use reference prices and subtract a separate cost allowance. In a test based on executable bid-and-ask prices, the spread is already reflected in the entry and exit. Do not subtract it a second time. Add only costs not already included, and preserve commission and execution records so that the calculation is auditable.

The CFTC’s forex customer advisory discusses leverage and dealer-dependent trading conditions. Its US registration framework does not establish authorisation in Ghana. For this entry method, the immediate lesson is to understand your account’s contract size, quote basis, stop behaviour, and execution terms rather than assuming that a clean candle guarantees a clean fill.

10. How to test IOFED without selecting only the winners

Build the candidate list at the moment each FVG becomes eligible. Store a chart showing only information available at that moment, then replay the subsequent prices. For each candidate, record the context, departure close, three candle timestamps, boundaries, near-edge offset, stop, target, expiry, and session convention.

Evaluate a near-edge model and a midpoint model on the same candidates. Use the same risk budget per filled trade, because the different stop distances imply different position sizes. Record no-fills as zero realised P&L at the opportunity level. Do not quietly remove them from one model’s denominator while retaining them in the other’s.

Measure opportunity outcomes as well as trade outcomes

Report fill rate, the proportion of filled trades that win, average gain, average loss, and average net outcome per eligible candidate. The average across filled trades answers a different question from the average across all opportunities. A method can have attractive results when filled but spend much of the sample unfilled.

Also record maximum adverse movement after entry, maximum favourable movement before exit, time to fill, and whether price later reached the midpoint. Those observations can explain where the two models differ. They do not justify rewriting an old trade’s entry after seeing the result. Changes should be tested on a later, separate set of candidates.

Resolve ambiguous candle paths honestly

A five-minute candle can contain both the stop and target without revealing which came first. The same problem appears when entry and invalidation lie within one bar. Use sufficiently detailed timestamped data where available. If the order cannot be established, flag the outcome as ambiguous or apply a predefined conservative convention consistently.

Do not assume a winning sequence because the candle eventually closes in the anticipated direction. A price path can fill an entry, hit its stop, and then rally to the original target within one bar. A screenshot of the close cannot rescue the stopped trade. This is especially relevant to narrow stops and volatile announcement periods.

Keep the question small enough to answer

Begin with one pair, one session convention, one execution timeframe, and one offset rule. Store the research period and feed. If you vary the gap age, entry depth, stop reference, and time window together, you will not know which change affected the result. A detailed journal is more useful than a large collection of labelled charts without order timestamps.

No minimum number of attractive screenshots proves an edge. Look for stability across later observations and realistic cost assumptions, then use demo execution to identify differences between replay fills and actual order handling. If a small increase in costs removes the apparent advantage, that sensitivity belongs in the conclusion.

11. Frequently asked questions

Build the surrounding framework with the guides to fair value gaps, ICT order blocks, and market structure shifts. Use the London Kill Zone guide and Judas Swing explanation to organise session context. Then compare IOFED and deeper entries with fixed rules, realistic fills, and a record that gives missed orders and losing trades their proper place.