This article does not teach any single model in depth — each one has its own dedicated guide below. This is the map: the foundational AMD concept every other model assumes you understand, plus a full directory of every model and where to go to learn it properly. This follows the same function as the
ICT PD Array hub for the previous cluster.
The Foundation: What is the AMD Model?
The AMD model — short for Accumulation, Manipulation, Distribution — is the three-phase price cycle that underlies virtually all ICT trading models. Understanding it is the prerequisite for everything else in this cluster.
Accumulation — price consolidates in a range, building the energy and liquidity needed for the next move.
Manipulation — a false, deceptive move in the opposite direction of the eventual real move, sweeping resting liquidity.
Distribution — the real directional move, running toward the genuine liquidity target that was always the actual destination.
The critical insight from this diagram is the proportions. Accumulation is typically the longest phase by time but the smallest by range. Manipulation is sharp and brief — a spike or sweep that traps traders positioned in the wrong direction. Distribution is the payoff: the largest, most sustained move of the three phases, running cleanly once both sides of the liquidity equation have been addressed.
How AMD Connects to Every ICT Model
Every model in this cluster is AMD expressed with a specific refinement. The
Silver Bullet identifies an AMD cycle compressed into a specific kill zone window.
Turtle Soup specifically targets the Manipulation phase — the false breakout — as the trigger entry. The
MMXM model describes the full AMD sequence through a market-maker lens, labelling each phase by institutional intent. The
Judas Swing is the Manipulation phase of AMD named specifically for the early session false move. Even the
Unicorn Model (
Breaker Block plus FVG) is essentially looking for a PD array entry at the transition between the Manipulation and Distribution phases.
Understanding this is more valuable than memorising each model’s rules in isolation: once you see how every model maps back to AMD, you can adapt to variations in real market conditions without being rigidly dependent on one set of rules.
The Models & Sessions Directory
Every model below has a dedicated, full-depth guide. Descriptions here are intentionally brief — pick the model most relevant to your trading style and click through for the complete breakdown.
ICT Kill Zones
The time-defined windows — London Open, New York Open, and the Asian range — where institutional participation is highest and ICT setups are most reliable.
MMXM Model
Market Maker Buy Model and Market Maker Sell Model — the full AMD cycle described through the lens of how market makers accumulate, manipulate, and distribute positions.
SMT Divergence
Smart Money Technique Divergence — using correlated instrument pairs (EURUSD/GBPUSD, NQ/ES) to identify when one market fails to confirm a new high or low made by the other, signalling institutional repositioning.
ICT Silver Bullet
A specific three-step entry model designed around a precise kill zone window, identifying the AMD manipulation phase and entering on the FVG left by the displacement move that follows it.
ICT Turtle Soup
A specific strategy targeting the Manipulation phase of AMD — the false breakout beyond a prior range high or low — and entering on the reversal that follows the
liquidity sweep.
ICT Unicorn Model
A high-confluence entry model combining a Breaker Block with a
Fair Value Gap on the retest — two PD array tools confirming the same zone, applied at the Manipulation-to-Distribution transition.
OTE Trading
Optimal Trade Entry — the Fibonacci-based retracement framework for finding the highest-probability entry level within a confirmed displacement move, particularly within the Distribution phase of AMD.
Judas Swing
The early-session false move — typically the Asian session or the London Open — that sweeps liquidity in the opposite direction of the session’s genuine bias before the real move develops. This is AMD’s Manipulation phase named for the early-morning context in which it most commonly appears.
The right model depends on your schedule, risk tolerance, and chart-reading strengths — not on which one has the highest theoretical win rate in a backtest.
If you trade fixed, predictable hours → Kill Zones or Silver Bullet. These models require you to be at your screen during specific time windows and reward discipline around session timing.
If you prefer seeing the full institutional cycle before entering → MMXM Model. This model waits for the full AMD sequence to become visible before committing to the Distribution entry, which is more conservative but slower.
If you prefer trading false breakouts and reversals → Turtle Soup or Judas Swing. These models specifically target the Manipulation phase as the trigger, requiring you to be comfortable fading moves that look like breakouts to most traders.
If you trade correlated pairs or markets → SMT Divergence. This model adds an additional confirmation layer by requiring the correlation between two instruments to break down at a swing point, which can significantly increase conviction on entries.
Whatever you choose, study one model completely — including backtesting it across at least three months of historical data — before adding a second. See our
ICT Trading Strategy guide for how
daily bias and the five-step process apply regardless of which specific model you use.
Frequently Asked Questions
How to Choose the Right ICT Model for Your Style
ICT offers multiple trading models, each designed for different time commitments, risk tolerances, and market preferences. Choosing the wrong model for your actual trading circumstances is one of the most common reasons students plateau — they are studying a model that does not fit their available time or temperament.
If you have a full-time job and can only check charts briefly: the London Close Trade or Daily Bias model suits you best. These require 30-60 minutes of analysis in the morning, a brief check at a specific time window, and then minimal monitoring. The setup is defined before the session, and the trade either fills or it does not.
If you have 2-3 focused hours per day during a specific session: the Silver Bullet model is ideal. Three specific 1-hour windows (3-4 AM, 10-11 AM, 2-3 PM New York) provide defined opportunities without requiring all-day monitoring. Within those windows, the mechanical approach of marking the opening price and waiting for a FVG limits the time on-screen to the kill zone itself.
If you can trade full-time across both London and New York sessions: the MMXM or AMD model gives you the full daily delivery cycle to work with. These models are more complex, require more chart time, and demand a deeper understanding of session interplay — but they also provide more opportunities per day when applied correctly.
Combining ICT Models: What Works and What Doesn't
ICT models are not mutually exclusive — experienced traders combine them, using one for direction and another for entry timing. A common combination: use the AMD model for directional context (what phase is the day in — accumulation, manipulation, or distribution?), and use the Silver Bullet entry for the actual trade execution during the manipulation-to-distribution transition.
The effective combinations share a common logic: a higher-level model provides the framework (which direction, which phase), and a lower-level model provides the execution (when exactly, from which zone). AMD → Silver Bullet, MMXM → One Minute Mastery, and Power of 3 → Judas Swing are all examples of this framework-plus-execution combination.
What does not work: combining two framework-level models without resolving the conflict between them. Using MMXM AND Power of 3 simultaneously without a clear rule for which takes precedence creates decision paralysis when they point in different directions. Pick one framework model and one execution model. Add complexity only after you have thoroughly mastered both of your primary choices.
The Mastery Timeline: How Long Does Each Model Take?
Honest expectations for model mastery matter. Trading a model correctly in a live account requires not just understanding the rules but developing the pattern recognition to identify setups correctly in real time, under pressure, with real money at stake. This takes longer than most students expect.
A realistic timeline: 3-6 months of study and backtesting to understand the rules and identify historical examples. Another 3-6 months of forward testing (paper trading or very small live size) to develop real-time recognition. 6-12 months of consistent live trading to build confidence and refine the application. Total: 12-24 months from first learning a model to trading it with full conviction and consistent results.
This timeline assumes daily practice — reviewing charts, backtesting setups, and journaling every trade. Students who study intermittently or skip the journaling phase extend this timeline significantly. The mastery timeline is not discouraging — it is clarifying. It tells you that trading results in the first 6 months are training data, not performance data. Give yourself the full development period before judging whether a model works for you.
Watch: ICT Trading Models: AMD, MMXM, Silver Bullet, Turtle Soup & More