Smart Money Concepts (SMC) Trading: A Practical, Honest Guide

Smart Money Concepts (SMC) is a popular discretionary price-action framework. It organizes chart reading into a repeatable sequence: identify market structure, mark where orders may be clustered, locate specific price zones the trader expects a reaction from, and wait for a confirmation event before entering. The name comes from the idea that large, well-capitalized participants leave recognizable traces in price.

That last idea deserves care. A candlestick chart records transacted price and volume. It does not record who traded, why they traded, or what they intended next. SMC can still be useful as a structured way to read charts and generate hypotheses, but it should be treated as a framework for organizing observations—not as a window into institutional intent.

This guide is the hub for the SMC topic. It explains the vocabulary, walks through the four-layer workflow, presents the main criticisms fairly, and shows how to convert loose concepts into objective rules and tags you can measure. Deeper components have their own guides: market structure, BOS and CHOCH, order blocks, fair value gaps, killzones and session timing, and liquidity sweeps.

What is SMC in trading?

SMC is a set of chart-reading conventions built around three claims. First, price tends to move between areas where resting orders are likely concentrated. Second, those areas are often visible as obvious highs, lows, and untested zones. Third, a trade has a better structure when it is taken after price has interacted with such an area and then shown a change in behaviour.

Nothing in that description is exclusive to SMC. Support and resistance, supply and demand, Wyckoff accumulation and distribution, and classic breakout-pullback trading all share parts of it. What SMC adds is a dense vocabulary and a specific ordering of steps. Whether that vocabulary improves results is an empirical question that only your own recorded sample can answer.

Where the terminology comes from

Most SMC vocabulary spread through online trading education, notably material associated with Inner Circle Trader (ICT) and the communities that grew around it. Terms were adopted, renamed, and re-defined by many educators independently. That is why two SMC traders can use the same word and mean different things—an issue that matters enormously once you start testing.

The SMC vocabulary, in plain language

• Market structure: the sequence of swing highs and swing lows used to label an uptrend, downtrend, or range.
• Break of structure (BOS): price closes or trades beyond the most recent relevant swing in the direction of the existing trend.
• Change of character (CHOCH) or market structure shift (MSS): the first structural break against the prevailing sequence, interpreted as a possible trend change.
• Liquidity: in SMC usage, chart areas where stop and entry orders are assumed to cluster—above highs, below lows, around equal highs and lows.
• Liquidity sweep or grab: a move beyond such an area followed by rejection or a return back through it.
• Inducement: an obvious minor high or low that a trader expects other participants to trade, positioned before the level the trader actually cares about.
• Displacement: an unusually large, fast, one-directional move, often used as evidence that a level “mattered.”
• Order block: the last opposing candle or small cluster before a displacement move, treated as a zone that may produce a reaction on retest.
• Fair value gap (FVG) or imbalance: a three-candle pattern where the wicks of the first and third candles do not overlap, leaving an unfilled price range.
• Breaker block: a failed order block that price traded through, re-used as a zone in the opposite direction.
• Mitigation block: a zone revisited so that earlier positions can be exited near breakeven, by assumption.
• Premium and discount: the upper and lower halves of a chosen range, used to argue that buying is preferable in the lower half and selling in the upper half.
• PD arrays: the general term for the specific price zones—order blocks, FVGs, breakers, and similar—used as entry references.
• Killzones: fixed intraday windows, usually tied to London and New York activity, in which the trader looks for setups.

The four-layer SMC workflow

Almost every SMC approach, regardless of the educator, reduces to four ordered layers: structure, liquidity, PD arrays, and confirmation. Keeping them separate is what makes the framework testable, because each layer can be measured on its own.

Layer 1 — Structure: which direction are you allowed to trade?

Structure answers a single question: on the timeframe you have chosen as your bias timeframe, is the sequence of swings making higher highs and higher lows, lower highs and lower lows, or neither? Everything downstream depends on this label, so it must be defined mechanically rather than visually.

