Before anything else
ICT is a discretionary framework, not a proven edge. No independently verified backtest. Entry and target rules are the places where discretionary judgment matters most — and where Huddleston's own teaching is least precise. Every rule below is ICT-convention unless labeled otherwise.
Sources: ICT-HUB-RESEARCH-2026-06-29 (alansmith/guy-vault); innercircletrader.net; flowchart.html (deployed)
The Setup Chain ICT
The full ICT setup is a sequence, not a menu. Each step conditions the next:
- Liquidity sweep: price grabs equal high/low or stop cluster — BSL or SSL taken. This identifies where institutional orders were sitting.
- Displacement + MSS: a strong candle displaces through structure with an FVG. This is ICT's confirmation — institutional activity has been confirmed.
- Entry zone: price returns to the FVG or OB the displacement left behind. You enter on the retest, not on the sweep itself.
- Stop placement: beyond the swept liquidity wick + a buffer. Not at the structure edge — that gets wicked out.
- Target: the next liquidity pool / opposite end of the range.
Each step is a precondition for the next. A liquidity sweep without displacement = no entry. Displacement without a prior liquidity grab = weaker setup. The sweep identifies where the smart money is; the displacement confirms when it's moving. Enter the retest, not the move.
Entry — On the Retest ICT canon
Enter on the retest of the FVG or OB — not on the displacement itself, and not on the sweep
The entry is not "price broke out, I bought." It is: price swept liquidity, displaced, left an FVG and OB behind, and returned to that zone. That return is your entry signal.
What you're waiting for
Price returns down to the OB or FVG the displacement candle left behind. This retest is where liquidity is available for you to enter against.
The displacement told you institutions are pushing. The retest gives you a defined risk entry.
What you're NOT doing
Not entering on the sweep itself — that's where stops get taken. You want the liquidity grab to complete before you enter.
Not entering mid-displacement — by the time a strong candle closes with an FVG, the move may already be half over. Wait for the return.
Long entry — worked sequence (5M chart)
Label:
SSL swept
Retest entry
Target
Stop:
Below the SSL wick (beyond the sweep)
below SSL
Long entry: SSL grabbed → displacement up → price returns to entry zone → long. Stop beyond the sweep. Target = next liquidity pool above.
Ambiguity — "how deep must the retest go?" Huddleston teaches "trade the retest, not the gap" but doesn't define a depth rule. Some ICT traders enter when price merely touches the OB/FVG zone; others wait for a full candle close within it. Neither is canonically specified. Use a defined entry trigger (e.g., a rejection candle, a micro-structural break) and apply it consistently in your journal.
Best practice: Wait for price to trade into the zone and show a rejection response (e.g., a doji, pinbar, or displacement candle away from the zone) before entering. Entering on "price touched the OB" without confirmation is a common leak — you're guessing the retest will hold without evidence it already has.
Source: ICT-HUB-RESEARCH-2026-06-29; innercircletrader.net; flowchart.html (deployed)
Stop Placement — Beyond the Sweep ICT canon
Stop = beyond the swept liquidity wick + buffer — NOT at the structure edge
The most common ICT mistake is placing a stop "at the support" or "at the order block." ICT's own material lists "stopped out at the structure level" as its #3 listed error. The structure level is where institutions swept retail stops. If your stop is there, it gets taken.
The wrong stop
Stop placed at the OB, FVG edge, or structural high/low — the same level that just got swept by the liquidity grab. Institutions know those levels. Your stop is their liquidity.
The correct stop
Stop placed beyond the wick extreme of the liquidity sweep that triggered the move — past where the actual stops sat. This is beyond the sweep, not at the structure.
Add a small buffer beyond the wick (a few ticks on ES/NQ) to account for spread widening at the sweep.
Stop placement — long example
SSL sweep zone (sell-side liquidity below lows)
← SSL swept here (wicks through lows)
Stop at structure = WRONG → taken
Beyond the sweep wick
← Stop placed BELOW sweep wick
+ buffer = 2–4 ES ticks (example)
For short entries the logic mirrors: stop above the BSL wick + buffer. The principle is identical — go past the wick extreme, not to it.
Ambiguity — buffer size: ICT doesn't specify a canonical buffer. A reasonable starting point is 2–4 ticks on ES (e.g., $12.50–$25 risk per contract) and proportionally on NQ. Your buffer must account for spread widening at the exact moment of the sweep. Journal your stop distances and outcomes to calibrate.
Source: ICT-HUB-RESEARCH-2026-06-29 (canon); innercircletrader.net "Common Mistakes" (canon)
Target — Next Liquidity Pool ICT canon
Target = the next liquidity pool / opposite end of the dealing range — let R fall where it falls
ICT targets are not fixed R-multiples. They are structural: the next significant liquidity pool in the direction of the trade. R is the output, not the input. ICT's minimum is 1:3 — meaning your target distance should be at least 3× your stop distance.
