Smart Money Concepts (SMC): the complete guide

Smart Money Concepts, usually shortened to SMC, is a modern retail trading framework that reads price charts as the footprint of large institutional orders. Instead of moving averages and oscillators, SMC traders talk about fair value gaps, order blocks, breaks of structure, liquidity sweeps and kill zones: a whole vocabulary for describing where big participants supposedly bought, sold, and hunted the stops of everyone else. This page explains where the framework came from, how each concept works mechanically, how a rules-based engine can turn those concepts into exact, testable conditions, and where the framework's honest weaknesses lie.

Where it came from
Fair value gaps: anatomy of an imbalance
Market structure: BOS and CHoCH
Order blocks and breaker blocks
Liquidity: pools, sweeps and inducement
Premium, discount and the 50% level
Kill zones and sessions
How the engine mechanises it
Strengths and failure modes
Deeper reading
Glossary
FAQ
In the Systems Library

Where it came from

SMC is best understood as a modern retail reframing of much older institutional order-flow ideas. None of its raw material is new. Richard Wyckoff was writing about accumulation, distribution and the "composite operator" (his stand-in for the coordinated behaviour of large players) in the early twentieth century, and his schematic of quiet accumulation followed by a markup phase is a direct ancestor of the SMC "accumulation, manipulation, distribution" cycle. Auction market theory, developed around futures markets in the 1980s, contributed the idea that price moves to seek out liquidity and that markets rotate around areas of accepted value. Floor-trading folklore about stop runs, the observation that obvious stop clusters get tagged before the real move, is older still. If SMC feels familiar to a reader of the Wyckoff Method, that is because it largely is Wyckoff, restated in the language of the modern forex chart.

What is new is the packaging. The framework was popularised online, above all by Michael Huddleston, who teaches under the name Inner Circle Trader, usually shortened to ICT. Through a long-running series of free YouTube mentorship videos he built a very large online following and coined or standardised much of the vocabulary this page uses: fair value gaps, kill zones, the Judas swing, and many more. The wider SMC movement grew out of and around that material, with other educators renaming, remixing and extending the concepts, which is why terminology drifts between communities: one person's "inducement" is another's "minor liquidity", one person's "supply zone" overlaps heavily with another's "bearish order block".

Two things are worth stating plainly. First, crediting the lineage honestly does not diminish SMC; giving old ideas crisp names and chart-readable definitions is genuinely useful. Second, the framework's central narrative, that these patterns are the visible footprints of institutional dealing, is an interpretation layered on top of geometry. The geometry is real and testable. The story about who caused it is not something a price chart can prove.

Fair value gaps: anatomy of an imbalance

A fair value gap (FVG) is a three-candle pattern. When price moves fast enough in one direction, the middle candle of a three-bar sequence can be so large that the wicks of its neighbours fail to overlap. In a bullish FVG, the low of the third candle sits above the high of the first, leaving a window of prices that traded only once, on the way through, inside the big middle candle. That untraded window is the gap.

Three candles forming a bullish fair value gap between candle 1's high and candle 3's low, with price later returning to mitigate the zone.

SMC reads the gap as an imbalance: buying was so one-sided that sellers never got to transact through those prices. The framework then expects two possible behaviours. The gap may act as a magnet, drawing price back down to "rebalance" it, or as support, with price returning into the zone, finding buyers, and continuing higher. A return into the gap is called mitigation; trading all the way through it is a fill. A close through the far side of the gap invalidates it. There is also the inverted FVG, in which an old bearish gap is reclaimed from below and flips into support, a role reversal much like old resistance becoming support in classical technical analysis.

Two honest caveats. Price often revisits gaps simply because markets oscillate; "gaps get filled" is partly a statement about volatility, not a secret. And a gap has two contradictory readings, magnet and springboard, so any outcome can be narrated after the fact. A testable version must pick one behaviour, define it exactly, and measure it.

Market structure: BOS and CHoCH

Markets swing. An uptrend, in structural terms, is a sequence of higher highs and higher lows; a downtrend is the reverse. SMC formalises two events on that skeleton.

