What is the Turtle Soup trading strategy success rate?

This page reports no measured success rate. A defensible number needs a precise version, symbol, dates, costs, fill assumptions and count of closed trades. Report net expectancy and drawdown alongside it; win rate alone does not describe profitability.

Are RSI, EMA200 and volume required for Turtle Soup?

They are custom filters in this page. Test each independently before combining them, and use only values available when the entry decision is made. Fewer accepted trades do not prove that the retained trades are better.

Does a wick through a level qualify by itself?

No, not in the daily-close research version here. It also requires the level age, a close back through the reference and the other eligibility rules. The entry follows the completed signal; it is not backdated to the wick.

Can a daily-close test reproduce an intraday entry?

No. Waiting for the daily close changes the information, timing and possible fill. Label it as a separate version and do not attribute its results to an intraday rulebook.

TradingView strategy research

Turtle Soup trading strategy: failed breakouts and filters

The Turtle Soup trading strategy trades a failed breakout back through a prior price extreme. Its success rate depends on the exact rules, instrument, sample and trading costs; this page reports no measured win rate. RSI, EMA200 and volume filters below are separate research choices, with explicit comparison tests. They are not evidence that a reversal will hold.

Draft the specified strategy

Historical Turtle Soup rules and the Turtle comparison

Street Smarts by Larry Connors and Linda Raschke lists Turtle Soup and Turtle Soup Plus One as separate chapters. TradingMarkets describes Turtle trend entries as trading breakouts and Turtle Soup as trading their failure. Its historical long setup is summarized below; shorts reverse the directions.

Price makes a new 20-day low; the previous 20-day low occurred at least four trading sessions earlier.
After that break, place a buy stop 5 to 10 ticks above the old low, valid for that day only.
After entry, place a protective sell stop one tick below the session low. A live rule must use the low known when the stop is placed.
Turtle Soup Plus One moves the setup to the following day and has different conditions. Keep it as a separate experiment.

A daily-close research version with explicit rules

The following is a custom, untested specification. It replaces the historical intraday stop entry with daily-close confirmation and adds mechanical exits. That choice changes the trades and prices. Use standard daily candles, a fixed exchange/session and one tradable unit with sufficient capital. For comparable filter tests, wait for at least 400 completed daily bars and valid indicator values in every run.

Allow one position and no pyramiding. Evaluate a signal only if flat at that close; ignore signals while an entry is pending. If a session breaks both L and H, skip it even if only one side passes the age check.
For each accepted trade, R is the positive distance between the actual entry fill and the frozen stop. Place a target 2R from that fill in the trade direction. Keep both prices fixed, with no trailing or opposite-signal exit.
Make the protective stop active with the entry; activate the target when the fill is known. Document the implementation timing. Do not use later bar extremes to improve the initial stop.
Count the entry session as session zero. If still open at the close of session five, submit a market exit for the next available fill and cancel the remaining exit orders after the position closes.
If the entry gaps past the stop, record the actual fill and exit at the first executable opportunity. Do not remove that loss from the sample or assign a negative R target. Re-entry requires a new qualifying daily close.

Decision	Long setup	Short setup
Freeze the reference	L is the lowest low of the 20 completed sessions before today.	H is the highest high of the same preceding 20 sessions.
Check level age	The most recent occurrence of L is at least four sessions before today. A tied low yesterday fails.	Apply the same age rule to the most recent occurrence of H.
Confirm the failure	Today trades strictly below L and closes strictly above L.	Today trades strictly above H and closes strictly below H.
Submit the entry	After the qualifying close, submit a market buy for the next available fill.	After the qualifying close, submit a market sell for the next available fill.
Freeze the stop	One minimum tick below the signal session low.	One minimum tick above the signal session high.

A hypothetical reclaim, followed by two possible outcomes

Assume a fictional instrument with a 0.01 tick and a point value of 1 per unit. The preceding range is 100 to 108; the most recent 100 low was six sessions ago. Today has a high of 102, low of 99 and close of 100.20. It passes the long rule without breaking both sides. These prices are a worked example, not market observations or a backtest.

The two outcomes are alternatives, not two observed trades. The assumed cost is arithmetic only; actual commissions, spread and slippage need instrument-specific inputs.
If the signal close were 99.80, there would be no trade in this version. A later rebound does not retroactively qualify it.
If the next fill were 98.50, it would already be below the planned long stop. Record the gap and executable exit, not a fictional fill at 98.99.
A valid reclaim can still stop out when the original downward move resumes. Passing the pattern test is separate from making money.

Step	Calculation	Interpretation
Decision close	99 < 100 and 100.20 > 100	Submit the long order after the close.
Assumed next fill	Entry 100.25; stop 99 - 0.01 = 98.99	The fill differs from the signal close.
Planned distance	R = 100.25 - 98.99 = 1.26	This excludes costs and gap risk.
Target branch	100.25 + 2 × 1.26 = 102.77	If filled exactly, gross gain is 2.52 per unit.
Stop branch	98.99 - 100.25 = -1.26	If filled exactly, gross loss is 1.26 per unit.
Illustrative cost	0.04 per unit for the entire round trip	Net target gain 2.48; net stop loss 1.30.

