Deeply cheap + strong quality and clean accruals + momentum not collapsing. The cohort Piotroski identified.
Cheap stocks typically split into two populations: cheap-because-broken and cheap-because-temporarily-hated. Piotroski's F-Score (9 accounting checks) separates them; we approximate that screen with our quality and accruals factors rather than computing the literal 9-point score. The pattern fires when a deeply cheap name clears both, with price action not in freefall.
Piotroski 2000 showed the top F-Score cheap cohort outperforms the bottom cohort by 7-10% per year. The effect comes from correcting analyst over-reaction to short-term bad news on businesses that are actually improving.
Small-cap bias in the original study. Works best on stocks under $10B market cap where analyst coverage is thin. Mega-caps with the same factor profile generally have already been arbitraged.