Cluster Analysis in Trading: How to Read Footprint Charts [2026]

Cluster analysis in trading isn't just another indicator stacked on top of a chart. It's an X-ray of the candle: a tool that reveals what's happening inside every price move. While a regular trader sees a green or red candle, someone using cluster analysis sees the exact balance of forces — how much was bought, how much was sold, and at which precise level the real volume was concentrated. That difference is what separates good entries from mediocre ones.

Professional scalping is built on three data sources: the order book (the market's future), the tape (the present), and clusters (the past). Clusters accumulate the history of the battle for every price level — and they're the ones that answer where price is likely to go from any given zone. In this article, we'll break down cluster analysis in trading from the ground up to practical application.

What Is Cluster Analysis in Trading

Cluster analysis is a method for studying how trading volume is distributed across price levels within each individual candle. Instead of a single aggregated volume number below the candle, the trader gets a detailed map: at exactly which prices trades occurred, who initiated them (buyers or sellers), and what the final aggression balance was.

The tool is called a cluster (also "footprint" — literally a trace or imprint of activity). Each candle becomes a table where the rows are price ticks and the columns are volume on the Bid and Ask side.

If you're just starting to understand how the market works from the inside, it's worth checking out Secret Terminal's free YouTube course alongside this material. "Secret Terminal interface | order book, clusters, tape and workspace" is a good starting point — it shows how all three tools of the triad look in practice before diving into the theory of each. The lesson is part of a full free course on crypto trading and scalping.

Clusters vs Regular Candles: What's the Difference

A standard Japanese candle tells you four things: open, high, low, close. How price got from open to close — nothing. Was the move aggressive or sluggish? Who dominated at key levels? Was the large limit order actually absorbed, or pulled before price reached it?

A regular candle answers none of those questions.

A cluster answers all of them. Inside a single candle you can see:

• At which level the maximum volume is concentrated.

• How the buyer/seller balance shifted from tick to tick.

• Where price "got stuck" — because a real limit wall was sitting there.

• Where volume is anomalously high compared to average values.

That's why experienced scalpers call clusters an X-ray of the candle: the external appearance can deceive, but the internal content cannot.

What Clusters Show (Delta, Volume, Bid/Ask)

Three key metrics in cluster analysis:

Volume — the total number of contracts or coins that passed through a specific price level during the candle's lifetime. The higher the volume at a level, the stronger it is as a zone of interest.

Bid/Ask breakdown — each trade falls into one of two types:

• Buy at Ask (aggressive buyer hits the best ask).

• Sell at Bid (aggressive seller hits the best bid).

The breakdown shows who was the initiator: a market order is always aggressive, a limit order is passive. If at the $42,000 level, 500 BTC went through on the Ask side and only 80 BTC on the Bid side, that's a clear signal of buyer dominance at that exact point.

Delta — the difference between buy volume (Ask) and sell volume (Bid):

Delta = Ask Volume − Bid Volume

Positive delta: buyers were more aggressive. Negative: sellers. The cumulative candle delta is the final balance for the entire period. Delta analysis is precisely what lets you determine whether a large limit order in the order book was genuinely absorbed or simply pulled (spoofing).

Footprint Chart: Display Modes

There are several footprint display modes, each with its own use case:

Bid×Ask (classic footprint) — the most common format. Each row shows cells as Bid_volume × Ask_volume. Lets you instantly see the imbalance at every tick.

Delta footprint — each cell shows only the delta (positive = green, negative = red). Useful for quick visual assessment of who's dominating.

Volume profile — total volume distribution across price levels shown as horizontal bars. Doesn't show bid/ask, but clearly marks the Point of Control (POC) — the level with the highest volume.

Imbalance — a special markup that highlights cells where volume on one side is 3x or more the volume of the opposing cell in the adjacent row. These zones frequently act as magnets that pull price back.

How to Read Clusters in Trading

Clusters in trading aren't read like traditional technical analysis — not "pattern on pattern," but as a cross-section of real transactions. The logic: large players can't hide their activity — they leave a trail in the volume.

