An order block is the last opposing candle before an impulsive move; a fair value gap is a three-candle price imbalance the market skipped over. Both are zones traders expect price to return to. Both are also drawn by hand, and that is exactly where the trouble starts.
These two ideas sit at the centre of Smart Money Concepts, the price-action framework popularised by the trader known as ICT. This article defines each one plainly, shows how practitioners mark and trade them, and then does the part most tutorials skip: it names the subjectivity honestly and spells out what a testable version would require.
What an order block actually is
An order block is the last candle that closes against the direction of a strong move that follows it. A bullish order block is the last bearish (down) candle before an impulsive rally; a bearish order block is the last bullish (up) candle before an impulsive sell-off (Inner Circle Trader: ICT Order Block, FluxCharts: Order Blocks explained).
The zone is the candle's own range: for a bullish block you draw the box from the low to the high of that last down candle; for a bearish one, from the high to the low of that last up candle (FluxCharts). The story attached to it is that institutions filled orders inside that candle before the market ran, leaving "unfilled" interest behind. When price returns to the zone later, that is called mitigation — the market supposedly coming back to fill remaining orders, so the zone should act as support or resistance (TradingFinder: ICT Order Blocks).
That institutional story is unverifiable from a candle chart alone — you cannot see whose orders filled where. What you can observe is the mechanical pattern: an opposing candle, then a sharp move away from it.
What a fair value gap (imbalance) is
A fair value gap (FVG), also called an imbalance, is a three-candle pattern that marks a price range the market moved through so fast it left a gap between wicks (TrendSpider: Fair Value Gap strategy). In a bullish FVG, the middle candle is a strong up move, and the gap is the empty space between the high of the first candle and the low of the third candle — a zone the middle candle's body jumped over without trading through it (FluxCharts: Fair Value Gaps explained). A bearish FVG is the mirror image: price drops sharply and leaves a gap between the first candle's low and the third candle's high.
Bullish FVG (3 candles):
candle 1 high ──┐
│ <- the gap (imbalance)
candle 3 low ──┘
box top = candle 1 high, box bottom = candle 3 lowThe trading premise is that price often retraces to "rebalance" the gap before continuing. Practitioners wait for price to trade back into the zone — many aim for a reaction near the 50% fill — then enter in the direction of the prior move (TrendSpider). A common invalidation rule: once price trades fully through the gap, the imbalance is "filled" and the setup is spent. As one guide puts it, an FVG is "a one-time use area. Once price trades back into it and fills the imbalance, the edge is gone" (Daily Price Action: Fair Value Gaps).
How practitioners mark and trade them
Both concepts share one rhythm: find the zone, wait for price to return, trade the reaction. The mechanics differ by direction.
| Concept | The zone | Bullish version | Bearish version |
|---|---|---|---|
| Order block | The last opposing candle's range | Last down candle before an up impulse; expect support on return | Last up candle before a down impulse; expect resistance on return |
| Fair value gap | The 3-candle wick gap | Gap between candle-1 high and candle-3 low; buy the retrace | Gap between candle-1 low and candle-3 high; sell the retrace |
| Trigger | Return to the zone ("mitigation" / "fill") | Enter long on the retest | Enter short on the retest |
| Common stop | Beyond the zone | Below the block / gap | Above the block / gap |
Where the honesty problem lives
That last point is the crux, and reputable critics of SMC do not soften it. The framework has no single agreed definition of a valid order block. How strong must the "impulsive" move be? Does the block break structure, or just move sharply? Do you use the candle body or the full wick? Is the first down candle valid, or only the last? Every tutorial answers a little differently.
Critics point out two consequences. First, order blocks are largely a rebrand of classic supply-and-demand zones, dressed in an institutional narrative that a chart cannot confirm (EarnForex: Smart Money Concepts flaws). Second — and more damaging for anyone who wants proof — the discretion makes the method very hard to test. When a mechanical backtest of isolated FVGs performs poorly, the reply is usually that it was applied wrong, because "there is no objective way to use SMC; it is a framework that depends on how the person who uses it decides to use it" (Sentient Trading Society: SMC and backtest limitations). A method that can always be rescued by reinterpretation can never be cleanly disproven — and something that can never be disproven cannot really be validated either.
None of this means order blocks and FVGs "don't work." It means "it works" has no fixed meaning until the rules are pinned down.
What you'd need to make them testable
The fix is not to abandon the ideas but to remove the discretion. To backtest an order block or FVG the way you would any other rule, every judgement call has to become a number a machine can check:
- Exact candle logic. Define "the block" precisely — the last candle whose close is opposite the next candle, body or wick, with no ambiguity about which candle qualifies.
- A displacement threshold. Quantify "impulsive": the move away must span at least N times the recent average range within k bars. No eyeballing.
- A gap-size floor. Require the FVG to exceed a minimum width (points, or a multiple of ATR) so trivial one-tick gaps are excluded.
- A concrete entry. A fixed retrace level — the zone edge, or the 50% point — not "somewhere around it."
- A hard invalidation and expiry. Define the stop, the "filled" condition, and how many bars a zone stays live, so each zone has one unambiguous lifespan.
Once those are fixed, the zones stop being art and become a rule set — and a rule set can be run against history, checked on unseen data, and stress-tested for costs. That is the whole point of out-of-sample testing: a definition you cannot state precisely is one you cannot honestly claim an edge for. It is also the difference between a screenshot and something you could put through a documented cost-and-parity model.
Frequently asked
What is the difference between an order block and a fair value gap?
An order block is a single opposing candle before an impulsive move, and its zone is that candle's own high-to-low range. A fair value gap is a three-candle imbalance — the empty space between the first and third candles' wicks. They often appear together, but they are marked from different candles.
What does it mean when an order block or FVG is "mitigated"?
Mitigation is the moment price trades back into the zone after leaving it. Many traders treat the first return as the trade signal and consider the zone "spent" once price has fully passed through it.
Are order blocks and fair value gaps reliable, and can you backtest them?
There is no fixed answer, because there is no single agreed definition to measure. The zones are drawn discretionarily, so the same setup gets marked differently by different traders — which makes an apples-to-apples backtest difficult. You can only test them once every judgement (candle logic, displacement, gap size, entry, invalidation) is converted into an explicit rule.
The stubborn takeaway
Order blocks and fair value gaps are useful vocabulary for describing where price moved fast and where it might return. But a zone you draw by feel is a story, not an edge. The moment you can state the rule so exactly that a machine marks the same block you would — and then run it on data you have never seen — you find out whether the story was ever true. Test it; don't trust it.