The Black-Scholes model assumes a single constant volatility for every option on an underlying, which would make implied volatility (IV) flat across all strikes and expirations. In real markets it is not. When you plot the IV backed out of traded option prices against strike price for one expiration, you get a curve — a U-shaped smile, a downward-sloping skew (also called a smirk or "reverse skew"), or some asymmetric blend. The core tension is that this shape is the market's way of pricing in a return distribution that has fatter tails and is more negatively skewed than the lognormal one Black-Scholes assumes; the smile is the visible fingerprint of the model being wrong about the tails.

How it's formed

Implied volatility is the volatility input that makes the Black-Scholes price equal the observed market price. Because every option on the same underlying and expiry shares the same true forward distribution, any variation in IV across strikes is purely an artifact of forcing a lognormal model onto a non-lognormal world.

• Smile (symmetric U): OTM puts and OTM calls both carry higher IV than at-the-money (ATM). This reflects excess kurtosis — fat tails on both sides. Common in currency options and short-dated equity options.
• Reverse skew / smirk (equity, index): IV is highest at low strikes (OTM puts) and falls as strike rises. The downward slope reflects a left-skewed distribution — the market prices a larger probability of a big drop than of a big rally.
• Forward skew: IV rises with strike (OTM calls richest); seen in some commodities where supply shocks drive prices up violently.

Mapped across both strike and time-to-maturity, the full picture is the three-dimensional implied volatility surface (IV vs. strike vs. expiry). The skew typically flattens at longer maturities — the per-strike differences are largest in short-dated options. Practitioners usually parameterize a single expiry's slice with three numbers around delta: the ATM vol, the risk reversal (25-delta call IV minus 25-delta put IV, capturing the slope/asymmetry), and the butterfly/fly (average of the wings minus ATM, capturing curvature) (Wikipedia; [Natenberg via Wikipedia]).

How it's used in practice

• Pricing & hedging consistency: Dealers fit a surface so that all options price off one coherent model. Local-volatility (Dupire) and stochastic-volatility (e.g. Heston, SABR) models exist specifically to reproduce the observed smile instead of contradicting it.
• Reading sentiment: The 25-delta risk reversal is the standard one-number gauge of which tail the market fears. A negative equity risk reversal (puts bid over calls) is the normal "crash fear" state; an unusually steep or unusually flat skew can signal extremes (CBOE; CME CVOL).
• Tail-risk index: The CBOE SKEW Index translates the steepness of S&P 500 OTM-put pricing into a measure of perceived 30-day tail risk; ~100 implies a near-normal distribution, higher values imply richer tails (CBOE SKEW).
• Relative-value & structure selection: Because the OTM-put wing is "expensive," premium sellers often prefer to sell that rich downside vol (put spreads, ratio spreads). Skew also dictates the economics of multi-leg structures: it is precisely why a put spread, a risk reversal, or a collar (buy a richer-IV OTM put, sell a cheaper-IV OTM call) prices the way it does, and why those costs change as the skew steepens or flattens.

Adoption, debate & evidence

The smile is a universal, non-controversial empirical fact — it is taught in the CFA and FRM curricula and built into every dealer pricing system. What is genuinely debated is its cause and its predictive value.

• Why equities skew: The popular "crashophobia" story is Mark Rubinstein's: index options showed essentially no skew before October 1987 and a persistent steep skew afterward, attributed to fear of another crash and demand for downside protection (Wikipedia; analyst notes on Hull). A second explanation is the leverage effect (Black 1976): falling prices raise a firm's debt-to-equity ratio, raising volatility, producing negative return-volatility correlation. The honest state of the literature is that both are real but neither fully explains the magnitude — Black himself, and later Christie (1982) and Schwert (1989), found the volatility response too large to be leverage alone, and several recent studies conclude the negative return-vol correlation is "not due to leverage" in any mechanical sense; its economic source remains partly unresolved.
• Does skew predict returns? Some peer-reviewed cross-sectional work (e.g. Xing, Zhang & Zhao, 2010, JFQA; the broader "option-implied volatility spread" literature) finds steeper firm-level put skew predicts lower subsequent equity returns — interpreted as informed traders acting in options first. This is a measured edge, but it is modest, regime-dependent, and not a standalone retail signal. The smile's primary value is pricing/sentiment, not directional forecasting — do not overstate it.

Strengths & limitations

Strengths: The smile is the most honest single object in options — it directly reveals the market's risk-neutral distribution, including its fear of tails, in a way spot price cannot. Risk reversals and the SKEW index give clean, comparable sentiment reads.

Limitations & the #1 misuse: Risk-neutral skew is not a real-world probability forecast. It is inflated by a variance/tail risk premium — investors pay up for protection, so the implied left tail is systematically fatter than realized outcomes warrant. Treating a high SKEW reading as "a crash is coming" is the classic error; SKEW has many false positives and poor timing as a standalone signal. A second trap is assuming the surface is static — it shifts (sticky-strike vs. sticky-delta dynamics) as spot moves, which matters for any delta-hedged book.

Sources

• CBOE — The Power of the Risk Reversal: https://www.cboe.com/insights/posts/the-power-of-the-risk-reversal/
• CBOE SKEW Index overview: https://www.tradingview.com/symbols/CBOE-SKEW/
• CME Group — Introduction to CVOL Skew: https://www.cmegroup.com/education/courses/introduction-to-cvol/introduction-to-cvol-skew
• Wikipedia — Volatility smile (Hull 2003, Natenberg 2015, Rubinstein/crashophobia, sticky-strike vs sticky-delta): https://en.wikipedia.org/wiki/Volatility_smile
• AnalystPrep — CFA L3 Volatility Skew and Smile: https://analystprep.com/study-notes/cfa-level-iii/volatility-skew-and-smile/
• Leverage-effect literature (Black 1976; Christie 1982; Schwert 1989; "Black's leverage effect is not due to leverage"): https://www.researchgate.net/publication/228252913
• Xing, Zhang & Zhao (2010), What Does Individual Option Volatility Smirk Tell Us About Future Equity Returns?, JFQA; option-implied volatility-spread return-prediction literature.

Disputes flagged: (1) cause of equity skew — crashophobia/demand vs. leverage effect — unresolved, magnitude unexplained by leverage alone; (2) skew's return-predictive power is real but modest/contested as a tradable standalone signal.