Order Blocks

An order block is a specific way of marking a zone on the chart: the last opposing candle — or small cluster of candles — immediately before an impulsive move that breaks market structure. A bullish order block is the last down candle before an up-move that takes out a prior high; a bearish order block is the last up candle before a down-move that takes out a prior low. The premise behind the name is that large participants had to absorb opposing flow before price could leave the area with that much force, so the zone is treated as unfinished business rather than as an ordinary candle.

In practice the concept is a disciplined subset of supply and demand. What distinguishes it from generic zone drawing are the two conditions most versions require: displacement (the move away must be fast and one-sided, usually leaving an imbalance or fair value gap) and a structural event (a break of structure, BOS, or a change of character, CHOCH). Without those filters, almost any pullback candle qualifies and the concept stops being selective.

## How the mechanic works

The usual sequence runs: read trend and structure on a higher timeframe; wait for an impulsive leg that breaks a swing point; walk back to the last opposing candle before that leg and mark it; wait for price to return to the zone (mitigation); enter on the reaction with the stop beyond the block and the target at the next structural level or pool of liquidity. The surrounding vocabulary — mitigation, imbalance, liquidity sweep, displacement, premium and discount — comes from the same body of material, and most versions catalogued here use some part of it.

Two properties matter for anyone implementing this. An order block can only be identified after the impulse that confirms it, so there is always a lag between the candle that forms the zone and the moment a rule can see it. And the block is a region, not a price, so every version must state explicitly where the zone starts and ends.

## Main variants

Across the decoded versions linked from this page, the recurring families are:

- **Structure-gated blocks** — the block only counts if a BOS or CHOCH confirms it. The canonical smart-money-concepts form. - **Breaker and inversion blocks** — a block that failed and was traded through, then reused with its role flipped: former demand becomes supply. - **Imbalance-paired blocks** — a fair value gap or inefficiency must sit alongside the block, and entry is often placed inside the gap rather than at the block edge. - **Liquidity-first blocks** — a sweep of obvious highs or lows must occur before the block is considered valid; accumulation–manipulation–distribution and "power of three" framings belong here. - **Volume-qualified blocks** — volume at the origin candle, or agreement with a fixed-range volume profile or point of control, is used to filter zones. - **Confluence overlays** — Fibonacci retracement bands, moving-average ribbons or other trend filters, and top-down higher-to-lower timeframe workflows that decide which blocks are eligible. - **Automated detection versus manual marking** — indicator-driven versions plot blocks by rule; discretionary versions leave selection to the trader.

## What separates one implementation from another

The label hides a large amount of specification, and this is where versions genuinely diverge. Zone definition: body only, body-to-wick, or full high-to-low range. Which candle counts as "the last opposing one": the last opposite-coloured candle, the last down close, or a small consolidation treated as a single block. How displacement is measured: an ATR multiple, the presence of an imbalance, a minimum candle count, or a close beyond the swing. How structure is defined: swing or fractal lookback length, and whether a break needs a close or only a wick. Entry trigger: a limit at the proximal edge, the 50% mean threshold, or a lower-timeframe confirmation. Invalidation: a touch, a wick through, or a close through. Lifespan: how long a block stays valid, whether it can be used more than once, and how many are tracked at a time. Two versions described in identical words can behave very differently once these are fixed.

## Common mistakes

The most frequent problem is hindsight marking — choosing the block that already worked instead of applying a rule computable at bar close. A close second is dropping the displacement requirement, which floods the chart with candidates and turns the method into after-the-fact selection. Others: ignoring higher-timeframe context and taking blocks against dominant structure; treating mitigation as a reversal signal when a return to the zone is only a precondition; placing the stop inside the block so ordinary noise removes the trade, or drawing zones so wide that risk becomes unmeasurable; and reading the institutional narrative as evidence the zone will hold, when the narrative is an explanation rather than a test.

## Evaluating and backtesting a version

Start by writing the rules until they are mechanical: detection, validity window, entry, stop, target, invalidation, and how many blocks are tracked simultaneously. Respect the confirmation lag — the zone must be timestamped from the moment the impulse confirms, not from the origin candle, or the results are unreachable in live trading. Check whether an indicator repaints: if detection changes as new bars arrive, the backtest is not measuring the rule you would actually trade.

Then test out of sample, and on more than one instrument and session, since zone detection is scale-sensitive and often implicitly tuned to one market's volatility. Vary one parameter at a time and prefer broad plateaus over isolated peaks. Include spread, commission and slippage; limit entries at zone edges are optimistic in fast markets, and some fills simply will not happen when price gaps through the level. Finally, compare against a baseline using the same stop and target — a plain pullback to a moving average, or a zone drawn without the structural filter. If the order block version does not separate from that baseline, the behaviour is coming from the exit and risk model rather than from the block. For discretionary versions, turn the description into a checklist and grade adherence trade by trade before judging the method itself.

Strategies in this concept (67)

Frequently asked questions

What exactly is an order block?

It is the last opposing candle, or small group of candles, immediately before an impulsive move that breaks market structure — the last down candle before an up-move that takes out a prior high, or the last up candle before a down-move that takes out a prior low. The zone drawn from that candle is then watched for a reaction when price returns to it. The name reflects an interpretation about where large orders were absorbed; as a trading rule, what matters is the mechanical definition, not the interpretation.

How is an order block different from a supply and demand zone?

Order blocks are a stricter subset. A supply or demand zone can be any area where price previously turned or consolidated. An order block normally requires two extra conditions: displacement — the move away must be fast and one-sided, usually leaving an imbalance — and a structural event such as a break of structure or change of character. Remove those filters and the two concepts converge, which is why implementations that skip them tend to generate far more zones than they can filter.

What is a breaker block, and how does it relate?

A breaker block is an order block that failed. Price traded through it instead of reacting, and the zone is then reused with its role inverted: a former demand block is treated as supply, and vice versa. It is the same drawing rule applied after invalidation, and it is why an order-block system needs an explicit invalidation definition — whether a wick or a close through the zone kills it determines whether a level becomes a breaker or simply disappears.

Do order blocks work the same on every timeframe and market?

The drawing rule is scale-free, but the parameters usually are not. Displacement thresholds, swing lookback lengths and zone widths that make sense on a 15-minute forex chart often produce too many or too few blocks on daily equities or on crypto, where volatility and gap behaviour differ. Session structure also matters, since several variants depend on liquidity sweeps tied to specific trading hours. Treat any timeframe or instrument change as a re-test rather than a transfer.

Do order block indicators repaint?

Some do, and it is the first thing to check before backtesting one. An order block is only identifiable after the impulse that confirms it, so a correct implementation plots the zone with that lag. Implementations that redraw or remove blocks as new bars arrive show a cleaner chart in hindsight than the one you would have traded. If the indicator has no fixed confirmation point, any historical result it produces is unreliable regardless of how it looks.

How do you tell whether an order block version has any edge?

Make the rules mechanical, respect the confirmation lag, then test out of sample across more than one instrument with realistic costs. The decisive step is a baseline comparison: run the same entry management, stop and target on a simpler rule — a pullback to a moving average, or the same zone without the structural filter. If results do not separate, the behaviour you are measuring belongs to the risk and exit model, not to the order block itself.

All strategy concepts · Explore all strategies