Volatility skew is the pattern of implied volatility across strikes at one expiration. Equity options normally price downside puts above upside calls, a persistent asymmetry born of crash demand; the information is in deviations from the usual shape, which research links to the direction of subsequent moves.
Two persistent flows: institutions buy downside puts as portfolio insurance regardless of price, and covered-call programs sell upside calls regardless of price. Both press the same direction, richening puts and cheapening calls, and the 1987 crash taught option markets that equity downside is where the jumps live. The result is the standard equity smile: a steep put wing, a shallower call wing, with the at-the-money strike near the bottom.
Deviation from a stock's own normal skew carries information. A put wing steepening beyond its usual slope means someone is paying up for downside protection faster than usual; research on skew and subsequent returns finds that pattern precedes weakness more often than chance. The disciplined reading of that fact is asymmetric: a steep put smile argues against selling bullish structures into it, but steep skew is not separately harvestable premium to sell, because the shape predicts direction. That is why a careful conviction adjustment lets skew subtract from a score and never add, the convention Options Scanner adopted after reviewing the research.
Twice, mechanically. In strike selection, the smile's richest strike near a target delta collects extra premium for the same risk profile, so a skew-aware search sells the smile's high point rather than the default strike. In probability math, each strike's own implied volatility, not the at-the-money number, prices the odds at that strike, and the smile's local slope corrects probability-of-profit estimates further. Ignoring skew in either place quietly misprices the trade the model claims to understand.