Smart Money Concepts

Any order can only fill against the liquidity that exists at that moment. An order larger than the resting interest on the other side has two choices: push price against itself, or wait for a moment when a crowd of opposite orders arrives at once.

Stop orders create those moments. Traders who are long place stop losses just below an obvious low, and breakout traders place sell orders in the same spot. When price trades there, those orders turn into market sells together. A buyer with size can fill against that burst of selling, and the fired stops often push price further in a quick cascade before it turns.

Above highs the picture is the mirror image: short sellers' stop losses and breakout buy orders.

BSLBuy-side liquidity

Buy stops resting above highs. They supply the buyers a large seller needs.

SSLSell-side liquidity

Sell stops resting below lows. They supply the sellers a large buyer needs.

This part is measured, not assumed. Order records from a major currency dealing bank showed buy stops clustered just above round numbers and sell stops just below, and prices moved unusually fast once they reached those clusters. Chapter 03 separates what is measured from what is only convention or story.

Price tends to travel from one pool of liquidity to the next. The rest of SMC is vocabulary for the steps along the way.

On every timeframe, the trip between pools runs through the same five phases. Each SMC term belongs to exactly one of them, so once the phases are clear, every new term has an obvious place to go. The small charts are slices of the master chart, with the phase's bars shaded.

1. 1 Build

   Stops collect beyond highs and lows. Equal highs and equal lows collect the most.

   BSL, SSL, equal highs and lows, inducement

   WyckoffTrading range

2. 2 Sweep

   Price runs one pool. The fired stops supply the other side of a large order.

   Sweep, stop run, SMT divergence

   WyckoffSpring

3. 3 Displace

   Price moves hard the other way and closes through a swing.

   Displacement, CHoCH, CISD

   WyckoffSign of strength

4. 4 Return

   Price pulls back into the zone the fast move left behind.

   FVG, order block, breaker, OTE, premium and discount

   WyckoffLast point of support

5. 5 Expand

   Price travels to the opposite pool, which becomes fuel for the next cycle.

   BOS, liquidity targets, take profit

   WyckoffMarkup

The fifth phase feeds the first. The pool that price runs to becomes the edge of the next range, and the cycle starts again. Compressed into one session or one day, the sequence is often called the Power of Three: accumulation (build), manipulation (sweep) and distribution (displace, return, expand).

The sequence is much older than SMC. Richard Wyckoff described the same shape for stocks in the early twentieth century as accumulation: a trading range, a spring below its low, a sign of strength, a last point of support, then markup. The bottom line of each card gives his name for the phase.

SMC mixes three kinds of statements: things measured in market data, conventions that make charts readable, and stories about who is trading. They deserve different amounts of trust.

Most of the vocabulary was popularized by Michael J. Huddleston's Inner Circle Trader (ICT) material. The ideas underneath are older: Wyckoff's accumulation and distribution schematics and classic supply and demand zones. The practical rule is simple. Rely on the measured parts, use the conventions as definitions, and test everything else before risking money on it.

Obvious levels attract the most stops, and equal highs or equal lows are the most obvious of all. Two touches at the same price look like strong support or resistance, which invites stops just beyond them. In stocks, traders watch the prior day's high and low, the premarket high and low, the opening range (the first 15 or 30 minutes), and round numbers such as $500.

"Equal" needs a tolerance. A fraction of the average true range scales across stocks and timeframes: 0.1 × ATR(14) is about $0.05 on the master chart, which is exactly how far apart its three lows sit.

When price trades through a pool, the close tells you what happened.

Structure is the record of swing points. A swing high is a bar whose high sits above the highs on either side of it, and a swing low is the mirror image. An uptrend prints higher highs (HH) and higher lows (HL). A downtrend prints lower highs (LH) and lower lows (LL).

Two events matter, and both require a candle close beyond a swing.

BOSBreak of structure

A close beyond a swing in the direction of the trend. The trend is still being delivered.

CHoCHChange of character

The first close beyond a swing against the trend. Control has changed hands, at least for now.

Only closes count. A wick beyond a swing means liquidity was taken, while a close beyond it means structure changed. That rule alone removes many false signals.

