12 Smart Money Concepts: What They Measure and How to Test Them

Video: I’ve Studied Hundreds of Smart Money Concepts. These 12 Stand Out.
Creator: Fractal Flow – Pro Trading Strategies
Original Video: Watch the original video
Difficulty: Intermediate to advanced

The Big Idea

The video’s useful contribution is a broader vocabulary for looking at markets. Some concepts describe the geometry of candles. Others describe participation, auction activity, related markets, or options exposure. They can produce similar-looking support and resistance areas while measuring very different things.

Imagine price returning to the same level. One trader sees a swing overlap; another sees concentrated volume; another sees aggressive selling failing to push price lower. Their conclusions may agree, but their evidence is not interchangeable. A candle pattern cannot manufacture the transaction detail needed for a footprint, and an options strike does not become a reliable resistance level simply because it has a memorable name.

Our framework is observation → hypothesis → data requirement → decision rule → test. The presenter selects 12 ideas he considers useful, not a definitive ranking or one complete trading system. This guide turns that selection into a map of what each idea can reasonably establish and where research should begin.

If You Remember Only One Thing

Know which data a concept requires before treating a chart label as evidence of an edge.

60-Second Summary

• Structure shifts and sweeps describe price events; they do not identify the intent of market participants.
• Implicit fair value zones and balanced price ranges can be defined through intersections of price intervals.
• Volume profiles measure activity by price, while time-based profiles measure a different form of auction participation.
• Anchored VWAP depends on both the selected anchor and the price source used.
• Stacked imbalances and CVD require a clear definition of aggressor-side volume or an explicitly labeled estimate.
• Intermarket divergence requires aligned observations across markets whose relationship can change.
• GEX requires options information and position assumptions beyond ordinary candles.
• The executable lab below tests one overlap-zone interpretation; it does not implement all 12 concepts.

A Map of the 12 Ideas

Structure: When Does a Warning Become Confirmation?

The double shift adds a second structure change after an intervening pullback or consolidation. The presenter argues that this can improve reliability and reduce time spent waiting after an early signal. The transcript supplies illustrative examples, not a controlled first-shift versus second-shift comparison.

Waiting introduces a tradeoff. In our hypothetical long, the first signal offers entry at 100 with a stop at 96 and a chart target at 112: 4 points of risk and 12 of potential reward, or 3R. The second signal arrives at 106 with those levels unchanged: 10 points of risk for 6 of potential reward, or 0.6R. Better directional evidence can arrive with worse trade economics.

A meaningful test compares net expectancy, missed moves, drawdown, and time in the market. Define what counts as a structure break, whether a wick or close is required, and how much intervening action separates the shifts. Otherwise, two reviewers may select different patterns after seeing the outcome.

Overlap Zones: Precise Geometry, Uncertain Predictive Power

The implicit fair value zone is unusually suitable for a transparent baseline. Identify three candles whose middle candle forms a swing low. Their common price interval is an inferred demand zone. For a swing high, the same operation produces an inferred supply zone.

Our mathematical definition treats each candle as its full high–low interval. If the lows are L1, L2, L3 and highs H1, H2, H3, the intersection is [max(L1,L2,L3), min(H1,H2,H3)]. It exists only when the upper bound exceeds the lower bound. The full-wick convention and strict swing comparisons are our explicit implementation choices; the narration does not settle every edge case.

Our example has ranges [99,105], [97,104], and [100,106]. The middle low is lower than its neighbors, and the common interval is [100,104]. The zone becomes knowable only when the third candle completes. Buying an earlier touch inside the formation would use information that did not yet exist.

“Fair value” here is a label for price overlap. The calculation does not estimate a company’s intrinsic worth or directly observe order flow. Objective boundaries make the hypothesis reproducible; they do not establish profitability.

The balanced price range applies similar intersection logic to opposing gaps. Our example combines a bullish gap [100,103] with a later bearish gap [102,105], producing overlap [102,103]. An empty intersection produces no zone. A research rule must specify which gaps can pair, how long they remain valid, and when the second gap becomes known. These are different zones from the three-candle swing overlap.

