Skip to main content
Most investing claims are tested on the same data that suggested them, which makes them look better than they are. Insights does it the other way round.

The method

1

Write the rules down first

Before any results are seen, each study fixes the trading rule, the shares it applies to, the time periods and the pass mark, and commits them with a date. The rules post is published first.
2

Develop on one period

The idea is tried on a development period (for example 2005–2016). This is where it’s allowed to be explored.
3

Test once on years it hasn't seen

A final check runs once, on a later period kept aside from the start (a “holdout”). No adjusting afterwards.
4

Count the costs

Brokerage and the gap between buying and selling prices are subtracted, so the result is what a private investor could actually have earned.
5

Publish either way

The result goes out whether the idea passed or failed. The original rules post is kept, unchanged, above the results.

The data

  • Every ASX-listed common stock, including companies that later delisted, so results aren’t flattered by only looking at survivors.
  • Shares need enough daily trading (a liquidity floor) so the results could be traded in practice.
  • Returns are compared with the market as a whole.

Reading a study

Things to keep in mind

  • “No edge found” means the rule didn’t beat the market after costs in the years tested. It doesn’t prove the idea can never work.
  • A study tests the exact rule written down. A different holding period or filter is a different idea.
  • Past results, even honest ones, don’t guarantee future returns.