CRT Trading Strategy: Rules, Entries, and Trade Management

A CRT trading strategy is a rules-based way to trade Candle Range Theory: treat a higher-timeframe candle as a range, wait for price to sweep one side of it, demand a close back inside, and trade toward the other side. Every step has a definition. None of it requires you to feel anything about the chart.

That last part is the point. Most traders who lose with CRT lose because they run it on vibes — a sweep that wasn't really a sweep, a confirmation they didn't wait for, an entry with no defined invalidation. This article lays out the full workflow as mechanical rules, then covers the failure modes honestly.

The strategy in one paragraph

Pick a higher-timeframe (HTF) candle to act as the parent range. Wait for a later candle's wick to take out the parent's high or low. If that candle closes back inside the parent's range, you have a confirmed CRT: the swept side was a liquidity grab, and the opposite side of the range becomes the target. Enter on the order block left behind at the sweep, stop beyond the sweep's extreme, target the far side of the parent. One setup, one trade.

Everything below is that paragraph, expanded into rules you can actually test.

Step 1 — Pick a timeframe pair

CRT runs on two timeframes: the HTF defines the parent range, the LTF gives you entry resolution. Common pairs:

• H4 parent / M15 entry — the standard swing-intraday pair
• H1 parent / M5 entry — faster, more setups, more noise
• D1 parent / H1 entry — slower, fewer setups, larger ranges

The only hard rule: the entry timeframe must be strictly finer than the parent timeframe, and it should divide evenly into it. Pick one pair and keep it fixed for the duration of any test. Switching pairs mid-sample is how traders convince themselves a broken strategy works.

Step 2 — Identify the parent candle

The parent is the earliest closed HTF candle whose high and low have not yet been taken out by any later closed candle. Not the biggest candle, not the one that looks important — the earliest one still untouched on both sides.

Two details matter:

• Wick level, not body. A later candle's wick reaching the parent's high counts. Equal price counts — a touch is a sweep.
• Closed candles only. A candle still forming is never the parent.

Once a CRT plays out against a parent — target hit, range midpoint touched, or a close outside the range — that parent is done. It never becomes the parent again. The candle that ended the setup becomes the next parent.

Step 3 — Wait for the sweep

A sweep happens the moment a later candle's wick touches or exceeds one of the parent's extremes. High swept means the setup, if it confirms, is bearish. Low swept means bullish. The mechanics of why swept extremes matter — resting stops, engineered liquidity — are covered in liquidity sweeps.

The sweep alone is not a trade. It is a candidate. What the sweep candle does next decides everything.

Step 4 — Demand confirmation

Confirmation is binary: the sweep candle closes back inside the parent's range, or it doesn't.

• Close back inside — confirmed CRT. The wick beyond the range was a grab, the range held, and the opposite extreme is now the draw.
• Close outside — no CRT. That is expansion, not reversal. Price took the level and kept it. The parent is finished and the breakout candle becomes the new parent.

This single rule filters out the most expensive mistake in range trading: shorting a breakout because the high got "swept." A sweep that closes outside was never a sweep. It was the move.

Step 5 — Enter at the order block

You do not chase the close. The sweep leaves a footprint on the lower timeframe: the LTF candle that pushed into the parent's extreme, then engulfed by an opposite-direction close. That two-bar structure is the order block, and it only exists once the engulfing close prints — the reversal is structural, not assumed.

The entry rule: wait for price to trade back into the order block zone. That retrace is your fill. After a high sweep, the OB is the bullish candle that drove deepest into the high, and you sell its mitigation. After a low sweep, mirror it.

If price never comes back to the zone, you miss the trade. That is the correct outcome. A missed entry costs nothing; a chased entry has no defined risk.

Step 6 — Stops and targets

• Stop: beyond the most extreme point of the sweep structure — the wick that took the parent's level, including the reversal bar's wick. If price trades back through the full sweep, the reversal thesis is dead and you want out.
• Target: the opposite side of the parent's range. High swept, target the parent's low. Low swept, target the parent's high. The range defines the trade; the trade does not outlive the range.

One more management line worth knowing: the midpoint of the parent's range. If price sweeps an extreme and then trades all the way back through 50% of the range, the setup has expended its energy — no new entries on that CRT. It is a lifecycle boundary, not a stop for trades already open. Define up front whether you hold through it.

And the discipline rule that makes the whole thing testable: one trade per CRT. The first valid mitigation of the order block is the trade. There is no second entry, no re-entry after a stop, no averaging. One setup, one decision, one outcome.

Time of day matters

Not every sweep carries the same intent. In CRT and the broader ICT framework it descends from, sweeps during the London open (roughly 02:00–05:00 ET) and the New York open (07:00–10:00 ET) are the ones worth the most attention — those windows are where engineered liquidity grabs cluster, because that is when the volume to fuel the reversal shows up.

Be honest about what this is: context, not a mechanical filter. If you want time-of-day in your ruleset, write it as a rule before you trade it — "setups only when the sweep prints inside a kill zone" — and test it against the unfiltered version. Do not apply it after the fact to explain losers.

What goes wrong

Three failure modes account for most blown CRT trades.

Dual sweeps. One HTF candle takes out both sides of the parent's range. There is no clean directional read — whichever side went first, the other side was the target, and the whole lifecycle resolved inside a single candle. When the two crosses land on the same LTF bar, direction cannot be resolved at all. Rule: an outside bar consumes the parent and becomes the new parent. You stand down and wait for the next range.

Expansion candles. The sweep that closes outside the range. Covered above, but it deserves repeating because it is the trap traders fall into most: the wick takes the high, it looks exactly like the start of a reversal, and then the candle settles above the range. Without the close-inside rule you are fading a breakout. With it, you simply never take that trade.

Chasing unconfirmed setups. Entering on the wick touch itself — no reversal structure on the LTF, no order block, no defined invalidation. Sometimes it works, which is the problem: intermittent reinforcement teaches the habit. The rule set exists precisely so that "it touched the level" is never, on its own, a reason to be in a trade.

Rules you can replay

The reason to make every step mechanical is that mechanical rules can be checked. A discretionary CRT trade that loses teaches you nothing — maybe the setup was bad, maybe the execution was. A rules-based CRT trade that loses is data.

CRT Terminal runs this exact detection logic — parent selection, wick-level sweeps, confirmation closes, order blocks, signals — as a live engine on real market data, with frame-by-frame replay and backtesting so you can see every setup the rules produced and why each one won or lost. A terminal, not a charting app.

Start a 7-day trial — no card required. Plans are on the pricing page.

FAQ

What is the CRT trading strategy?

CRT (Candle Range Theory) treats a higher-timeframe candle as a tradable range. When a later candle's wick sweeps one side of the range and closes back inside it, the strategy enters on the order block left at the sweep and targets the opposite side of the range.

What timeframes work best for a CRT strategy?

Any pair where the entry timeframe is strictly finer than the parent timeframe — H4/M15 and H1/M5 are the common choices. Pick one pair and keep it fixed while you test.

Where do the stop loss and take profit go in CRT?

The stop sits beyond the most extreme wick of the sweep structure; the target is the opposite side of the parent candle's range. Both levels are defined by the range itself, not by discretion.

What invalidates a CRT setup?

Three things: the sweep candle closing outside the parent's range (expansion, not reversal), price returning through the 50% midpoint of the range after the sweep, or both sides of the range being taken in one candle (a dual sweep).

Is CRT the same as ICT?

CRT grew out of ICT liquidity concepts but narrows them to one repeatable structure: the swept-and-reclaimed higher-timeframe range. See CRT vs ICT for the full comparison.

Educational content, not financial advice.