What an SMC trading strategy actually is
An SMC trading strategy is the smart money concepts method assembled into a single setup, and it stacks a directional bias, a time window, a liquidity event, a structure break and an entry object rather than trading any one pattern on its own. ICT and SMC traders use the term for this layered read of price, built on the Inner Circle Trader framework of Michael Huddleston (The Inner Circle Trader, ICT methodology).
The strategy is the capstone of the cluster because every other concept feeds into it. Bias comes from the dealing range and the trend, timing from the killzones, the trigger from inducement and the sweep, confirmation from the break of structure, and the entry from the order block and the fair value gap.
I frame the strategy as a method rather than a system, because the same stack can be described perfectly and still lose if the entry zone fails. The discipline of waiting for the layers to align is the method's genuine contribution, separate from whether any single assembled setup pays.
The layers of an SMC setup
An SMC setup runs through six layers in a fixed reading order, and the order is the point. Traders set a higher-timeframe bias, wait for the right time window, let the inducement and sweep resolve, confirm with a break of structure, enter at the order block or fair value gap, and target the draw on liquidity (The Inner Circle Trader, ICT methodology).
| Layer | Question it answers | ICT concept |
|---|---|---|
| 1. Bias | Which direction am I trading? | Dealing range premium or discount, plus the trend |
| 2. Time | When am I watching? | The killzone window |
| 3. Liquidity | What gets swept first? | Inducement and the sweep |
| 4. Structure | Is the move confirmed? | A break of structure or change of character |
| 5. Entry | Where exactly do I get in? | The order block with a fair value gap inside it |
| 6. Target | Where am I aiming? | The draw on liquidity |
I list the layers as a reading order rather than a checklist to tick. A setup that has an entry object but no bias is not a partial SMC trade, it is a candle traded in a vacuum, and the method treats the missing layers as reasons to sit out rather than reasons to shrink the position.
Step one: set the higher-timeframe bias
The strategy starts with direction, and ICT traders read it off a higher timeframe than the one they trade. The dealing range splits into a premium and a discount half at its midpoint, and a sequence of higher highs or lower lows sets the trend, so the bias is bullish when the higher timeframe sits in discount and bullish structure, and bearish when it sits in premium and bearish structure.
I set the bias before I look at the entry timeframe, because the entry only makes sense in the direction the higher timeframe is already leaning. A long inside a higher-timeframe premium is trading against the bias, which the methodology reads as the wrong side of the larger range.
The bias is a filter, not a forecast. It tells me which setups to take and which to skip, and the discipline of trading only with the higher-timeframe direction is what stops the method from firing on every counter-move that looks tempting on the lower timeframe.
Step two: wait for the killzone
With the bias set, the strategy narrows to a time window. ICT traders watch the killzones, the session opens when expansion is most likely, and they wait for the setup to form inside the window rather than searching for it around the clock (The Inner Circle Trader, ICT methodology).
The window is a when, not a what. Showing up at the London or New York killzone does not produce a trade on its own, but it concentrates attention on the hours when the liquidity events and structure breaks the strategy depends on actually tend to happen.
I treat the killzone as patience enforced by the clock. Most hours produce nothing worth trading, and the window is the method's way of saying to wait for the session open instead of forcing setups in dead hours.
Step three: let the inducement and sweep resolve
This is the step where most retail traders get caught, because the obvious level looks exactly like a tradeable breakout right up until it isn't. An obvious high or low acts as inducement, pulling entries against the bias, and the setup does not begin until price sweeps that level and traps those entries (The Inner Circle Trader, ICT methodology).
The sweep is the trigger condition, and entering before it is the trap the concept describes. I let the obvious level get taken and then watch what price does next, because the move only becomes a setup once the inducement has failed.
Once the inducement is run, the entries trapped behind it are under water, and the methodology expects price to reverse toward the draw on liquidity. That reversal is the leg the rest of the stack is built to catch, which is why the strategy waits for the sweep instead of trading the bait.
Step four: confirm with a break of structure
After the sweep, the strategy asks for confirmation, and ICT traders read it as a break of structure or a change of character. A strong break closes beyond a swing point with displacement, a large body and minimal overlap, which the method treats as the sign that the move in the bias direction is real (The Inner Circle Trader, ICT methodology).
The break separates a genuine sweep from a stop run that keeps going. A weak break with no body and no follow-through is the method's own signal to distrust the move, and waiting for a strong break is what filters out the false moves that look like setups in real time.