Value Areas: Volume and Time Answer Different Questions

The presenter proposes narrowing a volume profile’s value area from 70% to 40%. That may create a more precise location, but precision can also reduce coverage and exclude valid reactions. A 40% area refers to a selected share of volume, not 40% of the price range.

According to TradingView’s volume-profile documentation (paid link), a value area covers a chosen percentage of profile volume, and the point of control is the price with the most volume. Row size and profile period affect the result. Some instruments use tick volume rather than exchange trade volume.

For a useful comparison, lock the profile’s start, end, row settings, and value-area algorithm before comparing 40% with 70%. Selecting the historical impulse only after its subsequent bounce introduces hindsight into the location.

The video’s value-reversal discussion concerns time-based market profiles. A TPO-style profile counts price–time opportunities. Its value area is not simply a stopwatch interval containing 70% of elapsed session minutes. Distinguish TPO counts from contract volume, and preserve the same session definition when comparing value-area migration.

Value moving upward after a sequence of lower areas can support a reversal hypothesis. It still does not guarantee continuation. A developing session profile also differs from a completed profile: an end-of-session value area cannot be treated as known during the morning.

AVWAP: The Anchor and the Source Are Part of the Rule

An anchored volume-weighted average can be written as sum(price source × volume) / sum(volume), starting at the chosen anchor. The presenter uses lows when anchoring support at a swing low and highs when anchoring resistance at a swing high.

Our calculation uses low prices 100, 102, 101 and volumes 10, 20, 10. The low-source anchored average is (100×10 + 102×20 + 101×10)/40 = 101.25. If the corresponding highs are 104, 105, 103, the high-source result is 104.25. Different sources deliberately create different lines.

The low-source line is a volume-weighted average of bar lows, not the average execution price of all buyers. Likewise, a swing selected with future bars must be recognized only after confirmation. A historical line drawn back to the anchor does not authorize a trade before the anchor was identifiable.

Aggression: Why Delta Is Not the Number of Buyers Minus Sellers

Every matched trade has both a buyer and a seller. Delta distinguishes the side initiating the transaction, usually through a bid/ask or other trade-classification method. It does not count people. CVD accumulates those classified differences over a chosen period.

Stacked imbalance adds price-level detail. For our illustration, ask-side executions of 90, 120, and 150 contracts face comparison quantities of 30, 40, and 50. Each ratio is 3:1. A researcher might require three consecutive levels plus a minimum quantity. That threshold is ours, and diagonal versus same-price comparisons can produce different results.

In a CVD divergence example, price makes a higher high while cumulative delta makes a lower high. The presenter interprets this as buying exhaustion. Other explanations include changing passive supply, liquidity conditions, or the classification method. Delta disagreement is evidence about the measured flow; it is not proof that the total volume is too small to “justify” the price.

Ordinary OHLCV bars cannot reconstruct exact bid/ask executions at every price. Some platforms estimate delta from lower-timeframe price direction or provide specialized footprint data. Label the data source and its limits rather than presenting an up-candle-volume approximation as measured aggression. Weekly reset rules also matter: comparisons across resets can create misleading divergences.

Intermarket Divergence and Sweeps: Observable Disagreement, Inferred Intent

The presenter compares related markets at synchronized extremes. In our example, index A moves from 5,000 to 5,010 while index B moves from 20,000 to 19,980. One confirms a higher high and the other does not. That can motivate investigation, but shared drivers do not force identical performance. Constituents, sector exposure, session timing, and changing correlation matter.

A liquidity grab describes a price probe beyond a known swing followed by rejection. Define “quick” in bars or seconds and “rejection” with an explicit close or displacement rule. A chart can establish those events. It cannot establish who intentionally triggered stops or why.

The source distinguishes stop-related liquidity around swings from resting limit liquidity in the book. Keep that distinction, while avoiding an overstatement about stop orders: a stop-market becomes a market order after triggering, whereas a stop-limit becomes a limit order and can remain unfilled. The exact behavior depends on the order type and venue.