Large Buys and Sells

In the context of cluster analysis, a "large buy" isn't buying at market price upward (anyone can do that) — it's accumulating a position through limit orders on the Bid side under pressure from market sellers.

Classic accumulation scenario in the cluster:

• Price drops to a level.

• At that level, Bid volume spikes sharply — someone is "standing" and absorbing sellers.

• The candle delta is negative (sellers are active), but price isn't going lower.

• This is a sign that a large buyer is working at that level.

The same logic applies to sells: price climbs to a level, Ask volume spikes anomalously, and price won't go higher. A seller is standing and absorbing buyers.

These zones — where large capital makes its decisions — become strong support and resistance levels going forward. A trader reading clusters gets those levels from primary data, not from Fibonacci levels or subjective markup.

Delta: The Balance Between Buyers and Sellers

Delta is the main indicator for understanding cluster analysis. But you don't read it straight — you read it through the lens of divergences.

Bullish delta divergence: price is falling (making a new local low), but the delta is positive or less negative than at the previous low. Signal: despite price moving down, buyers are becoming more aggressive. Likely reversal upward.

Bearish delta divergence: price is rising, making a new high, but the delta is negative. This means sellers are dominating on the way up — someone is aggressively selling into strength. Likely reversal downward.

Absorption: delta and price are moving in the same direction, but price "freezes" at a level. Limit orders are absorbing the aggression — the level is being held by real capital.

An important nuance: delta doesn't predict moves directly. It shows who was the aggressor at the moment of trading. A large buyer may be working passively through limit orders — and the delta will be negative even though buying is actually happening. That's exactly why cluster analysis always needs to be paired with the order book.

Anomalous Volume: What It Means

Anomalous volume in a cluster is a level where volume exceeds the average by 3–5x or more. There are typically 2–4 such anomalies per session, and each carries information.

Three types of anomalies and their meanings:

High Volume Node (HVN) — a zone of maximum volume. This is where an intense battle between buyers and sellers played out. Price tends to return to these zones (as a "fair value" area). A solid base for bounce trading.

Low Volume Node (LVN) — a zone of minimum volume (an "empty" zone). When price enters an LVN, it moves through it fast — no buyers, no sellers. The goal: find the LVN between the current price and the next HVN — that's the space price will slice through without resistance.

Imbalance — a situation where Ask volume at one level is 3x or more the Bid volume at the adjacent level. According to the Fair Value Gap concept, the market tends to return to these zones to "close the imbalance." In practice: an imbalance in price's path is a magnet that price will likely reach. For more on imbalance as a standalone tool — Imbalance in Trading

Cluster Analysis in Practice

Clusters deliver maximum value when combined with other microstructure tools. In isolation, a cluster is only half the picture.

Clusters + Order Book

The "cluster + order book" combination solves the main problem in liquidity analysis: distinguishing real orders from manipulation (spoofing). For a full breakdown of how to read the order book on its own, see the dedicated article: Order Book

Scenario: a $2M limit buy order is sitting in the order book. Price approaches. Three possible outcomes:

• The order holds and price bounces upward — the cluster confirms Bid-side absorption. This is a real wall.

• The order is pulled before price touches it — the cluster shows zero volume at the level. This is a spoofer; you can't trade from that level.

• The order holds but price rips right through it — the cluster shows the order was "eaten" by a massive market order. The trend is strong; a continuing impulse is possible.

The cluster gives you the answer: was the density level in the order book genuinely executed, or just pulled? Without that understanding, trading from order book levels is a lottery.

Practical workflow:

• Find a large density level in the order book that's been sitting for 30+ minutes.

• Wait for price to approach.

• Watch the cluster: if Bid volume starts growing at the level (absorption) — go long ahead of the density. If Ask volume is explosive and the density is melting away — prepare for a breakout.

Clusters + Tape

The tape (time & sales) shows the market's present — the stream of executed market orders in real time. The cluster shows the past — accumulated volume. Together, they give you the full picture. The technique of reading the tape is covered in detail in: Tape Reading

The tape accelerates with large prints → the cluster starts "filling up" at specific levels → you can see exactly where real money is going right now. If the tape is flying toward a level break and the cluster confirms a growing imbalance — the breakout entry becomes high-probability.

