Should you invent a strategy or borrow a tested one?
Borrow first. The classic setups have already been run through walk-forward testing and a 100-run random-entry gauntlet across 28 tickers, so you can start from something with a measured track record instead of an idea you like the sound of. The catch is that several of the most famous setups busted, so borrowing only helps if you can tell the survivors from the folklore.
Based on the same process behind every Library report. How we test.
Breadth is the number to trust
The most tempting number on any report is the biggest return, and it is usually the least useful one. A single spectacular result often means one volatile ticker trended hard during the test window. The number that travels is breadth: how many separate tickers the setup held an edge on. A setup that cleared the bar on many names is describing something about markets. A setup that cleared it on one is describing that name.
Four survivors from four different families
| Setup | Ticker | Return | Sharpe | Win rate | Edge held on |
|---|---|---|---|---|---|
| Williams %R bounce | SPY | +35.5% | 0.69 | 74.2% | 12 / 28 |
| SuperTrend follower | QQQ | +199.8% | 1.11 | 52.5% | 11 / 28 |
| Parabolic SAR flip | SPY | +99.0% | 1.13 | 60.0% | 10 / 28 |
| SMA 20/50 cross | SPY | +103.4% | 0.83 | 61.9% | 9 / 28 |
Each of these held an edge on a meaningful share of the 28 tickers, and they work in different ways: an oversold bounce, two trend followers, and a moving average cross. Return shown is a representative index, not the best single ticker. Breadth is the column that matters.
Two famous setups that did not survive
This is why borrowing requires reading rather than trusting. Both of the setups below are staples of trading education, and both failed the same test the survivors passed. The receipts are pulled live from the reports:
Best showing 95 of 100 vs random on HOOD, but only 6 trades in a decade. A high score on too few signals is a coin flip, not an edge, so it clears the bar on no ticker we ran it on.
Best showing 88 of 100 vs random on MSTR, but only 3 trades in a decade and just 2 of 4 periods positive. A high score on too few signals is a coin flip, not an edge, so it clears the bar on no ticker we ran it on.
How to pick a starting point you will actually stick with
Fame is not evidence, and neither is a big return on one ticker. Filter the shelf by breadth first, then choose among the survivors by what suits you: how often it trades, how long it holds, and whether you can watch it on your schedule. A slightly weaker setup you can execute consistently beats a stronger one that needs attention you do not have. Then test your own version, because the shelf tests the rules as commonly stated and your version will differ.
How to use this
- Sort by breadth, not by return. How many tickers held an edge tells you more than the single best result.
- Check the busted list before you adopt a famous setup. Several of the best known ones failed on trade count.
- Pick a survivor whose trade frequency and hold time match your actual schedule, not your ideal one.
- Read the rules as stated, then write down your own version explicitly. Small parameter differences change results.
- Treat the borrowed setup as a starting point to test, not a finished system to trade.
Common questions
Is it better to use a proven strategy or create your own?
Start from a tested one. The classic setups already have walk-forward results across many tickers, so you begin with a measured baseline instead of an untested idea. You can adapt it afterwards, and then test the adaptation.
Do classic trading setups actually work?
Some do and several famous ones do not. In our testing, setups that fire often enough to build a real sample tended to hold an edge across multiple tickers, while several celebrated signals cleared the bar on none.
What makes a backtested setup trustworthy?
Breadth and trade count. A setup that held an edge on many separate tickers, with enough trades to measure, is far more trustworthy than one enormous return on a single volatile name.
Should I change a tested setup's parameters?
You can, but then it is your setup and it needs its own test. The published results describe the rules as commonly stated, and small parameter changes can move the outcome more than most traders expect.
Every setup on the shelf has a full public report: the rules, the walk-forward result, how it did against random timing, and how many tickers it held up on.