What is Ito's Lemma? The Calculus of Randomness

Ito's Lemma is the foundation of derivative pricing. Learn how to derive the change in an option's price using stochastic calculus and Taylor expansions.

Ito's Lemma represents the fundamental bridge between deterministic calculus and the stochastic world of finance. In standard calculus, the chain rule is simple; in stochastic calculus, the presence of a random walk term means we must account for second-order effects that usually vanish.

If we have a stochastic process $X_t$ that follows an Ito process: $dX_t = \mu_t dt + \sigma_t dW_t$, and a twice-differentiable function $f(t, X_t)$, Ito's Lemma tells us that the differential $df$ is: $df = \left( \frac{\partial f}{\partial t} + \mu_t \frac{\partial f}{\partial x} + \frac{1}{2} \sigma_t^2 \frac{\partial^2 f}{\partial x^2} \right) dt + \sigma_t \frac{\partial f}{\partial x} dW_t$.

The 'magic' of Ito's Lemma lies in the term $\frac{1}{2} \sigma^2 \frac{\partial^2 f}{\partial x^2}$. This arises because in Brownian motion, the square of the increment $(dW_t)^2$ is equal to $dt$. In deterministic calculus, $(dt)^2$ is too small to matter and is ignored ($0$), but in the stochastic world, the volatility $\sigma$ ensures that the variance contributes to the mean drift of the function.

This is precisely why options are priced using Gamma. Gamma represents the convexity of the option price relative to the underlying. Without Ito's Lemma, we would fail to account for the 'volatility tax' or the 'convexity gain' in a rebalanced portfolio.

To derive the Black-Scholes-Merton model equation, we apply Ito's Lemma to the price of a call option $C(S, t)$. By creating a delta-hedged portfolio consisting of one long option and $\Delta$ short shares, we can eliminate the $dW_t$ (random) term entirely. This transformation turns a stochastic problem into a deterministic partial differential equation.