RSI, EMA200 and volume are independent filter experiments

All thresholds here are illustrative choices fixed before testing. Evaluate them at the signal close in the daily-close version above. Each filter starts from that same baseline; do not change entry, stop or target rules while comparing it. RSI describes price momentum, and an EMA smooths price history. Neither measures the probability of a successful Turtle Soup trade.

Skip the volume-filtered candidate if any required volume value is missing or the prior mean is zero. Record exclusions. State whether the feed reports traded volume or tick activity; do not compare unlike volume series as if they were the same measure.
In the numeric example, suppose RSI is 34, EMA200 is 101 and the volume ratio is 1.30. RSI and volume pass, EMA fails because 100.20 is below 101, so the combined run rejects the trade. These values are also hypothetical.
Do not transfer a same-day closing RSI or full-day volume condition to the historical intraday entry. Those values were not yet available. An intraday adaptation needs its own information timestamps.

Run	Additional condition	Question to test
Baseline	No additional filter.	What are the net results and trade count of the written rules alone?
RSI only	RSI of daily closes, length 14: at most 35 for longs; at least 65 for shorts.	Does selecting these momentum readings change net expectancy on untouched dates?
EMA200 only	Signal close above the 200-day EMA for longs; below it for shorts. Equality fails.	Does trading the rebound in this broader price direction help, or just remove trades?
Volume only	Signal volume at least 1.2 times the mean volume of the previous 20 completed sessions, excluding today.	Does the activity threshold add information for this particular feed?
Combined	Require all three conditions at the same signal close.	Does the smaller intersection retain enough observations to evaluate?

Measure success rate with sample and cost controls

Define success as a completed trade with positive net profit after the stated costs. Win rate is net winners divided by all closed trades; keep flat trades in the denominator. Also report average net profit per trade, drawdown and trade count. A filter can raise win rate while reducing average profit or leaving too little data to judge.

For the hypothetical fixed outcomes above, the arithmetic break-even win rate is 1.30 / (2.48 + 1.30), about 34.4%. That is not an estimated Turtle Soup success rate. Time exits, gaps and varying costs invalidate this two-outcome shortcut.
Daily OHLC does not reveal every intraday price sequence. TradingView documents its broker-emulator fill assumptions and historical intrabar options; check those settings before treating a simulated stop or target as executable.
A different symbol or a 20-bar intraday window is a new experiment. Twenty intraday bars do not represent twenty daily sessions.

Validation	Freeze or record	Required comparison
Sample identity	Symbol, exchange, session, timeframe, data dates, adjustments and script version.	Same data availability and warm-up start across all five runs.
Selection sample	An earlier period used for rule and threshold decisions.	Keep every attempted setting and its number of closed trades.
Untouched sample	A later period reserved before examining outcomes.	Run frozen baseline and each filter once; report their results separately.
Costs	Commission units, spread treatment and slippage per fill.	Compare the intended cost scenario with a predeclared higher-cost scenario.
Fill sensitivity	Order activation, gaps and bars touching both stop and target.	Inspect ambiguous trades with finer price detail; retain the assumed sequence in the record.
Coverage	Accepted signals, filter rejections and missing-data exclusions.	Show how much of the baseline sample each filter retains.
Uncertainty	Net wins, losses, flats and open trades at the cutoff.	Report sample size beside win rate; a few wins do not establish a stable rate.

Turn the chosen version into a reviewable strategy

Give Pine Script AI Coding Agent the daily-close specification, tie rule, skip conditions, exit timing and filter table. Ask for separate filter switches and a timestamped signal log. Review the generated source and verify several accepted and rejected examples against the written rules before reading performance metrics.

Use Strategy Optimizer to compare exposed input settings over a small range chosen in advance. Keep the later evaluation sample separate and inspect sensitivity around a setting instead of selecting only the highest win rate.
Save the script, settings, costs, signal log and complete results for every run. A passing syntax check proves parsing, not faithful rules or profitable trading.

Where Pineify fits

Pine Script AI Coding Agent

Specify the Turtle Soup version and ask for a Pine Script strategy draft with separate filters. Review the source, order timing and signal log before evaluating results.

Also useful: Strategy Optimizer. Compare exposed parameters and filter sensitivity. A parameter search does not establish a universal success rate or guarantee future returns.

Draft the specified strategy

Related resources

TradingView strategy guides Audit backtest accuracy Separate backtest date ranges

Frequently asked questions

Educational strategy research, not investment advice. The custom rules and examples are illustrative and have not been performance-tested. Trading can lose money; historical results and parameter searches do not guarantee future results.

Sources and verification

Last verified: 2026-10-17

Published by Pineify. Pineify is an independent product and is not affiliated with TradingView. Sources support the technical explanations; examples do not establish trading performance.