An important sign of a weak move: the tape is active, prints are large, but the cluster shows absorption (high volume, no price movement). That means a real seller is sitting at the level and absorbing the aggression. The move is about to stop.

Clusters for Finding Levels

Cluster analysis is one of the most precise methods for identifying significant price levels. The process:

• Find High Volume Nodes in the history — zones where total volume is anomalously high. These are levels where large capital clashed at scale.

• Mark the Point of Control (POC) for the session, the week, the month — the level with the highest volume for the period.

• Find LVNs (empty zones) — the space between HVNs through which price moves fast and without resistance.

• Overlay these zones with key chart levels and order book density levels.

A level that's simultaneously an HVN on the cluster + a large limit order in the order book + a psychologically round number — that's your strongest support level. You can trade from it with a minimal stop.

Cluster Entry Strategies

Entry from Anomalous Volume

The core idea: find a zone where anomalous volume occurred and enter in the direction that volume is "defending."

Conditions for a long:

• Volume passed at 4–6x the average at this level.

• Delta was negative (price was dropping), but price didn't break the level.

• This is a sign of absorption — someone was sitting with a limit order and bought everything sellers threw at them.

Entry: on price's return to that level, with confirmation from the tape (sell-side print slowdown).

Stop: beyond the HVN level with a 0.1–0.2% buffer (outside the absorption zone).

Exit: at the next HVN or on a delta shift.

Entry on Delta Reversal

This strategy is based on price/delta divergence.

Short algorithm:

• Price makes a new high (HL/HH structure continues).

• Candle delta at the new high is negative — sellers are dominating.

• Volume at the level is above average (activity confirmation).

• The order book shows a large density level on the Ask side (above price).

Enter short on the pullback to the previous HVN. Stop above the high. Target: nearest HVN below or the LVN between levels.

Long algorithm — mirror image: price makes a new low, delta is positive, volume is high, order book shows large density on the Bid side.

Trade Examples (Case Studies)

Case 1: Absorption at a Support Level (ETH, session)

Context: ETH is ranging; price is approaching the $2,480 level for the third time.

Cluster at $2,480: Ask volume — 420 coins, Bid volume — 2,100 coins. Delta: −1,680. Price isn't dropping.

Interpretation: despite aggressive selling (high Ask), a large buyer is absorbing everything through Bid limit orders. Real support.

Entry: long at $2,482 (ahead of the density level). Stop: $2,474. Take-profit: $2,510 (next HVN).

Result: price moves from $2,480 to $2,515. Risk-to-reward ratio — 1:4.

Case 2: Delta Divergence at a High (BTC)

Context: BTC is rallying from $64,000 to $66,800. The last two candles on the 5-minute chart are making new highs.

Cluster of the high candle: Bid volume (sells) — 580 BTC, Ask volume (buys) — 210 BTC. Delta: −370 BTC. Price went up, but sellers are more aggressive than buyers.

Order book: at the $67,100 level there's a $4.2M density level — the Density Map filter shows it's been sitting there for 40 minutes.

Entry: short at $66,750. Stop: $67,200 (beyond the density). Take-profit: $65,800 (session HVN).

Result: BTC pulls back to $65,900. Risk-to-reward ratio — 1:1.9.

Case 3: Breakout with Confirmation (ALT)

Context: an altcoin is trading at a horizontal resistance level. Price has tapped it three times — a cascade.

Cluster of the last candle before the breakout: Ask volume spikes sharply, delta turns strongly positive (+12,000 contracts). The tape accelerates with large green prints.

Order book: the density level at $0.382 (resistance) starts "melting" — volume is decreasing, meaning it's not holding.

Entry: long the moment Ask volume in the cluster exceeds the size of the density level. Stop: below $0.379. Take-profit: next order book level at $0.415.

Result: the coin impulses to $0.418. Trade closed at $0.413. Risk-to-reward ratio — 1:5.2.

Where to View Clusters

ATAS, Bookmap, and Other Platforms

Proper cluster analysis requires specialized software — standard exchange interfaces don't provide footprint data.