Liquidity inducement adds a proposed causal story to an apparent break and reversal. It can be a useful hypothesis, but labeling every failed breakout “manipulation” makes the theory impossible to challenge. Record the observable failure first; treat intention as a separate evidentiary question.

GEX: Position Sign Matters More Than the Call/Put Label

The presenter introduces concentrated options exposure as a possible source of support and resistance. Gamma measures how delta changes as the underlying price changes. The Options Industry Council’s gamma explanation distinguishes positive gamma for long calls and long puts from negative gamma for short calls and short puts.

Consequently, calls are not inherently resistance and puts are not inherently support. In a simplified delta-neutral hedging model, positive-gamma positions can lead to selling underlying into rises and buying into declines; negative-gamma positions can require the opposite. This is our explanation of the hedge mechanism, not proof of actual dealer positioning.

Options open interest alone does not identify which contracts dealers own or are short. GEX estimates need position-sign assumptions, expirations, underlying prices, and valuation inputs. Intraday flows and changing volatility can alter the picture. The transcript supplies no options-chain dataset or reproducible exposure calculation, so this guide does not manufacture gamma walls from price candles.

What the Source Establishes

The video offers a selection of concepts and narrated examples across markets and timeframes. It does not provide complete trade records, standardized entry and exit rules for all 12 ideas, controlled comparisons, or independently reproducible profitability statistics. Charts mentioned in the transcript were not supplied as visual evidence here.

Claims about reliability, institutional popularity, exclusivity, and historical origins remain the presenter’s claims unless separately established. Combining several labels does not guarantee independent confirmation. For example, a structure break, gap, and sweep can all arise from the same move.

Executable Research: Do Swing-Overlap Zones Produce Useful Rejections?

//@version=6
strategy("CandleChasers: implicit fair value demand-zone experiment", overlay=true,
     initial_capital=100000, pyramiding=0,
     commission_type=strategy.commission.percent, commission_value=0.01,
     slippage=1, backtest_fill_limits_assumption=1,
     calc_on_every_tick=false, calc_on_order_fills=false,
     process_orders_on_close=false, use_bar_magnifier=true,
     margin_long=100, margin_short=100)

// Source concept: intersection of three candle ranges at a swing low.
// Experimental additions: first future touch, close confirmation, ATR bracket.
int maxAge = input.int(100, "Maximum zone age (bars)", minval=1)
int minWidthTicks = input.int(1, "Minimum overlap width (ticks)", minval=1)
int breakBufferTicks = input.int(0, "Close invalidation buffer below zone (ticks)", minval=0)
bool requireBullish = input.bool(true, "Require bullish close on rejection")
int atrLength = input.int(14, "ATR length", minval=2)
float stopMultiple = input.float(1.0, "Stop distance in signal-time ATR", minval=0.1, step=0.1)
float rewardMultiple = input.float(2.0, "Target multiple of rounded stop distance", minval=0.1, step=0.1)
int maxHold = input.int(20, "Maximum completed holding bars", minval=1)
int units = input.int(1, "Fixed shares/contracts/units", minval=1)

float atr = ta.atr(atrLength)
bool swingLow = bar_index >= 2 and low[1] < low[2] and low[1] < low
float overlapLow = math.max(low, math.max(low[1], low[2]))
float overlapHigh = math.min(high, math.min(high[1], high[2]))
bool newZone = swingLow and overlapHigh - overlapLow >= minWidthTicks * syminfo.mintick

var float zoneLow = na
var float zoneHigh = na
var int createdBar = na
var bool used = false
var int riskTicks = na
var int targetTicks = na
var int firstTouches = 0
var int submissions = 0
bool touch = false
bool signal = false

if barstate.isconfirmed
    // Each newly confirmed zone replaces the older one, even if untested.
    if newZone
        zoneLow := overlapLow
        zoneHigh := overlapHigh
        createdBar := bar_index
        used := false

