Variations

Before you trust any backtest — including ours — look at this. We took the four most famous strategies, ran them with 9,393 different settings across 101 stocks, then scored every single one on years of market data the search had never seen.

Beat buy & hold on unseen data 21.3%

Only 86 of 404 best-in-class settings stayed ahead of simply buying and holding the stock. That low number is the finding — pick the winner out of a big search and you're usually picking an accident.

9,393settings tested
404best-of finalists
86still won on unseen data
101stocks
4,392days of history

Bollinger bands

Trade the band: buy the lower band, exit at the middle or upper.

26 of 101 beat buy & hold 25.7%

Donchian breakout

Buy an N-day high, sell an N-day low — the classic turtle breakout.

15 of 101 beat buy & hold 14.9%

Moving-average cross

Buy when the fast average crosses above the slow one, sell when it crosses back.

19 of 101 beat buy & hold 18.8%

RSI mean-reversion

Buy when RSI says oversold, sell when it says overbought.

26 of 101 beat buy & hold 25.7%
🏆 Winners only All finalists Bollinger bands Donchian breakout Moving-average cross RSI mean-reversion
StockStrategy & settings Strategy returnBuy & hold DifferenceSharpeWorst drop
V Bollinger bands
period 20 · width 1.5
83% 66% +17 pts 0.86 -14% beat
PG Bollinger bands
period 30 · width 2.0
44% 34% +9 pts 0.80 -14% beat
MA Bollinger bands
period 20 · width 1.5
70% 43% +27 pts 0.75 -18% beat
DE Bollinger bands
period 50 · width 1.5
82% 69% +13 pts 0.75 -24% beat
UNP Bollinger bands
period 20 · width 1.5
57% 38% +20 pts 0.71 -15% beat
VZ Bollinger bands
period 50 · width 1.5
49% 12% +37 pts 0.66 -25% beat
SBUX Bollinger bands
period 20 · width 1.5
61% 6% +55 pts 0.59 -26% beat
O Bollinger bands
period 50 · width 1.5
34% 32% +2 pts 0.59 -19% beat
IWM Bollinger bands
period 30 · width 2.0
40% 36% +4 pts 0.55 -20% beat
HON Bollinger bands
period 30 · width 2.0
31% 27% +4 pts 0.48 -13% beat
BA Bollinger bands
period 20 · width 3.0
13% -12% +25 pts 0.48 -9% beat
D Bollinger bands
period 50 · width 1.5
25% 17% +8 pts 0.38 -31% beat
APD Bollinger bands
period 20 · width 1.5
26% 17% +9 pts 0.35 -28% beat
ABT Bollinger bands
period 50 · width 1.5
22% -19% +41 pts 0.33 -32% beat
BMY Bollinger bands
period 50 · width 1.5
21% 9% +12 pts 0.32 -25% beat
UNH Bollinger bands
period 50 · width 1.5
22% 20% +2 pts 0.29 -47% beat
DIS Bollinger bands
period 30 · width 2.0
20% -46% +66 pts 0.29 -28% beat
CRM Bollinger bands
period 20 · width 1.5
14% -26% +40 pts 0.23 -32% beat
PFE Bollinger bands
period 20 · width 2.0
11% -5% +16 pts 0.22 -24% beat
DHR Bollinger bands
period 50 · width 1.5
4% -7% +11 pts 0.13 -32% beat
AMT Bollinger bands
period 20 · width 1.5
2% -4% +6 pts 0.11 -34% beat
MDT Bollinger bands
period 50 · width 1.5
-1% -22% +21 pts 0.06 -27% beat
PYPL Bollinger bands
period 20 · width 3.0
-1% -43% +42 pts 0.01 -18% beat
CCI Bollinger bands
period 50 · width 1.5
-17% -34% +17 pts -0.11 -40% beat
CMCSA Bollinger bands
period 50 · width 1.5
-20% -49% +29 pts -0.14 -30% beat
NKE Bollinger bands
period 30 · width 1.5
-34% -65% +31 pts -0.22 -45% beat

Showing 26 of 26. Sorted by Sharpe — return per unit of risk taken.

How to read this honestly

Settings were chosen on the older 70% of each stock's history, then scored — untouched — on the newer 30%. Every number on this page comes from that unseen stretch. "Difference" is how far the strategy finished ahead of or behind simply buying the stock and waiting, in percentage points. A strategy that beat buy & hold here is not a recommendation: 101 stocks × four families is a big enough search that some winners are luck. That's exactly why Chartments lets you put an idea on trial going forward, where luck can't be backfilled.

Snapshot from the sweep engine · 101 stocks · 4,392 trading days. Research & education only — not investment advice.