How we test
Every belief on this site gets the same treatment. We write down an exact rule, run it on the longest clean price history we have, charge the costs a real trader would pay, and compare it with the simplest alternative. Then we publish the result, whether it is exciting or not.
1. The belief becomes a rule
We test sayings that traders repeat: seasonal patterns, chart signals, rules of thumb. Each one is turned into an exact rule that says what to buy or sell, when, and when to get out. If a saying can be read in several ways, we test the common reading and say which one we used.
2. The prices
We use daily closing prices of stock indices, funds, currencies, commodities and cryptocurrencies, mostly from 2004 onward, and back to 1970 for the S&P 500. Fund prices include reinvested dividends where we have them. Index prices do not, and every chart says which kind it uses. Before a price series is used we check it for gaps and for jumps that are data errors rather than market moves.
3. The costs
Every trade pays something: the gap between the buying and selling price, and any fee. We charge a cost a small investor could actually get and state it in the article. Taxes differ from country to country and are not counted.
4. A fair comparison
A rule has to beat the simple choice it replaces. For a stock market rule that is usually buying the index and holding it. Making money is not enough. A rule that earns less than doing nothing has failed.
5. Before and after it became famous
Many patterns fade once they are published and traded. Where we know when a belief became widely known, we show the years before and after separately, so you can see whether it still holds today.
6. The verdict
- Edge
- The rule beat the simple alternative after costs, over the whole test and in both halves of it.
- Half true
- The pattern shows up in the prices, but trading on it did not beat the simple alternative after costs, or it worked in one half of the history only.
- Myth
- The pattern is not there, or it is too small to tell apart from chance.
What we do not do
We do not adjust a rule's settings until it looks good, and we do not test a hundred versions and show you the best one. We do not sell trading signals, take money from brokers for results, or tell anyone what to buy. Advertisers and sponsors never see a result before it is published and have no say over what we test.
How the work is done
Software does the number work. It loads the prices, runs the rule and draws the charts, and every figure in an article is filled in from the program's output, so a typing slip cannot change a number. AI tools help us draft the text. A person oversees the site and reviews what it publishes.
Corrections
If you find a mistake, write to [email protected]. We correct the article and note the change and its date at the bottom of the page.
Everything on this site is research on past prices. It is not advice to buy or sell anything, and past results do not promise future ones.