In code, a swing is usually a pivot: a bar with N lower highs on each side (N = 2 on the master chart). That definition has a cost. A pivot is only known N bars after it forms, so an indicator that draws it on the pivot bar is showing information that arrived later. Any backtest has to account for that lag.

Internal and external structure

External structure is the set of swings that define the current leg on your chosen timeframe, like the ones labeled above. Internal structure is the smaller swings inside a leg. Read direction from the external swings and use internal breaks only to time an entry. An internal CHoCH against the external trend is usually just a pullback.

After a sweep, price often moves one way in large-bodied candles that barely overlap. This is displacement. A usable definition is a candle whose range is at least 1.5 times the recent average true range and that closes near its extreme. Bar 27 on the master chart spans $1.10, 2.3 times the 14-bar ATR of $0.48, and closes 10 cents from its high.

Displacement leaves a footprint, a three-candle pattern called a fair value gap (FVG). Prices inside the gap traded only during one fast candle, almost entirely in one direction.

The midpoint of the gap is called consequent encroachment (CE). The usual explanation for returns into a gap is that the fast move left little two-way trade there. Treat it as a tendency to measure, not a law: in strong trends many gaps are never revisited. In stocks, the overnight gap between sessions, especially after earnings, is the largest gap you will see.

An order block is the last opposing candle before displacement: the final down candle before a strong rally, or the final up candle before a strong drop. SMC reads it as the last point where the other side was in control. Its practical value is simpler. It is a small, well-defined zone that gives the entry and the stop a clear location.

A candle only earns the name when three things are true. It swept liquidity or sits right beside a sweep. The move away from it displaced and left a fair value gap. That move broke structure. Without all three, it is just the last red candle.

Breakers

When price closes through an order block instead of respecting it, the usual explanation is that traders who entered there are now losing, and many exit near breakeven when price returns to the block from the other side. Whatever the cause, the failed block is called a breaker, and it is traded in the opposite direction.

A leg with a clear swing low and swing high defines a dealing range. Its midpoint, equilibrium (EQ), splits it in two. Above equilibrium, price is at a premium, and below it, at a discount. The reason to buy in discount is arithmetic: the deeper the entry, the shorter the stop and the longer the distance to the target, so the same idea pays more R.

The optimal trade entry (OTE) narrows the discount to the 62% to 79% retracement band, with 70.5% as the traditional midpoint. Treat those ratios as a definition of a deep pullback rather than as precise levels. What matters is that the band overlaps a real zone.

Liquidity builds while the market is quiet and gets taken when volume arrives. For US stocks, that rhythm follows the session clock and the calendar.

A classic study of currency markets found that a major scheduled release, the US employment report, produced a jump of roughly 576% in five-minute volatility that faded within the day. The lesson carries over to stocks. The first minutes after scheduled news belong to whoever is fastest, and a limit order resting in a zone can fill at a much worse price than planned.

Put the pieces in order and a complete setup appears. Assume MSFT's daily and 1 hour charts are in an uptrend, so the morning selloff is a pullback and only longs are considered. Each number on the chart matches a step below.

1. Build. Three equal lows at $496.70 to $496.75 (bars 15, 20 and 23) hold a pool of sell stops.
2. Sweep. Bar 26 (11:35) runs the pool to $496.15 and closes back above it at $496.80.
3. Displace. Bars 27 and 28 rally $1.35. Bar 28 closes at $498.15, above the $497.80 lower high (CHoCH), and leaves a gap from $497.05 to $497.60.
4. Return. Bar 32 (12:05) pulls back to $496.95, into the gap, the order block and the 62% level. A limit buy at $497.05 fills, with the stop at $496.05, 10 cents under the sweep low.
5. Expand. Bar 34 (12:15) closes at $498.60, above the $498.25 swing high (BOS), and takes the first pool of buy stops at $498.65.
6. Target. Bar 40 (12:45) trades to $500.10, through the $500.00 high and the buy stops above it. The trade closes at the target.

At almost 3R per winner, a trader who is right one time in three still comes out ahead, before costs. Chapter 11 adds costs and shows what that edge feels like over 100 trades.

Two ways to enter

Limit at the zoneused here

Rest a limit order where the gap, the order block and the OTE band overlap. Best R and no chasing, but some trades never fill, and you buy while price is still falling.

