Most ideas do not survive the test. These are the ones from the first research runs that lost to the index, or added risk without adding much else.
Hypothetical backtest, not live results
Hypothetical backtest, not live results. These figures come from testing rules on past prices. Nobody invested this money, and past performance does not guarantee future results. Stocks Not Stuff is research and education, not financial advice: nothing here is a recommendation to buy or sell anything.
The question
Before the lab had strategies, it had a longer list of ideas. Most of them did not make it. Showing only the ones that worked would give a false picture of how often a sensible-sounding idea beats the index.
So which ideas failed, and by how much?
How we tested it
These figures come from the original research runs of 13 Oct 2026, done before the site's own test engine was written, on a public set of S&P 500 price data. They are not from the site's own stored backtest, and small differences between the two are normal.
The test had the same shape: $100 every trading day for ten years, from October 2016 to October 2026, bought at the next day's opening price, with a small cost on every trade. The same money in the S&P 500 made 16.0% a year.
The result
Yearly return over the ten years, research runs
Mixed score, buy and hold▼ 14.2% a year
Mixed score, trimming big holdings▼ 14.3% a year
Mixed score, no breakout bonus▼ 12.1% a year
Mixed score, selling at double▼ 11.7% a year
Steady stocks only▼ 9.4% a year
The S&P 500 over the same ten years: 16.0% a year
From the original research runs of 13 Oct 2026, not the site’s stored backtest. Hypothetical, not live results.
A mixed score. The first idea blended four things into one score: the past year's return, how steady the price had been, how far it was below its high, and a bonus for breaking out of a dip. It made 14.2% a year and was behind the index on 89.2% of days. Without the breakout bonus it made 12.1%.
Taking profits. Selling any stock once it had doubled cut that to 11.7%. Trimming any holding that grew past 10% of the portfolio gave 14.3%, which is no real change.
Steady stocks only. Buying the calmest stocks that had still risen over the year made 9.4%, the lowest of any rule that picks from the whole index.
Looser limits. Letting one stock reach 20% of the portfolio and one sector 40% lifted a plain past-winners rule from 24.1% to 27.5% a year. Its worst drop went from −40.8% to −45.2%. More return, bought with more eggs in one basket. We kept 10% and 30%.
One sector at a time. We ran the past-winners rule inside each sector by itself. Only three of the 11 sector portfolios beat the index (Energy, Industrials and Information Technology), and the Energy one fell 69.9% at its worst.
What it means
In these runs, adding parts did not help. The blended score made less than two of its own ingredients did when used alone: past winners at 24.1% a year, and stocks in a dip at 21.9%.
Rules for selling either hurt or made no difference. That is why every strategy in the lab buys and holds.
What it does not prove
Dropped means dropped in this form. The mixed score was tested on prices alone, because company financials were not in the research data. A version that uses them has not been tested yet.
The sector runs use today's sector labels for past years. A company that changed sector along the way is counted under its current one, so those runs are rougher than the rest.
One decade. A period that was hard on fast-rising stocks could rank these ideas differently.
Nobody invested this money. These are tests on past prices, not live results.