ATAS is considered the de-facto standard for professional footprint analysis in futures markets. Bookmap is strong on order book visualization through its Heatmap, but the footprint functionality is less detailed. Sierra Chart requires technical configuration but is highly flexible.

For crypto markets, there's a distinct class of tools: terminals with native exchange API integration. Secret Terminal falls into this category — it was built specifically for the characteristics of crypto market microstructure, not adapted from futures markets.

Cluster Analysis in Secret Terminal

Secret Terminal implements the analysis "triad" as a unified tool: the order book (Orderbook), tape (Tape), and clusters (Footprint) all operate in synchronized mode. This is a fundamental distinction from most platforms, where this data is spread across separate windows.

Key cluster analysis capabilities in the terminal:

Delta analysis — real-time visualization of the balance of forces. Lets you track delta changes candle by candle and see the accumulated cumulative delta.

Order book verification through the cluster — when price approaches a large limit order, the cluster immediately shows whether it's being absorbed (Bid volume rising) or sitting without reaction (spoofing). This is the key tool for trading from order book density levels.

Cluster transparency toggle — a function that switches the display from visual (color map) to numerical (exact volume values at each level). Used for verifying support and resistance zones against specific numbers.

Synchronization via Links — when working with multiple instruments simultaneously, the cluster window synchronizes with the order book and chart through the Links mechanism: changing the ticker in one window switches all connected modules.

Data update frequency — up to 20–80 ms for the tape and 100 ms for the order book. This is critical for scalping: slower updates mean a "lagging" cluster, which reduces entry precision.

Common Mistakes When Working with Clusters

Cluster analysis in trading is a powerful tool, but its very depth generates a predictable set of mistakes in traders just starting to work with volume.

Mistake 1: Trading Only on Delta

Delta is a derived metric, not a primary one. Looking only at delta and ignoring absolute volume values is like reading wind direction without knowing wind speed. Delta +500 on 600 units of volume is completely different from delta +500 on 50,000 units of volume. In the first case, buyers are barely more aggressive than sellers; in the second, the dominance is overwhelming.

The rule: always evaluate delta in the context of absolute volume. Only anomalously high volume with a skewed delta constitutes a tradable signal.

Mistake 2: Ignoring the Time Context

Cluster analysis works differently at different times of day. During low-liquidity periods (the Asian session for most assets), anomalous volume may be the result of a single large algorithmic order, not a genuine reflection of market demand. During the US session (from 15:30), the same numbers mean something entirely different — they represent full institutional flow.

Practical rule: cluster patterns formed during high-liquidity sessions (Europe + US) have significantly higher predictive value.

Mistake 3: Failing to Distinguish Absorption from a Weak Market

At first glance, "absorption" and a "weak market" look the same: high volume, price not moving. The difference is in who is doing the absorbing.

Seller absorption (bullish): price drops to a level, Bid volume spikes sharply, delta is negative, but price has stopped. A buyer is sitting with a limit order and buying everything.

Weak bull market: price moves up slightly, Ask volume is high, delta is positive, but price barely moves. Buyers are aggressive, but they're meeting powerful supply from above.

The distinction is critical: the first scenario is a reversal; the second is continued downward pressure.

Mistake 4: Trading Illiquid Instruments — When Cluster Analysis Doesn't Work

Cluster analysis requires sufficient trade flow to form statistically significant patterns. On instruments with daily volume below $10–20M, a single large order can paint "anomalous volume" that's actually just one whale's trade — not market structure.

Specific cases where clusters won't give a reliable signal: altcoins with a 24-hour volume below $20M; BTC on Sunday nights from 00:00 to 06:00 UTC (volume drops 3–5x from average); any asset 2–3 minutes before and after major macro data releases (Non-Farm Payrolls, Fed rate decisions) — in those moments, a single algorithm can generate all the visible "volume."

Minimum threshold for crypto volume analysis: daily trading volume from $50–100M, trade count from 500,000.

Mistake 5: Over-Complicating the Analysis

Beginning traders often try to factor in too many cluster patterns at once. Delta + imbalance + HVN/LVN + session cumulative delta + POC + ... The result is analysis paralysis exactly when you need to make a decision in seconds.