    if not na(createdBar) and bar_index > createdBar
        bool expired = bar_index - createdBar > maxAge
        bool broken = close < zoneLow - breakBufferTicks * syminfo.mintick
        if expired or broken
            zoneLow := na
            zoneHigh := na
            createdBar := na
        else if not used and low <= zoneHigh and high >= zoneLow
            // Formation candles cannot count as future touches.
            // First contact consumes the zone, even if already in a trade.
            touch := true
            firstTouches += 1
            used := true
            bool rejection = close > zoneHigh and (not requireBullish or close > open)
            if rejection and strategy.position_size == 0 and not na(atr)
                riskTicks := int(math.max(1, math.round(atr * stopMultiple / syminfo.mintick)))
                targetTicks := int(math.max(1, math.round(riskTicks * rewardMultiple)))
                strategy.entry("Long", strategy.long, qty=units)
                // Submit bracket with entry; relative ticks anchor to actual fill.
                strategy.exit("Bracket", from_entry="Long", loss=riskTicks, profit=targetTicks)
                signal := true
                submissions += 1

if strategy.position_size > 0
    strategy.exit("Bracket", from_entry="Long", loss=riskTicks, profit=targetTicks)
    int entryBar = strategy.opentrades.entry_bar_index(0)
    if bar_index - entryBar + 1 >= maxHold
        // Protective bracket remains live until market close can fill.
        strategy.close("Long", comment="Holding timeout")

plot(zoneLow, "Newest overlap zone low", color=color.teal, style=plot.style_linebr)
plot(zoneHigh, "Newest overlap zone high", color=color.teal, style=plot.style_linebr)
plotshape(newZone, title="Zone becomes known", style=shape.square, location=location.belowbar, color=color.blue, size=size.tiny)
plotshape(touch, title="First unbroken contact", style=shape.circle, location=location.belowbar, color=color.orange, size=size.tiny)
plotshape(signal, title="Entry submission, not fill", style=shape.triangleup, location=location.belowbar, color=color.lime, size=size.tiny)
var table counts = table.new(position.top_right, 2, 2)
if barstate.islast
    table.cell(counts, 0, 0, "First unbroken contacts")
    table.cell(counts, 1, 0, str.tostring(firstTouches))
    table.cell(counts, 0, 1, "Entry orders submitted")
    table.cell(counts, 1, 1, str.tostring(submissions))

Our primary experiment tests first future contacts with implicit demand zones. It asks whether a bullish rejection close is a useful filter, compared with requiring a close above the zone without the additional bullish-body requirement. Net results, not the attractiveness of individual examples, determine whether the filter earns its place.

This code tests one objective interpretation of the source’s concept. Any mechanical definitions introduced here are ours and are not necessarily the creator’s exact rules.

What This Code Actually Tests

The third candle’s close confirms the swing and intersection. The script starts looking for contacts only on later candles. A completed close below the zone, adjusted by the optional invalidation buffer, invalidates it. Otherwise, its first overlapping candle consumes the setup even if confirmation fails or another trade is open.

A qualifying close submits a market entry and its associated bracket. The rounded ATR-based stop distance and target distance attach to the actual simulated fill through relative ticks. They remain fixed throughout the trade. New zones cannot change an open trade’s risk distances. The TradingView strategy examples (paid link) explain these relative exit parameters.

What This Code Does Not Test

It does not reproduce all 12 concepts, identify institutional intent, estimate economic fair value, or use a structural stop under the swing low. It excludes supply-zone shorts, later retests, overlapping zone portfolios, limit-entry queue behavior, footprint imbalances, CVD, profiles, AVWAP, correlated markets, and options exposure. These omissions bound the research question rather than invalidate every possible use of the ideas.

Execution Assumptions Worth Inspecting

Signals use completed candles. Market entries and timeout exits ordinarily fill on the next available simulated tick, often the next bar’s open. A price gap can change the entry or exceed the stop. The bracket is submitted with entry rather than waiting another completed bar. Tick rounding can make the target slightly different from an exact 2R multiple.

