The Martingale Representation Theorem

In a world driven by a single martingale source — a Brownian motion W — there is only one way to be a martingale. That is the content of the martingale representation theorem: every martingale adapted to the Brownian filtration is, secretly, an Itô integral against W.

Precisely: if (\mathcal{F}_t) is the filtration generated by a Brownian motion W, and (M_t) is any martingale adapted to it (square-integrable), then there is a unique adapted process (H_t) with

M_t = M_0 + \int_0^t H_s\, dW_s.

No drift term, no other noise source — just a single stochastic integral. The Brownian motion is the only raw material, and H is the recipe for how much of it to use at each instant.

From representation to a hedge, line by line

This abstract fact is the secret engine of replication — the reason every option in a complete market can be manufactured by trading. Follow the logical chain.

Step 1 — the discounted price is a \mathbb{Q}-martingale. Work under the risk-neutral measure \mathbb{Q}. The discounted value of a claim paying H_T is, by the pricing rule,

\tilde{V}_t = \mathbb{E}_{\mathbb{Q}}\big[e^{-rT} H_T \mid \mathcal{F}_t\big],

and a conditional expectation of a fixed random variable is automatically a martingale (the tower property: \mathbb{E}_{\mathbb{Q}}[\tilde{V}_t \mid \mathcal{F}_s] = \tilde{V}_s).

Step 2 — represent it as an integral against the \mathbb{Q}-Brownian motion. By Girsanov the discounted price is driven by a \mathbb{Q}-Brownian motion \tilde{W}; the filtration is Brownian, so the representation theorem applies to the \mathbb{Q}-martingale \tilde{V}_t. There exists a unique adapted H_t with

\tilde{V}_t = \tilde{V}_0 + \int_0^t H_s\, d\tilde{W}_s.

Step 3 — match the integrand against a trading strategy. A self-financing portfolio holding \phi_t shares has discounted value moving as d\tilde{V}_t = \phi_t\, d\tilde{S}_t, and the discounted stock satisfies d\tilde{S}_t = \sigma \tilde{S}_t\, d\tilde{W}_t. So the portfolio's increment is

d\tilde{V}_t = \phi_t\,\sigma \tilde{S}_t\, d\tilde{W}_t.

Step 4 — read off the hedge. Comparing the integrand from Step 2 with the portfolio's integrand from Step 3 — both multiply the same d\tilde{W}_t, and the Itô representation is unique — forces them equal, H_t = \phi_t\,\sigma\tilde{S}_t. Solving for the holding,

\boxed{\,\phi_t = \frac{H_t}{\sigma \tilde{S}_t}.\,}

The abstract integrand H_t is the hedge: it tells you exactly how many shares \phi_t to hold at each instant to track the option. The mathematics that guarantees a representation exists is the same mathematics that guarantees a hedging strategy exists — they are one theorem. This holding \phi_t is the option's delta.

Let (\mathcal{F}_t) be the filtration generated by a Brownian motion W, and let (M_t) be a square-integrable (\mathcal{F}_t)-martingale. Then there is a unique adapted process (H_t) (with \mathbb{E}\int_0^T H_s^2\,ds < \infty) such that M_t = M_0 + \int_0^t H_s\, dW_s \qquad \text{for all } t \le T. In a Brownian world, every martingale is an Itô integral against W — and that integrand is the replicating strategy.

The representation theorem is exactly why the Black–Scholes market is complete. The Step-2 integrand exists, so every claim can be replicated; and it is unique, so there is one and only one fair price and one and only one hedge. This dovetails with the second fundamental theorem: completeness ⇔ a unique equivalent martingale measure. One Brownian source, one martingale measure, one hedge — the whole edifice stands on a single noise dimension.

The catch is the hypothesis: the filtration must be Brownian. The theorem fails the moment there is a noise source the stock cannot span — a jump, a second uncorrelated shock, stochastic volatility. Then some martingales are not integrals against W, the integrand (the hedge) may not exist, the market becomes incomplete, and the martingale measure is no longer unique. Markets are complete precisely to the extent that there are as many tradeable assets as sources of risk.

Looking forward, the integrand H_t — equivalently the holding \phi_t = H_t/(\sigma\tilde{S}_t) — is the option's delta, the first and most important of the Greeks. Computing it explicitly for a call option recovers the \Phi(d_1) of the Black–Scholes formula, where delta-hedging becomes a concrete recipe a trader can run.

A martingale as a stochastic integral

The figure draws one Brownian path W together with the martingale M_t = \int_0^t H_s\, dW_s it generates through a (here, simple adapted) integrand H. Notice M carries no drift — it merely accumulates H-weighted Brownian increments and wanders around its start M_0 = 0. Every Brownian-adapted martingale looks like this for some H; the representation theorem says the converse too. Refresh for a fresh \omega.