Smart Money Concepts (SMC)

Smart Money Concepts (SMC) is a discretionary price-action framework that reads a chart as a map of where orders rest and where they were filled. Instead of deriving signals from indicators, an SMC trader annotates price itself: where structure broke, where an impulsive move left an unfilled gap, and where clusters of stop orders are likely to sit. The underlying premise is that large participants need counterparty volume to fill size, which is offered as the explanation for why price so often runs through an obvious high or low before turning. That narrative is not testable. The annotation rules built on top of it are, and keeping the two apart is the most useful habit when studying any version of this method.

## The shared vocabulary

Almost every SMC variant is assembled from the same components. **Market structure** is the sequence of swing highs and lows; a **break of structure (BOS)** continues that sequence, while a **change of character (CHoCH)** is the first break against it and is read as a possible shift in control. An **order block** is the last opposing candle, or candle cluster, before a displacement move. A **fair value gap (FVG)**, also called an imbalance, is a three-candle formation whose outer wicks do not overlap, leaving an area that traded in one direction without two-sided participation. **Liquidity** refers to zones where stops are likely to be clustered — equal highs and lows, session and daily extremes, obvious swing points, trendline touches — and a **sweep** or **raid** is a move through one of them that fails to hold. **Premium and discount** split a dealing range at its midpoint, so shorts are sought in the upper half and longs in the lower half. **Mitigation** describes price returning to an unfilled area; **inducement** describes a minor liquidity pool that attracts early entries before the intended move.

## How a setup is usually constructed

The workflow is top-down. A higher timeframe — daily, 4-hour or 1-hour — supplies directional bias and a dealing range. Points of interest are marked inside it: unmitigated order blocks, open FVGs, and the liquidity pools above and below. The trader then drops to a lower timeframe, often 15-minute down to 1-minute, and waits for price to reach the area and take out the identified pool. Confirmation is typically a lower-timeframe CHoCH followed by displacement, with the entry placed on the order block or FVG that leg leaves behind. The stop sits beyond the swept extreme or the far side of the zone, and the target is the opposing liquidity pool — which is why SMC setups are normally expressed in R multiples rather than fixed pip objectives. Many versions add a time filter, trading only around session opens or defined windows.

## Main variants

The versions collected on this page differ less in vocabulary than in which component carries the decision:

- **Structure-first**: BOS and CHoCH drive everything; zones only refine entry. - **Liquidity-first**: the trade begins with a sweep of a specific pool, and structure confirms afterwards. - **Imbalance-centred**: FVGs are the primary entry object, with order blocks treated as secondary. - **Multi-timeframe alignment**: the setup is only valid when two or three timeframes agree on direction and location. - **Confluence hybrids**: SMC is layered onto classical price action, supply and demand zones, session levels or moving averages, requiring two independent reasons before entry. - **Instrument-specialised**: rule sets tuned to one market — gold scalping, a single FX pair, or index futures and CFDs — where spread, session behaviour and volatility shape the parameters.

## What separates one implementation from another

The differences that matter are definitional. How exactly is an order block drawn — body or wick, single candle or cluster, refined to its internal FVG? Does a BOS require a body close beyond the swing, or is a wick enough? Is entry a resting limit order at the zone, or does it demand lower-timeframe confirmation first? Which timeframe pair is used, is higher-timeframe bias mandatory, and is there a session filter? Where does the stop go, and is the target a fixed R multiple or a specific liquidity pool? Two strategies using identical terminology can produce almost non-overlapping trade lists once these choices are pinned down.

## Common mistakes

The most frequent failure is hindsight labelling: on a completed chart, every reversal has an order block behind it. Related problems include marking so many zones that some level is always nearby, treating a lower-timeframe CHoCH as a trend reversal, assuming a sweep must reverse rather than waiting for the stated confirmation, and moving invalidation after entry. Cost is routinely underestimated — tight SMC stops on gold or around session opens are sensitive to spread and slippage in a way that a clean chart never shows. Finally, the institutional narrative is often used to explain outcomes after the fact; it adds no edge on its own.

## Evaluating and backtesting a version

Start by rewriting the rules until they are unambiguous enough to code, even if you intend to trade them discretionarily — if the order block definition cannot be expressed as an algorithm, it cannot be tested. Then replay bar by bar without scrolling ahead, logging every candidate setup including the ones you skip, since skipped trades are where discretion hides. Record results in R with entry, stop, target, and maximum adverse and favourable excursion, and include realistic spread, commission and slippage for the instrument. Run the sample across different volatility regimes and more than one period, and keep an out-of-sample segment untouched. Where possible, test components in isolation: does requiring a sweep improve results over entering at the zone alone? Does the higher-timeframe filter earn its place? For discretionary rule sets, re-label the same period weeks later and compare — if your own annotations do not agree with themselves, the backtest is measuring you, not the method.

Strategies in this concept (83)

Frequently asked questions

Is SMC different from supply and demand trading?

They overlap heavily. Both mark zones where an impulsive move originated and trade the return to them. SMC adds an explicit structural layer (BOS/CHoCH), a liquidity model that treats stop clusters as targets, and the fair value gap as a distinct entry object. In practice many supply and demand traders already use SMC logic under different names.

What is the relationship between SMC and ICT?

Much of the SMC vocabulary — order blocks, fair value gaps, liquidity raids, premium and discount — comes from the ICT body of material, which adds a strong time component such as session windows and specific intraday sequences. SMC is generally used as the broader, more generic label, and many versions keep the structural concepts while dropping or simplifying the time-based rules.

Which timeframes are typically used?

Most versions pair a higher timeframe for bias and zone selection with a lower one for entry: daily or 4-hour down to 15-minute is common for day trading, 1-hour down to 1-minute for scalping. The specific pair is a real parameter, not a detail — the same rules applied to a different pair produce a different strategy and need separate testing.

Can SMC be backtested or automated?

Yes, but only after the definitions are made mechanical. Order blocks, valid breaks of structure, and which liquidity pools count all have to be reduced to explicit conditions before any test is meaningful. Versions that keep a genuinely discretionary step can still be evaluated through logged bar-by-bar replay, provided skipped setups are recorded and the labelling is checked for consistency over time.

Does SMC work the same on every market?

The concepts are applied to forex, gold, indices, futures and crypto, but the parameters do not transfer unchanged. Session structure, typical range, spread and how cleanly levels are respected differ by instrument, so a rule set tuned to gold scalping and one tuned to a single FX pair on the 15-minute chart are separate strategies even when the terminology is identical.

What is the minimum needed to define an SMC setup precisely?

Six things: the timeframes used, how the directional bias is established, the exact drawing rule for the zone, which liquidity pool must be taken and how a sweep is confirmed, what counts as entry confirmation, and where the stop and target sit. If any one of those is left to judgement, two traders following the same description will take different trades.

All strategy concepts · Explore all strategies