The working approach: for scalping, pick 1–2 core signals and only execute on those. For example, just "absorption at an order book level" + "delta divergence." Add new patterns only after the core ones are second nature.

Most of these mistakes are easier to avoid when you can see from the start how professionals apply the tools in real conditions. In Secret Terminal's free YouTube course, "How professionals read the market | market analysis through order book and clusters" shows exactly that — not pattern theory, but the live decision-making logic behind volume data. The lesson is part of the full free course.

Market Psychology Through Clusters: How the Big Player Leaves Traces

Behind every cluster pattern is specific market mechanics — the actions of participants with real capital. Understanding this psychology transforms technical analysis into an understanding of market logic.

The Market Maker and Its Inventory

A market maker constantly holds positions on both sides — selling into strength and buying on weakness, earning on the spread. When its inventory gets skewed (accumulated too many longs or shorts), it needs to rebalance. This creates distinctive cluster patterns: a sharp volume spike with neutral delta — the market maker is "unloading" a position, using a moment of activity to do so.

Visually in the cluster: a candle with anomalously high total volume, but delta close to zero (±5–10% of volume). Price stays roughly at the same level. This isn't an entry signal — it's "technical" volume, neutral in terms of direction.

Institutional Accumulation: Patience vs. Impulse

A large institutional buyer can't purchase the needed volume in a single transaction — that would immediately move price against them. So they accumulate slowly, absorbing market sales through Bid-side limit orders.

How it looks in clusters:

• A series of candles with neutral or slightly negative delta.

• Price barely moves — a range.

• At one specific price level (a POC zone), consistently high Bid volume keeps printing.

• The order book shows stable density at that same level.

When the institution has built its full position — it removes the limit order from the book (the density disappears) and the move begins. The trader who was tracking the cluster sees that the level has "exhausted" — and enters in the direction of accumulation.

Spoofing and How the Cluster Exposes It

Spoofing is placing a large limit order with no intention of filling it — purely to create the illusion of a level. The spoofer manipulates other participants' perception: seeing a large wall in the order book, traders expect price to bounce off it. The spoofer pulls the order as price approaches and waits for a move in the other direction.

The cluster exposes spoofing immediately: if the large density level in the order book was "real," as price approaches, the cluster should show a sharp surge in Bid volume (absorption). If volume doesn't grow — the density was pulled before the touch. That's a spoofer. You can't trade from that level — it doesn't exist.

This cluster-based verification is the key competitive advantage of a trader working with volume analysis.

Large Stop Hunts and Traps

Another classic pattern is the stop hunt. Price sharply breaks an obvious technical level (say, a previous high), wipes out participants' stops, then reverses. In the cluster, a stop hunt looks like this: a sharp Ask volume spike (upside breakout), immediately followed by high Bid volume (return). The delta on the breakout candle is positive, but on the next candle it's already sharply negative. Price returns below the level. The "short after a false breakout" scenario is one of the technically cleanest setups out there, because the cluster captures the reversal point with exhaustive precision.

Clusters for Scalping

Volume analysis in crypto through clusters is especially powerful in scalping — trading on short timeframes (1–5 minutes). The reason: on higher timeframes, volume anomalies get "smeared," whereas on 1–5-minute candles every large order is clearly visible. The principles behind building scalping strategies are covered in: Scalping Strategies

The scalper's workflow:

• Context (30% of analysis): on the 15–60-minute chart, identify the key level — the prior session's HVN, a round number, or an order book level with density above $500K–$1M.

• Wait for price to approach: when price nears the level on the 1–5-minute timeframe, the trader goes into full readiness mode.

• Read the current candle's cluster: watch volume develop in real time. Is absorption growing? What is delta saying?

• Tape confirmation: the tape should either slow down (bounce signal) or accelerate with large prints in the breakout direction (signal to enter in the direction of the move).

• Entry with minimal stop: the stop is placed behind a specific price level in the cluster — beyond the HVN zone or the absorption level.

