Claim check: Turtle breakouts
“Buy 20-day or 55-day breakouts, the turtle way.”
What the data says
Partly Partly: the rule gave a higher Sharpe ratio than buy-and-hold in 5 of 10 markets over 2010-01-29 to 2026-11-10 and in both halves of the history in 2; the median CAGR was +3.0% against +4.5% for buy-and-hold and the median maximum drawdown -44.8% against -71.2% (before costs). It holds in some markets, not across the board.
It holds in some markets or periods, leans the right way without proof, or helps on risk but not on return.
Who says it
Richard Dennis and Bill Eckhardt's Turtle experiment (1983).
Origins are as commonly cited and were not independently verified.
How we tested it
Long from the day after the close exceeds the highest close of the previous 20 days; cash from the day after the close falls below the lowest close of the previous 10 days. The claim holds if the Sharpe ratio beats buy-and-hold in the full history and in both halves on most markets. The original Turtle system used intraday highs and lows, so this is a close-based approximation.
Results
| Market | Period | CAGR | Buy and hold | Sharpe | Buy and hold | Max drawdown | Buy and hold | In market | Trades | Reading |
|---|---|---|---|---|---|---|---|---|---|---|
| S&P 500 | 2010-02-02 → 2026-11-10 | +7.0% | +14.1% | 0.83 | 0.86 | -12% | -34% | 55% | 102 | No effect |
| Gold | 2010-02-02 → 2026-11-10 | +3.2% | +7.5% | 0.34 | 0.52 | -26% | -46% | 40% | 97 | No effect |
| Silver | 2010-02-02 → 2026-11-10 | +4.4% | +6.4% | 0.31 | 0.36 | -49% | -76% | 38% | 94 | No effect |
| Oil | 2010-02-02 → 2026-11-10 | +2.8% | -4.4% | 0.24 | 0.06 | -50% | -95% | 39% | 89 | Clear and stable |
| Natural gas | 2010-02-02 → 2026-11-10 | -13.4% | -24.9% | -0.35 | -0.34 | -91% | -99% | 30% | 92 | No effect |
| Commodities | 2010-02-02 → 2026-11-10 | +2.3% | +2.6% | 0.26 | 0.23 | -40% | -66% | 41% | 96 | Full history only |
| Agriculture | 2010-02-02 → 2026-11-10 | +1.0% | +1.0% | 0.17 | 0.15 | -20% | -62% | 37% | 92 | Full history only |
| EUR/USD | 2010-01-29 → 2026-11-10 | -0.8% | -1.3% | -0.15 | -0.10 | -28% | -35% | 33% | 94 | No effect |
| Bitcoin | 2014-10-07 → 2026-11-10 | +56.6% | +56.2% | 1.08 | 0.84 | -59% | -83% | 42% | 84 | Clear and stable |
| Ether | 2017-11-29 → 2026-11-10 | +51.9% | +19.1% | 0.87 | 0.53 | -52% | -94% | 38% | 65 | Full history only |
Sharpe ratio minus buy-and-hold
Robustness: other settings
| Setting | Verdict on its own | Markets with a stable win | Markets where it won |
|---|---|---|---|
| Main setting | Partly | 2 / 10 | 5 / 10 |
| 55-day entry, 20-day exit | Partly | 4 / 10 | 7 / 10 |
Stability across the two halves of history
A saying that only works in one half of the history is not much of a rule. Each row shows the effect in the full history, the first half and the second half; green means the claimed effect was present.
| Market | Full history | First half | Second half |
|---|---|---|---|
| S&P 500 | -0.03 2010-02-02 → 2026-11-10 | -0.24 2010-02-02 → 2018-06-20 | +0.14 2018-06-20 → 2026-11-10 |
| Gold | -0.18 2010-02-02 → 2026-11-10 | +0.07 2010-02-02 → 2018-06-20 | -0.42 2018-06-20 → 2026-11-10 |
| Silver | -0.05 2010-02-02 → 2026-11-10 | +0.28 2010-02-02 → 2018-06-20 | -0.29 2018-06-20 → 2026-11-10 |
| Oil | +0.18 2010-02-02 → 2026-11-10 | +0.08 2010-02-02 → 2018-06-20 | +0.23 2018-06-20 → 2026-11-10 |
| Natural gas | -0.01 2010-02-02 → 2026-11-10 | -0.02 2010-02-02 → 2018-06-20 | -0.03 2018-06-20 → 2026-11-10 |
| Commodities | +0.03 2010-02-02 → 2026-11-10 | +0.24 2010-02-02 → 2018-06-20 | -0.16 2018-06-20 → 2026-11-10 |
| Agriculture | +0.02 2010-02-02 → 2026-11-10 | +0.29 2010-02-02 → 2018-06-20 | -0.24 2018-06-20 → 2026-11-10 |
| EUR/USD | -0.05 2010-01-29 → 2026-11-10 | +0.11 2010-01-29 → 2018-06-20 | -0.26 2018-06-20 → 2026-11-10 |
| Bitcoin | +0.24 2014-10-07 → 2026-11-10 | +0.42 2014-10-07 → 2020-10-24 | +0.03 2020-10-24 → 2026-11-10 |
| Ether | +0.35 2017-11-29 → 2026-11-10 | +0.62 2017-11-29 → 2022-05-21 | -0.08 2022-05-21 → 2026-11-10 |
Context and cross-checks
Shown for context; not part of the verdict.