• Choose one method for defining a swing point and never change it mid-sample: an N-bar fractal, a percentage or ATR-based threshold, or a fixed ZigZag setting.
• Decide whether a structural break requires a candle close beyond the swing or only a wick.
• Decide which swing counts as “the relevant one” when several exist close together.
• Decide what happens in a range, where structure labels flip repeatedly and produce the most false signals.

These choices are covered in detail in the BOS and CHOCH guide, including how the same chart can be labelled bullish or bearish depending purely on swing settings.

Layer 2 — Liquidity: where might the market be drawn?

The liquidity layer marks levels that were visible before price reached them: prior-day highs and lows, session highs and lows, equal highs and equal lows, range boundaries, and obvious swing points. The hypothesis is that these areas attract price because orders are likely resting there.

Two cautions apply. First, market liquidity as measured by spread, depth, and volume is not the same thing as a chart-drawn liquidity zone; see what liquidity actually means in trading. Second, price trading through such a level proves nothing on its own. The observable, testable version of the idea—breach size, reclaim rules, and confirmation—is set out in the liquidity sweep strategy guide. This guide deliberately does not repeat that material.

Layer 3 — PD arrays: where exactly would you enter?

Once you know the direction you are willing to trade and the level price may be travelling toward, the PD array layer narrows entry to a specific price range instead of “somewhere in the pullback.” The two most common arrays are order blocks and fair value gaps.

• Order blocks give a zone anchored to a specific candle, which makes invalidation easy to define but depends heavily on how you select the candle. See the order block trading strategy guide.
• Fair value gaps give a mechanically identifiable three-candle range, which is easier to code but frequently fills without any reaction. See the fair value gap guide.
• Breaker and mitigation blocks are variations that require even stricter definitions, because they depend on a prior zone having already failed.
• Premium and discount positioning is a filter, not an entry: it only tells you which half of a defined range you are in.

The main risk in this layer is optionality. On any chart there are usually several order blocks and several FVGs available. If the rule is “enter at the one that works,” the strategy has no definition. Pick the selection rule in advance—for example, the first unmitigated FVG inside the discount half of the dealing range—and accept the trades it produces.

Layer 4 — Confirmation and timing

Confirmation is the event that turns a zone into a trade: a lower-timeframe structure break, a candle close back inside a level, displacement away from the zone, or a completed retest. Timing filters, such as trading only within defined session windows, sit alongside confirmation; the trade-offs are discussed in the killzones and session timing guide.

• Limit entry at the zone: best price, worst evidence, and the highest rate of zones that simply fail.
• Confirmation entry after a lower-timeframe break: more evidence, later price, wider stop or a second structural reference.
• Retest entry after confirmation: cleanest invalidation, but many valid moves never offer the retest.
• Session filter: fewer trades and a more homogeneous sample, at the cost of missing moves outside the window.

Criticisms and limitations of SMC

SMC attracts strong opinions in both directions. The honest position is that the framework contains reasonable price-action ideas wrapped in language that makes it unusually easy to fool yourself. These are the criticisms worth taking seriously.

1. The core premise is not verifiable from a chart

Claims about what institutions are doing cannot be confirmed from open, high, low, close, and volume. Even where order-book or exchange data exists, it rarely identifies participant type or intent. A framework can be useful without its narrative being true, but the narrative should not be used as evidence.

2. It is easy to make unfalsifiable

With enough timeframes, enough zone types, and the freedom to reinterpret after the fact, almost any move can be explained retrospectively. An explanation that can never be wrong also cannot be tested. The fix is to fix your timeframes, zone types, and selection rules before you look at the outcome.

3. Terminology is inconsistent across sources

CHOCH and MSS are used interchangeably by some educators and distinguished by others. Order block definitions vary between the last opposing candle, the last down-close candle, and a multi-candle base. Your own written definitions matter far more than which source you learned from.

