CRT Education · · 9 min read

Turtle Soup Trading Strategy: The Complete CRT Guide

What the turtle soup pattern is, why false breakouts happen, and how to trade it step by step with Candle Range Theory: levels, timing, entries and risk.

Turtle soup is a reversal setup built on one simple fact: obvious highs and lows are where the stops are, and price goes there to collect them before it moves the other way. A turtle soup happens when price runs beyond a well-known high or low, takes the liquidity resting there, then closes straight back inside the range. The breakout traders who jumped in are now trapped, and their losing positions fuel the move in the opposite direction.

In my Candle Range Theory model, the turtle soup isn't a side setup. It's the trigger. Candle 2 of a CRT sweeps one side of Candle 1's range, and the turtle soup on a lower timeframe is what tells me that sweep is finished. This guide covers the whole thing: where the pattern comes from, why it works, the exact rules I use, and the mistakes that cost people money with it.

Where the name comes from

In 1983, Richard Dennis and William Eckhardt ran a famous experiment. They recruited a group of beginners, taught them a strict set of trend-following rules and gave them money to trade. They called them the Turtles. One of the core entries was simple: buy when price makes a new 20-day high, sell when it makes a new 20-day low.

Breakout rules like that work in strong trends. The problem is that most breakouts fail. Linda Bradford Raschke noticed that as more traders bought every new 20-day high, more of those breakouts reversed right after triggering. In Street Smarts (1995), written with Larry Connors, she published the counter-strategy and named it after the people it traps: Turtle Soup. Wait for the breakout, then trade the failure.

The original rule set was mechanical: a new 20-day low, with the previous 20-day low made at least four sessions earlier, then a buy stop back at the old low once price returned above it. Turtle Soup Plus One waits one more bar: the breakout closes beyond the level, and the entry comes on the next bar if price comes back. The extra bar filters out some false signals at the cost of a slightly later entry.

The details changed over the years. The idea didn't: obvious levels attract stops, and stops attract price.

Why false breakouts happen: liquidity

Liquidity simply means orders waiting to be filled. Large participants can't buy or sell size without someone on the other side, so they need places where many orders sit together. The easiest places to find them are the levels every retail trader can see:

  • Previous day, week and session highs and lows: the Asia high, the London low, yesterday's high (PDH) and low (PDL).
  • Equal highs and equal lows: two or three touches at the same price look like strong resistance or support, which is exactly why stops pile up just beyond them.
  • Swing highs and lows on the higher timeframes.
  • Round numbers and textbook support and resistance.

Above every obvious high sits buy-side liquidity: the stops of traders who are short, plus the buy orders of breakout traders. Below every obvious low sits sell-side liquidity. When price pushes through one of those levels, all those orders trigger at once. A large player who wants to sell can sell into that burst of buying at a good price. Once the orders are used up, nothing is left to push price higher, and it reverses.

That's the stop hunt. Turtle soup is how you trade it from the right side instead of being part of the fuel.

Turtle soup inside Candle Range Theory

Candle Range Theory looks at three candles on a higher timeframe:

  1. Candle 1 sets a range: its high and its low.
  2. Candle 2 runs one side of that range, taking the liquidity there, and closes back inside it.
  3. Candle 3 delivers to the other side of Candle 1's range.

Read that again and you'll notice Candle 2 is a turtle soup on the higher timeframe. That's why the two fit so well. CRT gives you the map: which range, which side gets swept, and where price is likely to go next. The turtle soup on a lower timeframe gives you the timing: the moment the sweep is actually done.

Here's the definition I use in the CRT Radar. A turtle soup prints when a candle on the turtle-soup timeframe runs beyond Candle 1's level, beyond the previous turtle-soup candle's extreme, and closes back across it. For a short, that candle takes the high and closes back below. For a long, it takes the low and closes back above. The first lower-timeframe close back across the swept level is the signal. A wick through the level with no close back is not a turtle soup; it's just a sweep that may keep going.

Step by step: how I trade it

Step 1: Know your higher-timeframe bias and draw

Before anything else, decide where price is likely to go on the daily and weekly charts. Which side of the range has been swept already? Which pool of liquidity is still untouched? That untouched pool is the draw on liquidity, your eventual target. The DOL Map does this for me every day. A turtle soup that agrees with the higher-timeframe draw is worth taking. One that fights it usually isn't.

Step 2: Mark the levels before the session

Mark Candle 1's high and low, the previous day's high and low, the Asia and London session extremes, and any equal highs or lows. Note where these levels line up with a higher-timeframe key level (an order block, fair value gap or rejection block on H1 to daily). A sweep into a key level that also takes liquidity is the strongest version of the setup. My rule: the purge has to take the liquidity inside the higher-timeframe key level first, otherwise it's just noise.