Two cluster scenarios are especially important for scalping:

"Consolidation" scenario: price chops around on a level for several candles, volume accumulates, delta fluctuates near zero. This is accumulation before a move. Entry: on the breakout of the last consolidation candle in the direction of the delta imbalance.

"Volume spike" scenario: a single candle prints a sharp anomalous volume spike (5–10x the average) with negative delta, but price stays at the level. This is a climactic sell — "the last bear" is exiting, further downward pressure is unlikely. Enter long after that candle closes.

Cluster-based scalping requires strict asset selection criteria: 24-hour trading volume over $100M, trade count over 800K. On illiquid instruments, cluster analysis doesn't work — there isn't enough transaction flow to form meaningful patterns.

Position Management Through Clusters

Cluster analysis in trading applies not just to entries but to managing open positions:

• Stop shift: if after entering a long the cluster shows a new HVN above the entry price — the stop loss can safely be pulled up to breakeven. The new HVN is now "protecting" the position.

• Partial exit: when price approaches an LVN zone (an empty zone in the cluster), you can take partial profits — further movement is less predictable in an empty zone.

• Full exit: growing Ask volume with negative delta moving against your position — signal for a full exit. Someone large is starting to work against the current trend.

Recommended risk-to-reward ratio for scalping: minimum 1:1.5. With properly selected entries from absorption and delta divergences, real trades deliver 1:2–1:5, which ensures a positive expected value even with a 40–50% win rate.

FAQ

• What is cluster analysis in trading, in plain terms?

  It's a method for studying what happens inside a candle — how many trades occurred at each price tick, who was the aggressor (buyers or sellers), and where the maximum volume is concentrated. If a regular candle is a book cover, a cluster is its contents. A trader reading clusters works with the market's primary data, not a secondary processed version of it.

• How is a footprint chart different from standard volume analysis?

  Standard volume (the bars below the chart) shows total activity for the period — no price-level breakdown, no Bid/Ask split. Footprint shows volume at every price level, split into buys and sells. This is incomparably more detailed: you can see not just "how much," but "who" and "exactly where.

• What is delta and how do you use it?

  Delta = buy volume on Ask minus sell volume on Bid. Positive delta means buyers were more aggressive. The main way to use it: look for divergences. Price rallies to a new high, but delta is negative. That's a signal the move is weakening — buyers are running dry while sellers are getting more active. But always check delta against absolute volume.

• Can you use clusters without the order book?

  In theory — yes. In practice — no. The cluster shows the past (what already happened); the order book shows the future (players' intentions). Without the order book, there's no way to know whether a real limit order is "holding" a level or not. The "cluster + order book" combination is the minimum toolkit for crypto volume analysis.

• What timeframes does cluster analysis work on?

  All of them, but best on 1–15-minute timeframes for scalping. On higher timeframes, clusters are great for identifying large support/resistance zones (HVNs) and mapping weekly POC levels. For precise entries, a scalper needs the granularity of lower timeframes — every large order is visible in real time there.

• What is imbalance in a cluster?

  Imbalance is a situation where Ask volume at one level is 3x or more the Bid volume at the adjacent level (or vice versa). By market logic, these "voids" are something the market seeks to "close" — return to that zone and redistribute volume. An imbalance is a magnet, and it works especially well on liquid assets with high trade flow.

• Do you need special software for cluster analysis?

  Yes. Standard exchange interfaces (Binance, Bybit, etc.) don't provide footprint data. You need specialized software: ATAS, Bookmap, Quantower, Sierra Chart, or Secret Terminal. For crypto markets, the optimal choice is a terminal with native exchange API integration and tape/order book synchronization in a unified interface — exactly how it's implemented in Secret Terminal.

Cluster Analysis in Real Time — Secret Terminal

Seeing the market "straight through" — via the order book, tape, and clusters in a single interface — is what separates a professional scalper from a trader working blind. Secret Terminal unifies all three tools of the triad with data update frequency up to 20 ms, a Density Map for filtering real order book density levels, and cluster delta analysis for level verification.

Try cluster analysis under real market conditions and see for yourself: the difference between a candle and a cluster is the difference between a symptom and a diagnosis.