PromptDesk Pro ← Back to the blog

FVG, Order Blocks and liquidity: ICT/SMC concepts explained without the hype

Published July 6, 2026 · ICT/SMC · 7 min read

ICT/SMC concepts have a reputation problem: half the internet sells them as a secret bank code, and the other half dismisses them as astrology. The reality is more boring and more useful — they are a structured way of reading price: where unfilled orders were left behind, where stops have piled up, and when control of the market changed hands. Neither magic nor garbage: a language. And like any language, it works if you speak it precisely.

Structure: BOS and CHoCH

Everything starts with market structure — the sequence of highs and lows:

The distinction matters because it defines your bias: trading in the direction of the last BOS is trading with the trend; trading after a CHoCH is betting on the reversal, and that demands more confirmation, not less.

Liquidity: where the stops are

The central concept of the whole approach. Markets need counterparties: for someone to buy big, someone has to sell big. Where is selling guaranteed? Where the buyers' stops are — below obvious lows, below the support levels everyone can see, at equal lows.

This is where the most valuable practical rule of the approach comes from: obvious levels are not entry zones — they are magnets. The question is not "will support hold?" but "has the liquidity below already been swept, or not yet?".

FVG: the gap price left behind

A Fair Value Gap (imbalance) appears when price moves so fast that it leaves a three-candle gap: the high of the first candle never touches the low of the third (or the reverse). Almost no trading took place in that stretch — the move was so one-sided that orders were left unfilled.

The trading hypothesis: price tends to revisit that gap ("rebalance") before continuing. That's why the FVG is used as a pullback entry zone: instead of buying the impulsive candle (late and expensive), you wait for the return to the gap that candle left behind.

Order Block: the last opposing candle

An Order Block is the last bearish candle before a strong bullish impulse (or the last bullish one before a sell-off). The institutional reading: that's where the large operator built their position before moving price — and if price comes back to that zone, they're likely to defend it.

Not every opposing candle is an order block worth anything. The criteria that separate signal from noise: the impulse that follows breaks structure (BOS), the move swept liquidity before taking off, and it leaves an FVG along the way. An OB with none of that is just a red candle.

The sequence: how they connect (and why order matters)

Loose concepts are not a strategy. What makes them tradeable is the logical sequence — a classic long setup example:

  1. Context (higher timeframe): an uptrend defined by successive BOS on the 4H/daily. Your bias is long.
  2. Manipulation: on the pullback, price sweeps an obvious low on the 15m (liquidity sweep) — early buyers' stops fund the big money's entry.
  3. Confirmation: after the sweep, a bullish CHoCH on the entry timeframe (1m–5m): control returned to the buyers.
  4. Entry: on the pullback to the FVG or order block left by that confirmation move, with the stop below the low of the sweep.

Each piece answers a different question: the BOS defines which way, liquidity defines from where, the CHoCH defines when, and the FVG/OB defines the exact price. If your "ICT strategy" can't fill those four boxes with concrete rules, it's not a strategy yet — it's a vocabulary.

Turn these concepts into rules an AI can execute

The PromptDesk Pro generator builds your strategy step by step with 38 ICT/SMC concepts — each with its role in the sequence (context, manipulation, confirmation, entry) — and produces a professional prompt ready to paste into any AI and turn into rules or code. No ambiguity, no "it depends on context".

Create my strategy →

The honest warning

None of this carries a statistical guarantee on its own. An FVG is a hypothesis about a reaction, not a promise; serious backtests of these concepts show results that depend enormously on the instrument, the session and the exact execution rules. The only way to know whether your version of this works is the usual one: rules written without ambiguity, a large sample of trades, and statistical validation — not a YouTube video with three hand-picked examples. We cover that in the Monte Carlo guide and the risk of ruin guide.

In one sentence: ICT/SMC is a language for reading liquidity and structure — useful if you turn it into verifiable rules, noise if you leave it as jargon.

This content is educational and does not constitute financial advice. No concept or methodology guarantees results. Trading involves risk of loss.