Candle Range Theory (CRT) is a newer setup from the ICT lineage built on one elegant observation: every candle is a range. A 4-hour candle's high and low are a miniature session — and when the next candle sweeps one side of that range and closes back inside, the third candle often runs to the opposite side. Three candles: accumulation, manipulation, distribution.

If that sounds familiar, it should — CRT is the liquidity sweep pattern formalized to individual higher-timeframe candles, or equivalently a compressed failed breakout. Its appeal is precision: clear levels, clear invalidation, mechanical enough to backtest properly.

The three-candle sequence

Using 4H candles as the "range" timeframe:

1. Candle 1 — the range: any reasonably wide candle; its high and low become the levels
2. Candle 2 — the manipulation: price wicks below candle 1's low (sweeping the stops resting there) but closes back inside candle 1's range — the failed raid
3. Candle 3 — the distribution: with the downside fuel spent, price runs toward candle 1's high — the opposite side of the range

Bearish version: sweep of the high, close back inside, run to the low.

The strategy rules

• Timeframes: 4H or daily candles define the range; 15m chart executes the entry
• Setup (long): candle 2 sweeps candle 1's low and closes back inside the range
• Entry: drop to 15m — after the sweep, enter on a structure shift (break of the last 15m lower high) or on the pullback into the 15m FVG that shift creates; conservative version: enter at the reclaim of candle 1's low
• Stop-loss: below candle 2's wick low — the sweep extreme; if price trades back below it, the "manipulation" was just a breakdown
• Target: candle 1's high (the classic CRT objective); partials at 2R if the range is wide
• Quality filters: sweeps of levels that also matter structurally (equal lows, prior support) outrank random candle lows; CRT traders also weight setups forming into session opens (London/New York), where the volume to fuel candle 3 actually shows up

Why it can work — and where it breaks

The mechanism is the oldest reliable one in trading: obvious levels accumulate stops; raids of those stops that fail to follow through strand aggressive sellers, whose covering powers the reverse move. CRT's candle framing just gives the pattern fixed, checkable coordinates.

Where it breaks: in strong trends, "sweeps" are simply continuation — candle 2 closes back inside, and candle 3 breaks the low anyway. Counter-trend CRT longs in a heavy downtrend lose repeatedly; the setup performs best at higher-timeframe support in ranging or rotational conditions. Which conditions dominate your market? That's a backtest question, not a YouTube question.

How to backtest it

CRT's mechanical definition makes it one of the best replay projects:

1. In chart replay on 4H candles, advance until candle 2 sweeps and closes back inside — log the setup before seeing candle 3.
2. Record: did price reach the opposite side? How far did the wick extend first? Would your 15m entry have filled?
3. 50 setups, tagged by context (with-trend / counter-trend / at HTF level). Expect the edge to concentrate at meaningful levels — and the counter-trend subset to be the drawdown machine.

Common mistakes

1. Trading every candle's wick as a "sweep" — without stops resting at the level, there's no manipulation, just movement.
2. Entering before candle 2 closes — a wick below the range is nothing until the close confirms rejection.
3. Counter-trend CRT in strong trends — the classic failure mode; check higher-timeframe structure first.
4. Stops inside candle 2's wick — normal turbulence at the low will tag you before the run; the wick extreme is the line.

FAQ

What timeframes work best for CRT?

The common pairings are 4H range/15m entry and daily range/1H entry. Below 1H ranges, sweeps are so frequent the pattern loses meaning.

Is CRT the same as a failed breakout?

Structurally yes — CRT is a failed breakout of a single candle's range, with ICT-style entry refinement (structure shift, FVG). The formalization makes it easier to backtest consistently, which is its real value.

Does CRT work in crypto?

Yes — crypto's stop-hunt-heavy behavior around obvious levels suits it, especially on BTC/ETH 4H ranges around session opens. Verify on your pair with a 50-setup replay sample before trading it live.