About PairBook

PairBook exists because two numbers that belong together almost never appear together: how two assets move (correlation) and what they actually hold (overlap).

It is built and maintained by an independent developer. The site is intentionally simple: static pages, no tracking scripts, one clean dataset, and a methodology page that shows every formula so any figure can be reproduced independently.

The data pipeline recomputes correlations, covariances, betas and fund overlaps from end-of-day prices and issuer portfolio disclosures. Everything on the site is also available through the free JSON API.

The project is open source. The pipeline, the templates and the daily refresh workflow are published in the GitHub repository, so the way every number is produced can be audited line by line. Bug reports and data requests are welcome there too.

Questions and corrections are welcome.