Turtle Trading
The classic 1980s Turtle trend-following rules, run on 8 ETFs (SPY, QQQ, IWM, EFA, TLT, GLD, SLV, USO) with two $10,000 paper portfolios. Each buys breakouts to new highs, sizes every position by the ETF's own volatility (N), adds up to 4 units as the trend runs, and exits on a 2N stop or a break to new lows. Long only. Signals come from the close; trades fill at the next day's open.
The chart fills in as each trading day closes.
- Buy close above the prior 20-day high
- Sell close below the prior 10-day low, or the 2N stop
- Risk 1% of the account per unit
- Filter skip a breakout if the last one in that ETF was a winner (unless it is also a 55-day breakout)
No positions — $10,000.00 in cash, waiting for a breakout.
- Buy close above the prior 55-day high
- Sell close below the prior 20-day low, or the 2N stop
- Risk 0.5% of the account per unit
No positions — $10,000.00 in cash, waiting for a breakout.
| $10,000 grew to | Per year | Worst drop | Win rate | Profit factor | Trades | |
|---|---|---|---|---|---|---|
| Turtle System 1 | $58,058 | +17.7% | -40.0% | 36% | 1.71 | 372 |
| Turtle System 2 | $67,258 | +19.3% | -24.4% | 34% | 2.34 | 226 |
| SPY buy & hold | $44,703 | +14.8% | -33.7% | — | — | — |
Turtle systems win only about a third of their trades — the money comes from a few long trends. Backtest fills at the next open, uses up to 2x buying power and does not charge margin interest, which would lower the yearly return. Past results don't guarantee future ones; the live portfolios above are the real test.
No trades yet. Buys happen the morning after an ETF closes at a new breakout high.
