INSTITUTIONAL ORDER FLOW

Large orders can create execution challenges, including market impact and slippage, so some institutions use execution strategies that split orders across time or venues. The exact methods vary by market, instrument, liquidity and mandate. A chart by itself does not reveal the size of an order, who placed it, or whether a consolidation reflects accumulation. Treat AMD as a framework some ICT traders use to interpret ranges and directional moves—not as a verified account of every institutional transaction.

In ICT terminology, accumulation describes a range-bound phase in a proposed Accumulation-Manipulation-Distribution (AMD) sequence. A chart can show consolidation, repeated tests of levels, and changing range or volatility; it cannot establish that a participant is secretly building a directional position. These features may be recorded as candidate conditions in a model. Potential observations to log include range width, duration, session, volume where reliable, number of level tests, subsequent breakout direction, and the maximum adverse move after a breakout. Compare qualifying cases with non-qualifying ranges and include failed breakouts. A sweep may follow a consolidation, but it is not required and does not reveal the motive for the move.

A false breakout occurs when price moves beyond a visible level but fails to sustain the break under a trader's predefined rule. ICT material may interpret this pattern as a stop hunt or manipulation, but price action alone cannot show whether a bank caused the move, which orders were triggered, or why the price reversed. A testable example is to mark a prior high or low, define the required penetration and close-back condition, specify an entry trigger and invalidation level, and record every qualifying instance—including cases that continue in the breakout direction. Separate the observed pattern from the explanation assigned to it. Do not assume a reversal simply because price crossed a level.

Traders sometimes use displacement, gaps and wicks to describe candle patterns. A large-bodied candle with relatively small wicks is an observable feature; a Fair Value Gap is defined by the chosen candle-range rule; a wick beyond a prior high or low may meet a trader's definition of a sweep. None of these features identifies an institution or establishes intent. For a repeatable test, define candle size relative to recent volatility, specify the exact FVG and sweep rules, then record follow-through, retracement, invalidation, spread and execution costs. Compare performance with a baseline that uses the same instrument and periods. Avoid labeling all overlapping candles as retail noise or assuming that every displacement leads to continuation.

Some ICT traders describe a sequence such as a move beyond a prior low, a sharp rebound, a structure break and a retracement as a possible bullish scenario; the reverse sequence may be studied as bearish. These are chart interpretations, not a record of institutional intent or a confirmed accumulation/distribution cycle. Write down the levels and trigger conditions before the outcome is visible. Specify what would invalidate the scenario, and include examples where the sequence fails or does not develop. If the rule cannot be identified consistently in real time, it is not yet precise enough to evaluate.

Before entering, check whether price matches the conditions in your written setup. A pre-trade checklist might include a defined displacement condition, a move through a marked high or low, a specified lower-timeframe structure break, and a retracement into a reference zone. These checks document price behavior; they do not confirm institutional order flow. Define what each condition means, including candle-close or wick rules and the timeframe used. Test whether each condition adds value to the baseline setup, and record all valid signals whether they win, lose, or never reach the target. If one condition is absent, follow the prewritten rule rather than making an after-the-fact judgment about institutional intent.

In the AMD framework, distribution describes a directional move away from a range toward a potential objective. A trader can record directional candles, retracements, range expansion, and interaction with prior highs or lows. Those observations do not establish that institutions are executing a coordinated delivery plan. Define a possible end condition in advance—for example, price reaching a marked level, a structure change, a time limit, or a protective stop. Price may pass a target, reverse before reaching it, or continue beyond it. Use the same exit rule in backtests and forward tests, including costs and adverse outcomes.