Implied Volatility and Volatility Smile › Limitations and Practical Considerations

Implied volatility is model-dependent. The statement "implied volatility is 25%" means 25% makes the Black-Scholes formula match the market price. Different models like jump-diffusion or stochastic volatility models would extract different parameters. The Black-Scholes implied volatility is a convenient translation of dollar prices into a standardized metric, but it does not necessarily represent the "true" expected volatility.

Model dependency creates practical challenges. Deep out-of-the-money options have unstable implied volatilities because small price errors translate to large volatility errors. Very short-dated options present numerical difficulties because vega approaches zero, making Newton-Raphson unstable. Practitioners often use specialized algorithms, such as Jaeckel's "Let's Be Rational" approach, that remain stable on these edge cases.

Real option quotes come with bid-ask spreads that can be substantial for illiquid strikes. The "true" implied volatility lies somewhere between the bid-implied vol and the ask-implied vol. For liquid at-the-money options, the spread might be 0.5 vol points. For deep out-of-the-money options or those on less liquid underlyings, the spread can exceed 5 vol points. Traders must account for these transaction costs when implementing volatility trading strategies.

The volatility surface is not static. It shifts throughout the day as new information arrives and market participants adjust their views. During major economic announcements (employment reports, Federal Reserve meetings, or earnings releases), the surface can move dramatically in minutes. Risk managers must monitor the current surface, its dynamics, and its likely behavior under various scenarios.

Understanding smile dynamics as spot price moves is an active research area. Does the smile "stick to strike" (constant IV at fixed strike) or "stick to delta" (constant IV at fixed delta)? The answer affects hedging strategies and realized profit and loss as markets move. We'll revisit these dynamics when we discuss stochastic volatility models in Chapter 18.