Where this database becomes research.

• Research, not commentary. Every letter starts from something unusual in our own data — a divergence, a regime change, a chart that looks scary — and does the homework in public: what the full history says, what survives scrutiny, and what doesn't. When the honest answer is “this famous signal doesn't work,” that is the letter.
• Numbers you can check. Every chart is computed from the same public datasets that power this site, date-stamped, with a link to the live version that updates after each close. Nothing is asserted that you can't verify yourself.
• Falsifiable markers. Each letter ends with the specific, watchable conditions that would confirm or kill its thesis — and when a marker moves, the follow-up letter is how you hear about it.
• No fixed schedule, no filler. Letters ship when the data hands us something worth your inbox — typically a few times a month, never on a content calendar.

The site is the instrument panel: every tool updates daily, shows its full history, and explains its methodology — free, no signup. The Market Changelog is the weekly readout: a broad, ranked account of what changed enough to matter after Friday's close, with a frozen evidence receipt and live links.

The Research Letter goes narrow and deep. It starts with one thesis, tests it against history, and names the evidence that would prove it wrong. It ships only when the data earns the space, typically a few times a month. The Changelog keeps the weekly record; the letter makes the larger argument.

Recent letters have covered the Nasdaq's divergence from the S&P 500 across five independent gauges, and why the scariest breadth chart in our database — the Nasdaq's cumulative TICK, which peaked 16 days before the 2021 top — mostly doesn't predict what it looks like it predicts. Browse the full archive on Substack.

Yuriy Matso — the same person who builds and maintains this site, its data pipeline, and its methodology. The letter carries the site's editorial standards: dated claims, primary sources, corrections on the page, and transparent AI usage. It is research and education, never financial advice.