Implied Volatility and Volatility Smile › Why Do Smiles and Skews Exist?

The volatility smile reveals gaps in the Black-Scholes model's assumptions about market dynamics. The model assumes log-returns follow a normal distribution, but as we discussed in Part III, Chapter 1 on stylized facts of financial returns, actual returns exhibit fat tails (leptokurtosis), skewness, and volatility clustering. These deviations from normality explain the smile and skew patterns.

Real return distributions have fatter tails than the normal distribution. This phenomenon, known as leptokurtosis, means that extreme moves occur more frequently than Gaussian models predict. The normal distribution assumes returns decay exponentially in the tails. Empirical evidence shows financial returns decay more slowly, following a power law in many cases.

To price this tail risk correctly, the market demands higher premiums for out-of-the-money options. These options only pay off when extreme moves occur. Since such moves happen more frequently than the Black-Scholes model assumes, this translates to higher implied volatilities for strikes far from the current price. The smile pattern emerges naturally as the market's acknowledgment of fat tails within the Black-Scholes framework.

Consider a 3-standard-deviation move. Under a normal distribution, this should occur about 0.3% of the time (roughly once every 300 days). In reality, such moves happen several times per year in equity markets. Traders learned this lesson painfully: selling cheap out-of-the-money options that expired in-the-money during crashes led to massive losses. The survivors now price tail-risk options with appropriate premiums.

Equity returns are negatively skewed. Large down moves are more common than large up moves. This asymmetry in the return distribution justifies the equity skew pattern: out-of-the-money puts (which pay off in crashes) should be more expensive than out-of-the-money calls (which pay off in rallies), implying higher volatility for lower strikes.

The asymmetry arises from leverage effects and market dynamics. When stock prices fall, companies become more leveraged (debt-to-equity ratios rise), which increases their risk and volatility. Additionally, market panics create selling cascades and liquidity evaporation that amplify downward moves. These mechanisms are largely absent for upward moves. Markets tend to climb gradually but fall quickly, leading to the familiar pattern of slow, steady bull markets that are punctuated by sharp, sudden corrections.

The Black-Scholes model assumes volatility is constant, but we observe volatility changing over time and correlating with the underlying asset's price. When volatility is uncertain, option prices should reflect this additional risk. Models that incorporate stochastic volatility, which we'll explore in Chapter 18 on modeling volatility, naturally produce smile patterns even when the underlying asset follows a more general process.

The intuition is straightforward: if volatility can change, then an out-of-the-money option has two ways to end up in the money. Either the underlying can move directly there, or volatility can increase first (making larger moves more likely) and then the underlying can move. This second pathway adds value to out-of-the-money options, beyond what constant-volatility models predict. The possibility that volatility might spike before expiration creates additional value for options that would benefit from such an increase, and this additional value manifests as higher implied volatility.

Asset prices occasionally jump discontinuously. Earnings announcements, mergers, regulatory decisions, and market crashes can cause instant price changes that dwarf typical daily moves. These jumps represent information arriving suddenly rather than gradually, and they cannot be captured by continuous Brownian motion.

The Black-Scholes model with continuous Brownian motion cannot capture this behavior. Jump-diffusion models, where prices follow Brownian motion punctuated by occasional random jumps, generate smile patterns because the jumps add probability mass to the tails of the return distribution. The market prices options as if jumps might occur, even when such events are rare, and this additional risk premium appears as elevated implied volatility for out-of-the-money options.

Beyond pure modeling arguments, market structure matters. Institutional investors systematically buy downside protection (puts) to hedge their equity portfolios. This persistent demand raises put prices and implied volatilities for low strikes. Pension funds, insurance companies, and other fiduciaries have regulatory or policy mandates to protect against large losses, creating a structural bid for puts.

Conversely, many investors sell covered calls against their stock holdings, increasing supply and suppressing implied volatilities for high strikes. Retail investors and yield-seeking funds systematically write calls to generate income, adding to the supply of upside options. These supply-demand imbalances persist over time because they reflect the fundamental needs of different market participants, not temporary trading opportunities.