A break of structure (BOS) is a break beyond the most recent confirmed swing point in the direction of the prevailing trend: in an uptrend, price takes out the last swing high. It is a continuation signal, the trend confirming itself. A change of character (CHoCH) is the first break against the prevailing direction: a downtrend that has been printing lower lows suddenly takes out a prior lower high. It is the earliest structural hint that the trend may be turning, and correspondingly less reliable than a BOS, because first hints are often false.

Downtrend of lower highs and lower lows with a BOS marked at a broken low and a CHoCH marked at the first break above a lower high.

SMC also distinguishes internal structure from swing structure. The big swings visible on your chart define swing structure; the smaller zigzags inside each leg define internal structure. A common playbook is to let swing structure set the directional bias and use an internal CHoCH, inside a pullback, as the entry trigger. The distinction matters because it is also the framework's biggest source of ambiguity: with no fixed rule for which swings "count", two traders can look at the same chart and see opposite structures. A mechanical implementation removes the ambiguity by pinning swing confirmation to an exact fractal or lookback rule and never revising it, as discussed under timeframes, since structure on one timeframe is noise on another.

Order blocks and breaker blocks

An order block is usually defined as the last opposite-direction candle before a displacement: for a bullish order block, the last down-candle before the strong up-move that breaks structure. The SMC narrative is that this candle marks where large buy orders were being accumulated while price ticked down, and that unfilled orders remain in the zone, so a later return should find support there.

A gold order-block candle before a green displacement leg, its zone projected forward and holding on a later retest.

Practitioners refine the zone in several ways: using only the candle body rather than the full wick range, requiring the displacement to leave a fair value gap (evidence of genuine urgency), or dropping to a lower timeframe to isolate a tighter block inside the higher-timeframe one. A related pattern is the breaker block: an order block that failed, was traded through, and is then expected to act in the opposite role on a return, again a cousin of support-becoming-resistance. In practice an order block is a supply-or-demand zone with a stricter birth certificate: it must be the last opposite candle, and it must precede displacement, which is exactly what makes it definable in code.

Liquidity: pools, sweeps and inducement

Liquidity, in the SMC sense, means resting orders: the stop-losses and pending entries clustered at obvious chart levels. Above equal highs sit the stops of short sellers and the buy-stops of breakout traders; below equal lows, the mirror image. SMC calls these clusters liquidity pools: buyside liquidity above highs, sellside liquidity below lows. The reasoning is sound in outline: stops are real orders, and clustered stops are a pocket of guaranteed volume, useful to anyone who needs size filled.

A liquidity sweep (or stop hunt, or raid) is the signature pattern: price spikes through the obvious level, triggers the stops, then reverses back through it, often quickly. The wick through equal highs that immediately collapses is read as large players selling into the burst of forced buying. Inducement is the related idea of a smaller, nearer pocket of liquidity that gets engineered and swept first, luring early entries offside before the real level is run. ICT's Judas swing is a session-specific version: a false directional push around the London open that reverses and sets the real tone for the day.

Two equal highs with resting buy-stops marked above, a gold sweep spike through the level, then a sharp rejection lower.

The caveat belongs here more than anywhere: a sweep label is geometric. "The bar traded above the level and closed back below it" is a fact you can test; "institutions hunted the stops" is a story about intent that the chart cannot verify. Plenty of sweeps are just failed breakouts, and the same wick that SMC calls a raid, a breakout trader calls a false break and a mean-reversion trader calls an overextension snapping back. Three vocabularies, one shape.

Premium, discount and the 50% level

Take a dealing range, the distance from a significant swing low to a significant swing high, and mark its midpoint. SMC calls everything above the 50% level premium and everything below it discount, borrowing the language of value: buy at a discount, sell at a premium. The midpoint itself is often labelled the point of equilibrium, and it coincides with the 50% retracement of classical Fibonacci analysis. The practical use is as a filter: a bullish setup, say a CHoCH into an order block, is rated more highly if it forms in the discount half of the range, because the long is being taken from cheap prices rather than chased at expensive ones. It is a simple, fully mechanical idea, and one of the easiest SMC concepts to test, since it reduces to "where is price within the last N-bar range".