Step 3: Wait for the sweep in the right window

Don't anticipate the sweep. Let price take the level. Timing matters a lot here: stop hunts cluster around the session opens because that's when the big volume arrives. The London open and the New York open are where most of my turtle soups form. A sweep in the dead hours of the Asian afternoon carries much less weight. My CRT timings guide covers the exact windows.

Step 4: Confirm the turtle soup on the lower timeframe

Drop to the turtle-soup timeframe for your model: 15 minutes for the Classic 2-6-10 model and London, 30 minutes for the H6 and H7 models. Wait for the candle that runs the level and closes back across it. That close is your turtle soup. Until it prints, there is no trade, no matter how good the sweep looks.

Step 5: Enter on a requirement, not on a feeling

After the turtle soup, I wait for one of four entry requirements to print on the entry timeframe:

  • Purge order block: the last opposite candle before the move away from the sweep.
  • Inverse fair value gap (IFVG): a gap that price closes back through.
  • Balanced price range (BPR): where a bullish and a bearish gap overlap.
  • Change in state of delivery (CISD): price closes back through the open of the run that went into the purge.

Each one has to print after the turtle soup, not before. The full breakdown is in CRT entry model: 4 entry triggers after the turtle soup.

Step 6: Stop, target and management

  • Stop: just beyond the wick that made the sweep. If price closes back beyond it, the setup is wrong and you're out with a small, defined loss.
  • First target: the 50% of Candle 1's range.
  • Final target: the opposite side of Candle 1's range, or the higher-timeframe draw if it lines up.

Because the stop sits right behind the sweep, the risk is often small compared to the distance to the target. That's where the big risk-to-reward trades come from. It isn't from reaching for huge targets; it's from a precise stop. Most trades are far more modest, and some will lose. The structure is what keeps the losses small.

Five mistakes that ruin turtle soup trades

  1. Fading every breakout. Some breakouts are real. If the higher-timeframe draw is in the breakout's direction and there's no turtle soup close, step aside.
  2. Entering on the wick. Price can run much further past a level than you think. The close back across the level is the signal, not the first touch.
  3. Ignoring time. A turtle soup at the New York open and one at 2:30 pm are not the same trade.
  4. Taking sweeps of meaningless levels. If the level doesn't hold real liquidity, there's nothing to trap. Stick to session highs and lows, PDH and PDL, equal highs and lows, and higher-timeframe swings.
  5. Moving the stop. The sweep wick is the line. Widening it after entry turns a small loss into a big one.

When a setup breaks the rules, walk away. CRT setup invalidation: 6 signs to walk away lists the ones I use.

A turtle soup trading plan

Write your plan down before the session so there's nothing to decide in the heat of the moment:

  • Markets and times: the instruments you trade and the windows you trade them in.
  • Levels: which highs and lows count, marked before the open.
  • Trigger: the sweep, then the turtle-soup close, then one of the four entry requirements.
  • Risk: a fixed amount per trade. Many traders use 0.5% to 1% of the account; prop firm traders size from the daily loss limit.
  • Journal: a screenshot, the level that was swept, the time, the entry requirement and the result, every single trade. Review it weekly.

If you can't write a rule down, you can't follow it consistently. The plan is what turns the pattern into a process.

Turtle soup and Wyckoff

If you've studied Wyckoff, the turtle soup will look familiar. A spring is price dropping below the low of a trading range and quickly coming back in. An upthrust is the same thing at the top. Both are turtle soups seen through Wyckoff's lens: the sweep that ends accumulation or distribution. My Wyckoff guide shows how the two fit together.

Let the chart do the marking

I built the CRT Radar because marking all of this by hand every session gets old fast. It draws Candle 1's range, flags the sweep, prints the turtle soup when the close confirms it and then shows the entry requirements as they form. You still decide whether to take the trade; it just stops you from missing the sequence. The CRT Handbook covers the method in full, and CRT University has the free lessons in order.

FAQ

What is the turtle soup pattern in trading?

A reversal setup where price breaks beyond an obvious high or low, takes the stops resting there, then closes back inside the range. The failed breakout traps traders on the wrong side, and their exits push price the other way.

Is turtle soup an ICT concept?

No. Linda Raschke published it in 1995. ICT and other smart-money teachers later used the name for the same liquidity-sweep idea. In CRT, it's the lower-timeframe confirmation that Candle 2's sweep is finished.

What timeframe is best for turtle soup?

Use a higher timeframe for the range and a lower one for the confirmation. In my models, that means an H4, H6 or H7 range with a 15 or 30-minute turtle soup, then entries on 3 to 15 minutes depending on the model.

Where do you put the stop on a turtle soup trade?

Just beyond the wick that made the sweep. A close back beyond it means the breakout was real and the setup has failed.

What's the difference between a turtle soup and a liquidity sweep?

Every turtle soup includes a liquidity sweep, but not every sweep is a turtle soup. The turtle soup needs the close back across the swept level. Without that close, price may simply keep going.

Education only, not financial advice. Trading involves substantial risk of loss.