Bar Magnifier is requested where data is available. A bar hitting both exits remains sensitive to intrabar sequencing, and target touches do not prove real fills. The model requires one tick of target penetration. Review ambiguous trades rather than assuming the favorable order of events.

Default assumptions are 0.01% commission per side, one tick of slippage on applicable fills, $100,000 initial capital, and 100% emulator margin. They are illustrative research settings, not a broker quote. Change Strategy Properties for the selected instrument. Fixed units give variable dollar risk as ATR changes; spread, financing, partial fills, and all venue-specific charges are not separately modeled. Positions can span sessions.

A new zone replaces an older untested zone and takes priority if formation coincides with an older zone’s contact. Age is counted from confirmation. Formation candles do not count as retests. Contacts closing below invalidation are excluded from the first-unbroken-contact counter; it is not a complete count of all failed zones. Record replacements, invalidations, and expired zones separately when assessing the full opportunity set.

Validation status: This script has not been compiled in TradingView or run against historical data here. No performance results are claimed. Static review checks causal signal timing, fixed brackets, setup consumption, and HTML code preservation.

Backtest Lab

Begin with standard candles on one declared instrument, timeframe, and date range. Compile the script and inspect individual formation, contact, submission, and fill events. Then change one variable at a time:

Keep market, timeframe, period, unit size, and other rules constant. Record trade count, mean net trade, average winner and loser, win rate, profit factor, maximum drawdown, holding time, and ambiguous fills. For filter comparisons, include submitted trades relative to observed contacts, while recording rejected and invalidated zones in a separate ledger.

Develop rules on one period and evaluate them on another. Similar neighboring settings are more persuasive than one isolated historical peak. The newest-zone policy changes which contacts enter the sample; a multiple-zone implementation would be a separate experiment.

Try It Yourself: Convert a Label into a Falsifiable Rule

For ten three-candle swing lows, calculate the intersection by hand while hiding future bars. Record the time the third candle completes, both zone boundaries, whether the zone was later replaced or invalidated, first future contact, confirmation, and the resulting simulated fill. Include empty intersections and failures.

Then choose one additional concept from the table. Write its data source, exact definition, reset or session convention, entry, invalidation, cost assumptions, and a result that would weaken your confidence. For CVD or GEX, identifying missing data is a valid research finding; inventing substitute “smart money” values is not.

Expectancy exercise: 45% of trades win 2R, 55% lose 1R, and average friction is 0.10R. Net expectancy = 0.45×2 − 0.55×1 − 0.10 = 0.25R. If partial exits reduce the average winner to 1.4R, expectancy becomes −0.02R. Nominal targets alone do not establish actual trade economics.

Check Your Understanding

1. When does the three-candle overlap become knowable? After the third candle completes.
2. Does an objective zone guarantee an edge? No. It makes the hypothesis reproducible, not profitable.
3. Is a 40% value area 40% of the price range? No. For volume profiles it refers to a specified share of profile volume.
4. Does delta count buyers minus sellers? No. It classifies initiating-side traded volume.
5. Does a rejected swing break prove manipulation? No. The price event is observable; participant intent needs separate evidence.
6. Are calls always negative gamma and puts positive gamma? No. Long versus short position determines the gamma sign.
7. Why might a second structure shift hurt trade economics? Confirmation can arrive farther from a fixed stop and closer to the target.

Final Takeaway

The source encourages looking beyond one familiar chart pattern. The deeper lesson is to separate price geometry, participation data, and inferred mechanisms. Twelve compelling examples do not establish twelve profitable strategies. Start with one concept that has clear data and causal rules, test it under realistic costs, and add another only when it supplies information the baseline does not already contain.

Risk Disclosure: This material is for educational and research purposes only and is not financial or investment advice. Trading can result in substantial losses. Backtests are hypothetical and depend on assumptions. Fees, spreads, slippage, gaps, liquidity, data quality, order behavior, and changing market conditions can materially affect actual results.

This article is an independent educational companion to the video and is not affiliated with, sponsored by, endorsed by, or authorized by Fractal Flow – Pro Trading Strategies.