Academics found (and disputed) a lunar effect in stock returns for decades — Yuan, Zheng & Zhu's 2006 paper is the famous one. Instead of arguing, we computed it: every S&P 500 close (2022–2025) and every Bitcoin close (2014–2026), split by whether the moon was waxing (new → full) or waning (full → new). Side by side below.
Method: half-cycle return = close(last trading day before the next marker) ÷ close(first trading day at/after the marker) − 1, using the bundled S&P 500 (2022–2025) and Bitcoin (2014–2026) daily-close snapshots. Full-moon / new-moon dates from a mean-synodic engine (±1 day).