4. Hindsight bias is built into how it is taught

Most SMC teaching material is a chart annotated after the move completed. Marking the correct zone on a finished chart is a different skill from marking candidate zones in real time and accepting that most will fail. Replay-based practice, moving forward one candle at a time, exposes the difference quickly.

5. Selection effects in public results

Publicly shared SMC results are dominated by winning screenshots. This is not unique to SMC, but the visual appeal of the annotations amplifies it. Treat social proof as marketing and your own logged sample as evidence.

6. Complexity can crowd out risk management

Time spent refining zone taxonomy is time not spent on position sizing, reward-to-risk, and cost assumptions. An SMC entry with excellent structure and undisciplined sizing is still a losing system.

7. Not all of it is new

Order blocks resemble supply and demand zones. Liquidity sweeps resemble failed breakouts and spring or upthrust patterns from Wyckoff. Structure breaks resemble classic trend definitions. This is not a criticism of the ideas—older concepts can work—but it does mean the framework should be judged on the same evidence standard as anything else.

Turning SMC concepts into testable rules

The practical value of SMC appears only after each concept becomes a rule that produces the same answer twice. The conversion process is the same for every concept: replace an adjective with a measurement.

• “Strong displacement” becomes “a candle whose body is at least 1.5× the 20-period average body on the entry timeframe.”
• “Clean liquidity” becomes “two or more highs within 3 ticks of each other, formed at least 10 bars apart.”
• “Higher-timeframe bias is bullish” becomes “the 4-hour chart printed a close-based BOS above the last 5-bar fractal high and has not printed a close-based CHOCH since.”
• “Untested order block” becomes “the zone has not been traded into by any wick since its formation candle closed.”
• “Discount” becomes “price is below the 50% level of the range defined by the last confirmed swing high and swing low on the bias timeframe.”
• “Killzone” becomes an explicit clock window in a named timezone, with a rule for daylight-saving changes.

A rule template you can adapt

The following is an educational structure for building a testable variant, not a recommendation to trade. Every threshold should be chosen for your market and timeframe, then tested before any capital is risked.

1. Universe: one instrument or one tightly defined group, such as three major FX pairs or two liquid crypto perpetuals.
2. Bias timeframe and rule: the exact timeframe and the exact structure definition that sets the permitted direction.
3. Draw-on-liquidity: the specific level type you require price to be travelling toward, marked before the session.
4. Zone type: one PD array type only, with one selection rule when several are available.
5. Confirmation: one event, defined on a named timeframe, with a maximum number of bars it may take to appear.
6. Entry: one order type at one reference price.
7. Invalidation: a structural point plus a buffer, and a hard maximum risk in currency terms.
8. Target: one exit model—opposing liquidity, fixed R, or a defined partial-and-runner scheme.
9. Time stop: the bar count after which an unresolved trade is closed.
10. No-trade filters: news windows, spread thresholds, minimum reward-to-risk, and maximum trades per day.

A tag taxonomy for SMC trades

Tags are what let you answer “which part of this framework is actually contributing?” Keep the list short enough to apply consistently. Six to ten well-defined tags beat thirty vague ones.

• Bias timeframe state: trending, ranging, or unclear at the time of entry.
• Structure event that permitted the trade: BOS, CHOCH, or continuation without a fresh break.
• Liquidity reference: prior-day high or low, session extreme, equal highs or lows, or range boundary.
• Zone type: order block, fair value gap, breaker, or none.
• Confirmation model: limit at zone, lower-timeframe break, or retest after break.
• Session or killzone window.
• Zone freshness: first touch or a later touch.
• Rule adherence: fully followed, partially followed, or improvised.

Once 40 to 60 trades share the same tag scheme, setup analytics can show whether confirmation entries beat limit entries, or whether the edge is concentrated in one session. Be honest about sample size: see how many trades are needed to judge a strategy before drawing conclusions from a dozen results.