Confirmationlower timeframe

Wait for a 1-minute CHoCH inside the zone, then enter. Fewer losers and a clearer invalidation, but a worse price and some missed moves.

What cancels the setup

• A candle closes below the sweep low at $496.15. The idea is wrong, whether or not the limit filled.
• Price reaches $498.65 before returning to the zone. The move happened without you, so do not chase it.
• Earnings, 8:30 data or an FOMC decision is due before the trade can reasonably play out.

A correct read of the chart is worth nothing without the arithmetic around it. Three numbers decide whether the model survives: costs, position size and losing streaks.

Costs

Spread, commission and slippage come out of every trade. MSFT is one of the most liquid stocks in the world, but a one-cent spread and a cent of slippage still add up to $0.02 a share round trip. Costs rise at the open and around news, exactly when these setups tend to trigger.

Position size

Size follows the stop, never the other way round. With a tight intraday stop, buying power usually caps the position before the risk budget does. 20 shares in a cash account risk about $20, well under 1% of the account, and that is fine. Widening risk per trade to make a small account feel busy is not.

Losing streaks

This model has a real edge of +0.29R per trade after costs, and it still produces long runs of losses. At 1% risk per trade, the typical run of 100 trades draws down about 11% from a peak, and one run in ten draws down 18% or more. Plan for it before it happens: risk 0.5% to 1% per trade, set a daily and a weekly loss limit, and never raise size to win back a drawdown. Margin magnifies all of this, since at 2:1 or 4:1 buying power the same move against a full position costs two or four times as much of the account.

Managing the trade

The example holds to the final target. Two common alternatives change the numbers. Taking half off at the first pool ($498.65, +$1.60, 1.6R) lowers the payoff per winner but books profit earlier. Moving the stop to breakeven after the BOS protects capital but turns some eventual winners into scratches. Pick one rule and test it rather than deciding trade by trade.

SMC is taught as discretionary chart reading, which makes it easy to fool yourself: looking back, every sweep is obvious. The fix is to write each rule precisely enough that code could apply it. These starting definitions match the master chart.

1. Code the rules, or at least write them so two people would mark the same chart the same way.
2. Include spread, commission and slippage in every trade.
3. Collect at least 100 trades per variant before trusting a win rate. With a true 33% win rate, 100 trades still produce a measured rate anywhere from about 24% to 43% one time in twenty.
4. Design on one period and test on another that you have not looked at.
5. Forward test at small size before trading full risk.
6. Track expectancy, profit factor, maximum drawdown and the longest losing streak, not just the win rate.

My own version of these rules runs as an indicator on TradingView.

Most of the remaining vocabulary refines one phase of the cycle. Each term below is placed where it belongs.

Nearly every one of these comes from using a term outside its place in the cycle, or from skipping the arithmetic.

• AvoidMarking every gap and block on the chart.

  InsteadKeep only zones created by a sweep, displacement and a structure break.

• AvoidTreating a wick as a break.

  InsteadWait for a candle to close beyond the level.

• AvoidTrading a 5-minute CHoCH against the daily trend.

  InsteadTake direction from the higher timeframe and use the lower one only to enter.

• AvoidBuying in premium because the pattern looks clean.

  InsteadBuy only in the discount of the current dealing range, and sell only in its premium.

• AvoidPlacing the stop inside the sweep.

  InsteadPut it beyond the swept extreme, where the idea is proven wrong.

• AvoidChoosing targets by R multiple.

  InsteadTarget the next liquidity pool, then decide whether that R is worth taking.

• AvoidTreating the first minutes after 9:30 like any other time.

  InsteadSpreads are widest at the open. Let the opening range form, or size down.

• AvoidHolding through earnings or an FOMC decision by accident.

  InsteadCheck the calendar. Scheduled news can gap price through any level.

• AvoidSizing by feel, or adding size after losses.

  InsteadSize every trade from the stop at a fixed risk, and stop for the day after a set loss.

• AvoidBacktesting by eye.

  InsteadCode the rules, include costs, and test on data you did not use to design them.

Pre-filled with the example trade from chapter 10. Uncheck anything your own setup lacks.