Kill zones and sessions

The forex day has a rhythm. The Asian session is typically the quietest stretch, often carving out a narrow overnight range. London's open brings the first heavy European volume and frequently the day's first decisive move, sometimes a false one, the Judas swing, before the true direction asserts itself. New York overlaps London for a few hours in what is usually the most active window of the day. ICT's kill zones are named windows around these events, the London open and New York open above all, in which his methodology concentrates its entries: the idea being that displacement, sweeps and gap-creating moves need volume, and volume keeps office hours.

The Asian range itself becomes an object: its high and low are treated as the first liquidity pools of the day, with London expected to sweep one side before travelling. Mechanically, though, a kill zone is a clock filter and nothing more. It reads the time, not the price, and it only means anything on intraday charts; on four-hour bars and slower, a "session" is one or two bars, and the filter stops making sense. Time filters of this family also interact strongly with momentum-style entries, since both are really bets on when participation arrives.

How the engine mechanises it

Everything above can be narrated loosely or defined exactly. Our engine takes the second road: each SMC concept is pinned to a precise bar-by-bar condition that either fires or does not, with no discretion in the loop. A few examples, paraphrased from the exact rules our systems use:

Bullish fair value gap forms: this bar's low is above the high of two bars ago, and the gap is wider than 10% of the 14-period average true range. That width floor matters: without it, every microscopic three-candle non-overlap would count as an "imbalance".
Break of structure up: the bar breaks above the prior confirmed swing high on an up close, with its high-to-low range over one ATR and volume above 1.2 times its 20-bar average. The swing high is locked by the structure engine in advance, so the rule cannot quietly repaint which high "counted".
Sellside liquidity swept: the bar wicks below a confirmed sellside liquidity level and closes back above it on an up-bar. The label is geometric: it certifies the shape of the excursion and recovery, not that any institution acted.
CHoCH plus order block: the structure engine's bullish change-of-character state is active, price trades inside the active bullish order-block zone (the body of the last bearish candle before the break of structure, valid until price closes below it), and the bar closes above its open. Structure, location and confirmation in a single rule.
Kill-zone filter: the bar's server-time hour falls inside fixed session windows. On four-hour charts and slower the gate is lifted entirely, because hours stop meaning much at that scale, and session-flavoured rules fall back to pure price definitions such as a Donchian-style range break.

The honest fine print travels with the rules. Gap and sweep conditions describe geometry, and geometry is all a backtest can score. "Volume" on a forex chart is tick volume, a count of price updates used as a proxy for activity, not a true record of traded size, so volume-conditioned rules inherit that approximation. And no condition here predicts anything: each one classifies the present bar, and whether that classification carried any edge in the past is exactly what testing exists to measure, on data the rule never trained on.

Strengths and failure modes

SMC's genuine strengths are worth naming. It focuses attention on real market phenomena: stop clustering at obvious levels is real, session rhythms are real, fast one-directional moves that leave untraded windows are real. It supplies a shared vocabulary precise enough that most of its patterns can be written down as code, which is more than can be said for a lot of discretionary lore. And its emphasis on structure and location pushes traders towards asking "where am I in the range, and what just broke" rather than stacking lagging indicators.

The failure modes are just as specific. The core narrative, that a given wick is smart money hunting stops, is untestable: no retail chart shows who traded, and the claim survives any outcome. The framework is also prone to redefinition after the fact: a failed order block becomes a breaker, a failed sweep becomes inducement for the real sweep, a CHoCH that reversed becomes "internal, not swing" structure. Each renaming is individually reasonable, but together they mean the framework can explain every chart in hindsight, which is precisely the property that makes a method feel true while telling you nothing about tomorrow. A related trap is zone density: an active chart carries so many overlapping FVGs, order blocks and liquidity levels that hindsight can always find one that "worked". None of this makes the patterns worthless: it makes the stories unfalsifiable while the shapes stay testable. A three-candle gap, a locked structure break, a wick-and-reclaim through a level: these are exact events with countable historical outcomes. The defensible position is to test the geometry, keep whatever survives honest out-of-sample validation, and leave the mind-reading out, the same discipline that applies to trend following or any other style.

