Introductory Context
"This topic develops the Heston model conceptually -- the economic intuition behind stochastic volatility, the Heston model's structure and parameters, its key outputs (the implied volatility surface it generates), and its practical applications in Indian options markets. The mathematics is introduced at an intuitive level: the goal is to understand what the Heston model does and why, not to derive the pricing formula from first principles (which requires advanced stochastic calculus beyond this curriculum's scope). "
The Economic Intuition for Stochastic Volatility
In Black-Scholes, the underlying's price follows a single random process: dS = μS dt + σS dW, where σ is fixed. This means the size of daily price fluctuations is constant over time -- a 20% annual volatility produces the same daily variation whether the market is calm or panicked. Every options practitioner knows this is wrong: in 2020 (COVID crash), Nifty's daily movements were 3-5 times larger than in 2019 (a calm year). Volatility itself was random.
Stochastic volatility models add a second random process for the instantaneous volatility: dV = κ(θ - V)dt + ξ√V dW_V, where V is the current variance (σ² = volatility squared), κ is the mean reversion speed, θ is the long-run mean variance, ξ is the volatility of volatility (vol-of-vol), and dW_V is a second Wiener process (random shock). The two random processes (price process dS and variance process dV) are correlated with correlation ρ -- when price falls, volatility tends to rise (the leverage effect observed in equity markets: market crashes are accompanied by volatility spikes).
The Heston Model's Five Parameters
κ (kappa) -- Mean reversion speed: how quickly instantaneous variance returns to its long-run mean θ after being shocked. High κ: fast mean reversion (volatility spikes are short-lived). Low κ: slow mean reversion (volatility regimes persist for extended periods). For Nifty: historical calibration suggests κ in the range of 2 to 5 (moderate mean reversion, consistent with the 20-40 session volatility clustering observed in Indian markets). θ (theta) -- Long-run mean variance: the level to which variance mean-reverts. √θ is the long-run average volatility. For Nifty: √θ ≈ 14-17% (consistent with the historical long-run India VIX average). ξ (xi) -- Volatility of volatility (vol-of-vol): the random shock to variance at each timestep. High ξ produces a more pronounced volatility smile. Low ξ produces a flatter surface. V₀ -- Initial variance: the current instantaneous variance (√V₀ = current IV for ATM option). ρ (rho) -- Correlation between price and variance processes: negative for equity (when price falls, variance rises). For Nifty: ρ typically -0.5 to -0.8 (strong negative correlation, reflecting the put skew). The ρ parameter is the primary driver of the volatility skew in the Heston model.
What the Heston Model Generates - The Implied Volatility Surface
The Heston model's most important output for practical options pricing: it generates an implied volatility surface (different IVs for different strikes and expiries) from just five parameters -- versus Black-Scholes' single constant σ. The key qualitative features the Heston model captures: (1) Put skew: the negative ρ parameter creates asymmetry -- when prices fall, variance rises (correlation), making OTM puts more expensive than symmetric models predict. The more negative ρ, the more pronounced the skew. (2) Volatility smile: the vol-of-vol parameter ξ generates the 'wings' of the smile -- both deep OTM puts and deep OTM calls have higher IV than ATM options, because the high vol-of-vol increases the probability of extreme moves in either direction. (3) Term structure: the mean reversion parameter κ determines the term structure shape -- high κ means near-term options have higher IV than long-term options (fast mean reversion), creating the term structure inversion seen in crisis periods.
Heston Model Parameter Interpretation
ρ (correlation): primary driver of skew. ρ = -0.70 → strong negative skew (OTM puts expensive vs OTM calls). ρ = 0 → symmetric smile. ξ (vol-of-vol): primary driver of smile curvature. High ξ → pronounced smile/wings. Low ξ → flat smile. κ (mean reversion speed): primary driver of term structure. High κ → fast reversion → inverted term structure in high-vol regimes. Low κ → slow reversion → persistent vol regimes. θ (long-run variance): anchor for long-term IV. The long-dated ATM IV converges to √θ. V₀ (initial variance): sets current ATM IV. Calibration: fitting these 5 parameters to market option prices across strikes and expiries produces the model's implied volatility surface.
Heston vs Black-Scholes for Indian Market Pricing
For standard Nifty vanilla options used in retail income strategies (iron condors, credit spreads): the Heston model's improvements over Black-Scholes are modest for ATM and near-ATM strikes where the practical difference in pricing is small. Where the Heston model provides decisive improvements: (1) Exotic options pricing where path dependency creates sensitivity to the volatility process (Asian options, barrier options). (2) Deep OTM options (5-10% OTM puts) where the skew premium is significant and Black-Scholes consistently underprices. (3) Long-dated options (3-12 month) where the stochastic volatility dynamics produce material differences from constant-volatility pricing. (4) Volatility arbitrage (Topic 25.7) where precise volatility surface modelling is essential for identifying mispricings.
The Heston model is the options world's version of the upgrade from Newtonian to Einsteinian physics: for everyday situations (driving a car, throwing a ball), Newtonian physics works perfectly well. For extreme conditions (near the speed of light, near a black hole), the Newtonian model fails and Einstein's extension becomes necessary. Black-Scholes works perfectly well for standard vanilla options in normal market conditions. For extreme conditions (deep OTM options, exotic options, crisis pricing), the Heston model's stochastic volatility extension becomes necessary. Understanding when the upgrade is needed is as important as knowing the upgrade itself.
Sensibull's IV Percentile and VIX Analysis Use Heston-Like Stochastic Volatility Intuition
The volatility regime analysis in Sensibull's options analytics (showing when current IV is high or low relative to historical distribution) is conceptually grounded in stochastic volatility thinking: it recognises that volatility follows its own process with mean reversion (VIX mean reversion from Topic 18.6) and clustering (GARCH from Topic 18.12). When Sensibull shows 'IV at 15th percentile of 52-week range': it is providing a simplified version of the Heston model's V₀ vs θ comparison -- the current vol is below the long-run mean, suggesting the Heston model would predict vol to rise toward θ (mean reversion). This is directly actionable: low V₀ vs θ in the Heston framework supports long volatility positions.