CRT Education · · 5 min read
The CRT Asia Model: Trading the Asia Open From the 3 PM H7 Candle
How the Asia Model works: the 3 PM H7 candle sets the range, the Asia open sweeps it into an H4 key level, a 15-minute turtle soup confirms, and the entry comes on M5.
Most traders close the charts at 4 PM and come back for London. That's exactly why the Asia open is one of my favourite windows to trade. The first hours after 6 PM New York time regularly run an obvious high or low left over from the afternoon, and when that run fails, it fails fast and cleanly.
The direct answer: the CRT Asia Model uses the 3 PM H7 candle (New York time) as Candle 1. At the Asia open, Candle 2 sweeps one side of that range into a past H4-or-higher key level, a 15-minute turtle soup confirms the sweep failed, and the entry comes on the 5-minute chart. The targets are the 50% and the far side of the 3 PM range.
That's the whole model. Below is how each step works and why it's built this way.
Why the Asia open, and why the 3 PM candle?
The killzones guide says the dead hours after the New York close are a coin flip, and most of the time that's true. The Asia open is the exception, for one reason: liquidity left behind.
By 3 PM, New York has done its work for the day. The afternoon builds a range, and the stops of everyone who traded that range sit just above its high and just below its low. Then the market goes quiet, reopens at 6 PM, and Asia's first push almost always goes looking for that resting liquidity.
The H7 candle that opens at 3 PM New York time captures that whole afternoon-into-evening range in a single candle. That's Candle 1. Its high and its low are the range I care about overnight.
Step 1: mark the range (Candle 1)
- Candle 1 = the 3 PM H7 candle. Its high and low are the range.
- I cut the model off at midnight. Anything after that belongs to London's setups, and I don't want the two overlapping.
Nothing to trade yet. I just know where the liquidity sits.
Step 2: the sweep into a key level (Candle 2)
At the Asia open, price runs one side of the 3 PM range. On its own that means nothing; breakouts happen every night. What turns it into a CRT is where the run ends:
- It has to reach a past H4, daily, weekly or monthly key level: an order block, a breaker block, a rejection block, an FVG, an opening gap, or an old high or low like the previous day's high.
- If the run stops in the middle of nowhere, there's no trade.
This is the part most people skip. A sweep into nothing is just a breakout. A sweep into an H4 level that has already proven itself is a trap.
Step 3: the 15-minute turtle soup
Now I need proof that the run failed. On the Asia Model that's a 15-minute turtle soup: a 15-minute candle runs the previous 15-minute high (or low) beyond the 3 PM range, then closes back inside it.
That close back inside is the whole thesis of the trade: price went to the level, found nothing, and came back. No turtle soup, no trade, however good the level looks.
Step 4: the entry on M5
Once the turtle soup prints, I drop to the 5-minute chart for the entry. There are two:
- Entry #1, the purge order block: the last opposing candle at the extreme of the sweep. Price often comes back to it quickly. It's the tighter entry, which is where the big risk-to-reward numbers come from.
- Entry #2, the CISD (change in state of delivery): a 5-minute close back through the open of the candles that ran into the sweep. It comes a little later and with a slightly wider stop, but it's the safer of the two, because the market has already shown its hand.
I used to take Asia entries on the 3-minute chart. It was too fast and gave too many false starts, so the model moved to M5.
Step 5: stop and targets
- Stop: just beyond the extreme the sweep made, plus a small buffer for spread. If price trades back through it, the run didn't fail and I'm wrong.
- Target 1: the 50% of the 3 PM range.
- Target 2: the far side of the 3 PM range, or the next untaken key level beyond it.
The full logic for stops and targets is in the stop loss and take profit lesson.
A real example: MNQ short, November 17
On Tuesday, November 17, 2026, MNQ pushed up into the Asia open, ran the afternoon high into a key level and failed. I shorted at 27,569.75 with a 5-point stop at 27,574.75. Price came back through the range and kept going overnight, to my target at 26,977: 592.75 points, about 1:118.
Honest disclaimer
That's one of my best Asia trades, not a typical one. Most trades are far smaller, plenty lose, and a 5-point stop on a fast open can slip. Read it as an example of the structure, not a promise of results.
The Asia Model checklist
- Mark the high and low of the 3 PM H7 candle.
- Wait for the Asia open (6 PM New York) to run one side.
- The run must reach a past H4-or-higher key level.
- Wait for the 15-minute turtle soup: a close back inside the range.
- Enter on M5: purge order block, or the safer CISD.
- Stop beyond the sweep; targets at the 50% and the far side of the range.
- Nothing after midnight.
Let the Radar do the watching
Nobody wants to sit up every evening waiting for a sweep. The Asia Model is built into the CRT Radar as a preset: it marks the 3 PM range, the key levels, the turtle soup and both entries, and alerts you when the sweep starts. The full rulebook for every model is in the CRT Handbook, and I call these trades live in the trading room.