The site's own strategies
| Strategy | Market | Sharpe, 5 years | Benchmark, 5 years | Sharpe, full | Benchmark, full | CAGR, full | Benchmark, full | Max DD, full | Benchmark, full | Reading |
|---|---|---|---|---|---|---|---|---|---|---|
| TURTLE-01 Turtle Trading | SPY | 0.62 | 0.72 | 0.78 | 0.85 | +6.7% | +13.8% | -13% | -34% | No effect |
| TURTLE-01 Turtle Trading | GLD | 0.54 | 0.89 | 0.39 | 0.51 | +3.8% | +7.5% | -23% | -46% | No effect |
| TURTLE-01 Turtle Trading | USO | 0.23 | 0.68 | 0.23 | 0.05 | +2.7% | -4.8% | -52% | -95% | Full history only |
| TURTLE-01 Turtle Trading | EUR/USD | 0.10 | -0.05 | -0.23 | -0.13 | -1.3% | -1.5% | -36% | -35% | No effect |
| TURTLE-01 Turtle Trading | DBC | 0.27 | 0.63 | 0.31 | 0.21 | +2.7% | +2.2% | -32% | -66% | Full history only |
| TURTLE-01 Turtle Trading | SLV | 0.11 | 0.57 | 0.32 | 0.35 | +4.5% | +6.1% | -49% | -76% | No effect |
| TURTLE-01 Turtle Trading | DBA | 0.23 | 0.68 | 0.11 | 0.13 | +0.5% | +0.8% | -25% | -62% | No effect |
| TURTLE-01 Turtle Trading | UNG | -0.42 | -0.27 | -0.34 | -0.35 | -13.6% | -25.1% | -92% | -99% | Full history only |
| TURTLE-01 Turtle Trading | BTC | 0.32 | 0.25 | 1.05 | 0.80 | +54.1% | +52.0% | -63% | -83% | Clear and stable |
| TURTLE-01 Turtle Trading | ETH | 0.06 | 0.09 | 0.79 | 0.56 | +43.9% | +22.9% | -52% | -94% | Full history only |
| TREND-02 Donchian Breakout | GLD | 0.72 | 0.89 | 0.46 | 0.51 | +4.6% | +7.5% | -27% | -46% | No effect |
| TREND-02 Donchian Breakout | SPY | 0.45 | 0.72 | 0.55 | 0.85 | +4.4% | +13.8% | -18% | -34% | No effect |
| TREND-02 Donchian Breakout | USO | 0.66 | 0.68 | 0.32 | 0.05 | +4.3% | -4.8% | -36% | -95% | Full history only |
| TREND-02 Donchian Breakout | EUR/USD | 0.01 | -0.05 | -0.19 | -0.13 | -0.9% | -1.5% | -24% | -35% | No effect |
| TREND-02 Donchian Breakout | DBC | 0.79 | 0.63 | 0.40 | 0.21 | +3.6% | +2.2% | -29% | -66% | Clear and stable |
| TREND-02 Donchian Breakout | SLV | 0.57 | 0.57 | 0.38 | 0.35 | +5.8% | +6.1% | -45% | -76% | Full history only |
| TREND-02 Donchian Breakout | DBA | 0.36 | 0.68 | 0.36 | 0.13 | +2.4% | +0.8% | -18% | -62% | Full history only |
| TREND-02 Donchian Breakout | UNG | -0.16 | -0.27 | -0.19 | -0.35 | -7.9% | -25.1% | -77% | -99% | Clear and stable |
| TREND-02 Donchian Breakout | BTC | 0.37 | 0.25 | 0.98 | 0.80 | +48.3% | +52.0% | -57% | -83% | Clear and stable |
| TREND-02 Donchian Breakout | ETH | 0.30 | 0.09 | 0.70 | 0.56 | +32.8% | +22.9% | -56% | -94% | Clear and stable |
Sharpe ratio minus benchmark, full history · Benchmark = the buy-and-hold shown on the strategy pages.
The honest caveat
Uses closes, not intraday highs/lows as the original system did.
Descriptive statistics from historical data, before costs. Not a forecast and not investment advice.
Questions
Buy 20-day or 55-day breakouts, the turtle way.
Partly: the rule gave a higher Sharpe ratio than buy-and-hold in 5 of 10 markets over 2010-01-29 to 2026-11-10 and in both halves of the history in 2; the median CAGR was +3.0% against +4.5% for buy-and-hold and the median maximum drawdown -44.8% against -71.2% (before costs). It holds in some markets, not across the board.
How was this claim tested?
Long from the day after the close exceeds the highest close of the previous 20 days; cash from the day after the close falls below the lowest close of the previous 10 days. The claim holds if the Sharpe ratio beats buy-and-hold in the full history and in both halves on most markets. The original Turtle system used intraday highs and lows, so this is a close-based approximation.
Related strategies on Quant Arena
Markets used
S&P 500 · Gold · Silver · Oil · Natural gas · Commodities · Agriculture · EUR/USD · Bitcoin · Ether