Risk management does not come from SMC

SMC describes where to enter and where the idea is wrong. It says nothing about how much to risk, how correlated your open positions are, or what happens after four losses in a row. Those decisions come from a separate, boring layer that determines survival.

• Fix the risk per trade in currency terms first, then derive size from the stop distance.
• Record results in R-multiples so slippage and early exits stay visible.
• Estimate expectancy per variant rather than judging by the last few trades.
• Map the realistic outcome distribution with a probability tree before assuming a high win rate.
• Write the whole thing down in a trading plan that specifies what you will not trade.

How to backtest an SMC strategy without hindsight

1. Write every rule, threshold, and timeframe before opening a chart.
2. Use bar-by-bar replay so the outcome of the current move is hidden.
3. Mark candidate zones and levels before price reaches them, and leave the failures on the chart.
4. Log every signal the rules produce, including the ones you would have hesitated on.
5. Apply realistic spread, commission, funding, and slippage assumptions.
6. Keep each confirmation model, zone type, and session as a separate tag rather than a separate memory.
7. Review the distribution of results, not just the total: the worst run matters as much as the average.
8. Re-test the finalized rules on an untouched period or a second instrument before trading them live.

The full process, including how to avoid optimizing on the same data twice, is covered in the backtesting guide.

SMC across markets

Forex

Most SMC material is taught on FX charts, which makes session structure, rollover, and scheduled economic releases central variables. Record the pair, session, and spread context in a forex trading journal; the same rule set can behave very differently during the London–New York overlap and the Asian range.

Crypto

Crypto runs continuously, so session-based filters must be defined explicitly rather than inherited from FX. Funding, liquidation cascades, venue differences, and weekend depth all change execution. Keep spot and perpetual samples separate in a crypto trading journal.

Stocks and futures

Equities add gaps, halts, earnings, and a genuine opening auction, which changes how prior-day levels behave. Index futures offer near-continuous trading with a defined session structure. Use a stock trading journal or futures trading journal and avoid pooling them with FX results.

Common SMC mistakes

• Learning the vocabulary before defining a single rule mechanically.
• Switching timeframes until one of them agrees with the desired direction.
• Re-labelling structure after the trade goes wrong so the loss becomes “not a valid setup.”
• Marking zones only on charts where the reaction already happened.
• Treating every wick beyond a high as a sweep with no minimum breach rule.
• Stacking five confluences so that qualifying setups become too rare to ever build a sample.
• Using a tiny stop just below a zone because the zone “must” hold.
• Judging the framework after 15 trades, in one market, in one month.
• Ignoring spread, commission, funding, and slippage in backtest results.
• Explaining the market fluently while keeping no record of what was actually traded.

A realistic study path

1. Week 1: write mechanical definitions for swing points, BOS, and CHOCH, and label 20 charts using only those.
2. Week 2: add one liquidity reference and mark it before each session, recording whether price reached it.
3. Week 3: add one PD array type with a fixed selection rule, and log every candidate zone including failures.
4. Week 4: add one confirmation model and run bar-by-bar replay, recording entry, stop, target, and result in R.
5. Weeks 5 to 8: collect a live or replay sample under unchanged rules, and resist adding new concepts.
6. Week 9: review by tag, keep what the data supports, and remove or isolate everything else.

A structured weekly review is what turns this into learning rather than accumulation. The goal of each review is one decision: keep the rule, tighten the rule, or drop it.

How Traderizz supports SMC testing

Traderizz lets you define each SMC variant as its own strategy, apply custom tags for structure event, zone type, confirmation model, and session, attach before-and-after screenshots, and record planned versus realized R. Analytics can then be filtered by tag so you can compare variants instead of averaging them together. You can explore the interface in the live demo.

SMC is worth studying if you treat it as a vocabulary for organizing price-action hypotheses. It becomes a liability when the narrative substitutes for measurement. The framework only starts paying for itself at the point where two traders reading your written rules would mark the same chart the same way—and where your journal, not a screenshot, tells you whether that marking is worth anything.