Deeper reading

Market microstructure (Wikipedia): the academic field that actually studies how orders become prices.
Order flow trading (Wikipedia): reading the flow of trades directly, the data SMC tries to infer from bar shapes.
Richard Wyckoff (Wikipedia): the early-1900s analyst whose accumulation and distribution schematics prefigure much of SMC.
Orders and stop orders (Wikipedia): how stop-loss and stop-entry orders actually work, the raw material of "liquidity pools".
Market liquidity (Wikipedia): what liquidity means in finance generally, beyond the SMC usage.
Auction theory (Wikipedia): background on the auction framing that market-profile and value-area thinking grew from.
Technical analysis (Wikipedia): the broader discipline, its history and the evidence debate around it.
Foreign exchange market (Wikipedia): how the decentralised FX market and its sessions are actually organised.

Glossary

Fair value gap (FVG): A three-candle imbalance in which the wicks of the first and third candles fail to overlap, leaving an untraded price window inside the middle candle's range.
Mitigation: A later return of price into a gap or zone, partially or fully trading through the prices it skipped.
Break of structure (BOS): A break beyond the most recent confirmed swing point in the direction of the prevailing trend; a continuation event.
Change of character (CHoCH): The first structural break against the prevailing trend, read as an early hint of reversal.
Internal structure: The smaller swings inside a single leg of the larger (swing) structure; often used for entry timing while swing structure sets bias.
Order block: The last opposite-direction candle before a displacement that breaks structure, projected forward as a zone expected to hold on retest.
Breaker block: A failed order block that price traded through, then expected to act in the opposite role when revisited.
Displacement: A fast, large, one-directional move, often leaving a fair value gap, treated as evidence of urgent participation.
Liquidity pool: A cluster of resting stop and entry orders assumed to sit above equal highs (buyside) or below equal lows (sellside).
Liquidity sweep: An excursion through an obvious level that triggers resting stops and then reverses back through it; also called a stop hunt or raid.
Inducement: A nearer, smaller pocket of liquidity that gets swept first, drawing early entries offside before the main level is run.
Judas swing: ICT's term for a false directional move around the London open that reverses and sets the session's real direction.
Premium and discount: The upper and lower halves of a dealing range, split at its 50% midpoint; longs are preferred from discount, shorts from premium.
Kill zone: A fixed clock window around a session event (London open, New York open) in which entries are concentrated; a time filter, not a price signal.
Tick volume: A count of price updates per bar, used on forex charts as a proxy for traded volume, which retail FX data does not truly report.

FAQ

Is SMC the same thing as ICT?

Not exactly. ICT is a specific teacher, Michael Huddleston, and his body of material; SMC is the broader retail movement that grew out of and around it. They share most core concepts, but naming and exact definitions drift between communities.

Do fair value gaps always get filled?

No. A fair value gap is simply an untraded price window. Price often revisits gaps because markets oscillate, but there is no rule that it must, and a tested system treats a gap as one condition among several, never a certainty.

Can Smart Money Concepts be traded mechanically?

Yes, once each concept is pinned to an exact definition. A gap, a locked structure break, a wick-and-reclaim sweep and a session window can all be written as precise bar-by-bar conditions, which is what makes them testable at all.

Does SMC actually show what banks are doing?

No chart can. Retail price data carries no record of who traded or why, and forex volume is a tick-count proxy. SMC's patterns are real geometry; the institutional story attached to them is an interpretation that cannot be verified from a chart.

Is SMC just Wyckoff with new names?

Largely, though not entirely. Accumulation, stop runs and imbalance all predate SMC by decades, and Wyckoff's schematics map closely onto the accumulation-manipulation-distribution cycle. SMC's real contribution is a compact, chart-readable vocabulary and a set of definitions crisp enough to code.

In the Systems Library

The Orion RFX Systems Library tags mechanical systems by style, including SMC and ICT, and every system ships with a plain-English playbook of its exact rules, with each one validated on data it never trained on. If the mechanised conditions above interested you, read how automated trading systems work to see how a paraphrased rule becomes a running strategy. To place SMC in its family tree, compare it with the Wyckoff Method it descends from, and with breakout trading, the style whose failures SMC renames as sweeps.