Kill the "1.5–2R" rule. This is a non-ICT convention that has crept into ICT-adjacent material. ICT's own teaching is built on minimum 1:3 reward-to-risk. A 1.5–2R cap systematically clips the edge the framework is designed to capture. If your target is closer than 3× your stop, you are not trading ICT — you are trading something else.
What to target
- Equal high / equal low of the current session range
- Opposite end of the NY session dealing range
- Prior session's high or low
- Structural level identified on the 1H/4H
The target is the next pool, not a fixed dollar amount or R-multiple you pick before the trade.
What NOT to do
- Pre-set "I want 2R" and exit at 2R regardless of structure
- Move a target forward because "that's enough profit"
- Target an arbitrary price with no structural basis
Clipping a winner because it "looks like enough" is how you turn a 5R winner into a 1.5R winner — the math requires an average R well above 1 to compensate for realistic win rates in the 40–55% range.
| Target distance |
R-multiple achieved |
ICT verdict |
| 0.5× stop distance |
0.5R |
Non-ICT. Would require >67% win rate to be profitable. |
| 1× stop distance |
1.0R |
Non-ICT. Breakeven at 50% win rate. ICT calls for more. |
| 1.5× stop distance |
1.5R |
Non-ICT convention. Below ICT minimum. |
| 2× stop distance |
2.0R |
Non-ICT convention. Below ICT minimum. |
| 3× stop distance |
3.0R |
ICT minimum. With 45% win rate → positive expectancy. |
| 4–6× stop distance |
4–6R |
ICT target when structure supports it. Let winners run. |
* R = reward-to-risk. R-multiple = (target price − entry price) ÷ (entry price − stop price). Positive expectancy requires win rate × avg win > (1 − win rate) × avg loss.
Source: ICT-HUB-RESEARCH-2026-06-29; flowchart.html (deployed)
Trade Management
Partial Exits Best practice
Partial exits are not ICT canon — Huddleston does not prescribe them in his core methodology. They are a community best practice for managing large winners and reducing exposure as price moves in your favor.
Common partial-exit approach: take 1/3 to 1/2 off at a logical structural level before the final target, move the stop to breakeven (BE) after the first partial, and let the remainder run to the target. This locks in a gain even if the final target doesn't get hit.
ICT caveat: the structural level you use for a partial exit should be a real level (support/resistance, prior high/low, FVG), not an arbitrary R-multiple like "1R." Partial at structure, not at a number.
Moving to Breakeven (BE)
ICT canon
Move stop to BE only after price clears a real structural or liquidity level — a prior high/low, FVG boundary, or the opposite end of a confirmed range. Not at an arbitrary 1R.
ICT rule: if you move your stop to BE and price comes back to take it, the trade was wrong — but you lost nothing but time. Never move a stop against your position (i.e., deeper into the trade) after moving it to BE.
Ambiguity — "which level clears BE?" ICT specifies a principle (clear a structural/liquidity level) but not which specific level counts. Common approaches in the ICT community: move to BE after price clears the high/low of the candle that triggered entry, or after price takes out a micro-structure level within the trade direction. Label your BE rule in your journal and track whether it improves your results.
Best practice (not ICT canon): Some traders move a fraction of their position to BE at 1R (locking in a win on that portion) and leave the rest to run. This is not in Huddleston's canon — it's a risk-management layer adopted from mainstream position management. Use it if it helps you hold winners, but track whether it's hurting your R-multiple average.
Exiting Before the Target
Exiting before your target is the most financially damaging habit in trading. An exited-at-2R winner when the target was 5R doesn't just cost 3R — it costs the expectancy that the 5R scenario was delivering. If you set a target before the trade and then exit early because you're "happy with the profit," you are systematically clipping your winners and violating the method's own math.
The only valid reasons to exit before target:
- Price has reached a structural decision point (support/resistance flips, major level) — not a mood
- Your daily risk cap is reached — stop trading, not the current trade
- A news event is approaching that could cause a liquidity event (exit before, not after)
- The setup has been invalidated (price structure has broken against you, not just pulled back)
Quick Reference
| Element |
Rule |
Source |
| Entry |
On the retest of the FVG/OB left by the displacement — not on the sweep, not mid-displacement |
Canon |
| Stop |
Beyond the swept liquidity wick + buffer — not at the structure edge (that gets wicked out) |
Canon |
| Target |
Next liquidity pool / opposite end of range — ICT minimum is 1:3R |
Canon |
| R-multiple |
Output, not input — let it fall where the structural target lands; minimum 3R, not 1.5–2R |
Canon |
| Move to BE |
After price clears a real structural or liquidity level — not at 1R |
Canon |
| Partial exits |
Not ICT canon — community best practice. Take at structural levels, not arbitrary R values. |
Best practice (labeled) |
| Exit before target |
Only at real structural decision points, daily risk cap, or approaching news — not because you're "up enough" |
Canon + best practice |
Source: ICT-HUB-RESEARCH-2026-06-29; innercircletrader.net; flowchart.html (deployed)