ABOUT
Built to turn over more rocks.
The investor who turns over the most rocks finds the most.
The bottleneck was never ideas. It was reading.
I built RockTurners because I kept losing the same race against my own reading list.
I'd run a screen and come out with a hundred names that fit. That's where the screener quit and the work began. Every one of those hundred was a 10-K to read, footnotes to reconcile, numbers to check against the narrative. Done honestly, I managed maybe a dozen a year. The other eighty-plus stayed a pile — and the company I was hunting for might have been in it.
Capable AI models changed the math. Not by skipping the reading, but by doing far more of it, the same way every time. It started as a single page I wrote for myself. It became RockTurners.
This is a tool for people who invest in businesses, not tickers. If you're reading a 10-K to understand how a company actually makes money — and you measure your holding period in years, not afternoons — you're who this is for. It's not built for options expiries or the next fifteen-minute move. RockTurners reads filings and applies frameworks built for owners, not traders.
The frameworks that matter in investing are published. Graham wrote his down. So did Lynch, and Fisher, and the academics behind the momentum literature. What none of them could give you is the time to apply their discipline to thousands of companies — reading the filings, checking the numbers against the words, hunting for the reasons a thesis fails.
RockTurners exists to do that reading — identically, impersonally, with every claim cited back to the filing it came from. It doesn't have opinions about what you should do. It has evidence about what the filings say, organized by disciplines that have been public for decades.
The gaps stay visible, the citations stay clickable, and the verdict vocabulary stays honest: a Gemstone is a strong match to a framework — nothing more, and nothing less.
What RockTurners is not
No advice, no ratings on what to do, no knowledge of your circumstances. Research for educational purposes only.
No forecasts, no targets, no claims about future returns. Verdicts describe what filings show today.
Every claim carries its source, every gap is flagged, and every scan shows its cost before it runs.