Introductory Context
"But the Black-Scholes model's foundational assumption -- that the underlying asset's volatility is constant over the option's life -- is empirically false. Volatility is not constant. It varies with the underlying's price level (the volatility smile and smirk from Topic 18.2), with time (volatility regimes from Topic 18.5), and with market conditions (volatility clustering from Topic 18.12). The failure of the constant-volatility assumption produces systematic pricing errors -- options priced with Black-Scholes at the wrong implied volatility are mispriced, and the mispricing follows predictable patterns that sophisticated market participants exploit. Module 24 develops the mathematical extensions to Black-Scholes that address these failures: stochastic volatility models, local volatility models, jump-diffusion models, and their implementation through numerical methods. "
The Five Assumptions Black-Scholes Violates in Practice
Assumption 1 -- Constant volatility. The most critical violation: Black-Scholes assumes σ (volatility) is fixed throughout the option's life. In reality: volatility changes daily, follows clusters (high-vol periods follow high-vol periods), and depends on the underlying's price level (put skew, Topic 18.3). Consequence: at-the-money options are priced reasonably well by Black-Scholes (the ATM IV calibrates the model), but deep OTM puts are consistently underpriced (their actual market prices correspond to higher IVs than the ATM level -- the put skew premium). The entire volatility surface (Topic 18.2) is the visual manifestation of Black-Scholes' constant-volatility failure.
Assumption 2 -- Log-normal returns with no jumps. Black-Scholes assumes the underlying's price follows a continuous lognormal diffusion with no sudden discontinuous jumps. In reality: market crashes (COVID 2020, 2008 Global Financial Crisis, 1987 Crash) produce instantaneous price gaps -- jumps -- that the diffusion model cannot capture. Deep OTM put options command a 'crash premium' because the market correctly prices the non-zero probability of a sudden large downward jump that the Black-Scholes diffusion model assigns zero probability. Merton's jump-diffusion extension (Topic 24.5) addresses this.
Assumption 3 -- No transaction costs or taxes. Black-Scholes assumes frictionless trading -- any position can be continuously rebalanced at zero cost. In reality: bid-ask spreads, brokerage commissions, taxes, and market impact costs make continuous rebalancing impossibly expensive. Practical delta hedging uses discrete rebalancing intervals (daily or weekly), which introduces hedging error that accumulates over the option's life. The model's theoretical continuous-hedging perfection is never achieved in practice.
Assumption 4 -- Constant risk-free rate. Black-Scholes discounts at a single risk-free rate assumed constant over the option's life. In reality: interest rates are themselves stochastic (the interest rate options market, Topic 23.11, exists precisely because rates move unpredictably). For short-dated equity options: this assumption is benign (rate changes over 30 days are small). For long-dated options (1-5 year): stochastic interest rates create meaningful pricing errors that require separate modelling.
Assumption 5 -- Normally distributed log-returns. Black-Scholes uses the normal distribution for log-returns, which has zero probability of extreme events beyond 3 or 4 standard deviations. Real financial returns have 'fat tails' -- extreme events (crashes, rallies) occur far more frequently than the normal distribution predicts. Deep OTM options price the fat tails at higher IVs than the Black-Scholes constant-volatility model implies.
Black-Scholes Failure Map
Failure → Market Consequence → Model Extension. Constant volatility → Volatility smile/smirk → Stochastic Vol (Heston), Local Vol (Dupire). No jumps → Crash premium in deep OTM puts → Jump-diffusion (Merton). Fat tails → OTM options more expensive than BS → Variance Gamma, Lévy processes. No transaction costs → Practical hedging errors → Discrete hedging adjustments. Constant rates → Long-dated pricing errors → Interest rate / equity hybrid models.
The Implied Volatility Surface as a Map of Black-Scholes Failures
The implied volatility surface (Topic 18.2) is the most visible evidence of Black-Scholes' failures. If Black-Scholes were correct (constant volatility assumption), every option on the same underlying with the same expiry would have the same implied volatility, regardless of strike. The actual implied volatility surface shows: (1) The skew -- OTM puts have higher IV than OTM calls at the same distance from ATM (the crash premium and fat left tail). (2) The term structure -- near-term IV differs from long-term IV (volatility regime dynamics). (3) The convexity -- deep OTM options have even higher IV than near-ATM OTM options (the fat tails beyond what a mild skew would predict). Each feature of the surface corresponds to a specific Black-Scholes failure -- and each can be addressed by the model extensions covered in Topics 24.2 through 24.5.
Why Black-Scholes Is Still Used Despite Its Failures
Despite its known failures, Black-Scholes remains the universal market standard because it serves three functions that its successors cannot easily replace: (1) Communication: implied volatility (Black-Scholes' input that equates the model's output to the market price) is the universal language for comparing options prices across different strikes, expiries, underlyings, and market conditions. Traders say 'the 25-delta put is trading at 18 vol' -- a Black-Scholes concept that conveys information instantly to any options professional globally. No advanced model has achieved this communication efficiency. (2) Delta hedging: Black-Scholes' delta is a reasonable first-order hedge ratio for most vanilla options positions, even if it is not perfectly accurate. The more complex delta calculations from advanced models do not justify the computational overhead for routine delta hedging. (3) Intuition: the Black-Scholes framework provides direct intuition about how options respond to changes in the underlying, time, and volatility (the Greeks) -- the foundation of options strategy design and risk management in Modules 1 through 23.
Black-Scholes is a map of the options world that omits certain terrain features -- the mountain ranges of volatility clustering, the chasms of crash risk, the fjords of put skew. The map is still useful for navigation: it covers most of the terrain correctly most of the time. But the sophisticated navigator knows where the map is unreliable -- the deep OTM options, the near-expiry gamma, the pre-event volatility pricing -- and uses supplementary tools (the advanced models of Topics 24.2-24.5) when navigating those specific terrains. Knowing both the map and its limitations is more powerful than either knowing only the map or having only the advanced models.
Black-Scholes Failures Directly Create the Option Seller's Structural Edge
The put skew premium that is the option seller's structural advantage (Topic 17.1) is itself a consequence of Black-Scholes' constant-volatility failure. Because Black-Scholes underprices deep OTM puts (by assigning them too low an IV), the market corrects by pricing OTM puts at a premium to the BS ATM-calibrated IV level. This premium is what option sellers collect and what creates the positive VRP. The seller's structural edge is not despite Black-Scholes' failure -- it partly exists because of it.