I treat the break as confirmation rather than the entry. Trading the break itself is chasing, and the method enters on the retrace that follows, once structure has confirmed the direction the sweep implied.
Step five: enter at the order block or fair value gap
The entry comes last in the stack, and it sits at the order block the displacement left behind, ideally with a fair value gap inside it. Price retraces to that zone after the break, and the trader enters in the bias direction with a stop just beyond the block (The Inner Circle Trader, ICT methodology).
The displacement is what makes the entry worth taking. A forceful move out of the block validates it, the gap inside it confirms the force, and the retrace returns price to the level the method expects to hold, which is why the entry object and its confluence are taught together.
I size the position off the distance from the entry to the stop, because the method's edge, such as it is, lives in taking many small, planned losses and letting the aligned setups run toward the draw on liquidity. The entry is where the risk is defined, not where the profit is guaranteed.
A worked example of the full sequence
Picture a higher-timeframe dealing range in discount with a bullish bias, which tells me to look for longs. At the London killzone, price drops to sweep an obvious Asian session low, the inducement, and then reverses up through the most recent lower high with a strong, gap-leaving break of structure.
That break originated from a clean order block with a fair value gap inside it, so I mark the zone and wait. Price retraces into the block, fills the gap, and I enter long with a stop just below the block and a target at the next pool of buy-side liquidity above, the draw on liquidity.
The mirror image is a bearish short. With a higher-timeframe range sitting in premium and a bearish bias, I watch the New York killzone for price to spike above an obvious high, sweep the buy-side stops, and then break structure lower with displacement.
The retrace into the bearish order block is the short entry, with a stop just above the block and a target at the sell-side draw on liquidity resting below.
I describe both examples as templates rather than predictions. The same sequence fails when the block breaks, the sweep keeps going, or the draw on liquidity exhausts first, and the value of the template is that it tells me in advance where to enter, where to be wrong, and where to take profit.
What is actually proven about stacking confluence
The honest evidence sits one layer beneath the assembled strategy. Each observation the stack rests on is real and documented: prices revisit prior levels, which Carol Osler's Federal Reserve Bank of New York research explains through stop-order clustering; trends continue, which time-series momentum research documents, including the work of Tobias Moskowitz, Yao Hua Ooi and Lasse Heje Pedersen; and volume clusters at session opens, which the Bank for International Settlements turnover data confirms.
What is not documented is the assembled edge. No peer-reviewed or tier-1 study tests the full SMC setup, bias through target, and publishes a result, so the claim that stacking the layers produces a reliable profit is a reasonable hypothesis built on real ingredients rather than a measured finding.
I split the ingredients from the assembly because the gap between them is where retail traders get misled. The individual observations are sound, but combining sound observations does not automatically produce a tested system, and the only honest way to know whether your version of the stack works is to backtest it on your own data.
The win-rate claim nobody can source
You will read that the SMC strategy wins a high percentage of the time, or that the full stack is the most profitable setup in retail trading. I have not found a peer-reviewed or tier-1 study that tests the assembled SMC setup and publishes a win rate, and neither has any page I have seen quote one with a citation.
Every specific percentage circulating on forums and educator blogs traces back to an unsourced claim repeated until it reads like data. The honest statement is that no verified backtest of the full strategy exists in the public literature, and anyone quoting a number should link the study or retract it.
I would rather tell you the evidence is missing than invent a reassuring statistic. The absence of a clean win-rate study is itself the finding, and it means the only honest way to size the edge is to test the stack yourself on your own data and treat your result as a sample, not a certainty.
Where the strategy fails
The stack fails whenever any layer breaks, and I want the failure modes stated before the setup looks easy. The higher-timeframe bias can flip, the killzone can produce no expansion, the sweep can keep going straight through the order block, and the draw on liquidity can exhaust before price arrives, so every layer is a place the setup can die.
Over-confluence is the opposite failure, where a trader waits for so many layers to align that they never take a trade. The method is a balance between stacking enough confluence to raise the probability and stacking so much that no setup ever qualifies, and finding that line is a judgment the rules alone cannot settle.
The disciplined response is to take the setups that meet the stack, accept that a meaningful share will fail, and keep a record of both winners and losers. SMC is a framework for organising your reading of price, not a guarantee, and the traders who do well with it treat the layers as a filter that raises their odds rather than a formula that hands them profit.