LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
The Black-Scholes equation is what remains after you hedge a European option against the stock and require the hedged book to earn the riskless rate. The derivation assumes the stock follows geometric Brownian motion, applies Ito's lemma to the option value, and chooses a share holding that cancels the random piece. Solving that equation is a later step. The lecture stops at the partial differential equation whose solution prices European calls and puts inside the model.
Why it matters
The talk is explicit about the assumption. Real markets are not geometric Brownian motion. That objection is aimed at a different article. Here the process is the modeling choice that lets a hedging argument run. If you will not grant the process for the length of the derivation, there is no equation to reach. The portfolio is the device that turns a random stock into a locally riskless book, which is why the riskless rate appears at all.
The idea in plain words
Build a book, called pi, with P shares and Q option contracts. Its value is P times the stock plus Q times the option value V. Over a tiny step, P and Q are treated as constants. You chose them. They do not jump just because the stock goes from 100 to 101. So the book's change is P times the stock's change plus Q times the option's change.
The stock's change is known by assumption. The option's change is not, because V is an unknown function of time and of the random stock. Ito's lemma expands that function. It includes the time derivative, the first derivative with respect to the stock times the stock's change, and second-order pieces. Stochastic calculus then deletes most of the second-order pieces. A squared time step counts as zero. A time step times a Brownian increment counts as zero. A squared Brownian increment counts as a time step, otherwise continuous randomness would not accumulate.
Put that expression back into the book. Choose P to be the negative of that derivative, the quantity traders call delta, and set Q to one. The choice is the same idea if you hold Q contracts and scale the shares by Q. The random terms cancel. Nothing random is left.
A book with no risk left must earn the riskless rate. Its change equals the rate times the book's value times the time step. Substitute both expressions, cancel the time step, and rearrange. Delta here is the first derivative of V with respect to S. The drift of the stock does not remain. It was inside the random piece that the hedge removed. What remains is a partial differential equation, not a stochastic one. In the lecture's order it sets the time derivative, plus one half times volatility squared times the stock squared times the second derivative, plus the financing rate times the stock times delta, minus the rate times the option value, equal to zero. This transcript does not write the call formula. It stops when that equation is on the board.
How to use it
Use the derivation as a checklist when a model is extended. Identify the assumed stock equation. Identify the function being expanded. State which products of tiny increments you are setting to zero and which one you are keeping. Name the holding that cancels the random term. Only then impose the riskless growth. If a later model adds jumps, the cancellation step fails in the form used here, which is why this equation belongs to the diffusion case. If you need a number, solving the equation is a separate piece of work. Quoting the equation is not yet a price.
What usually goes wrong
Keeping the stock's drift in the final equation after the hedge has removed it. Forgetting the one-half from the second-order term. Treating P and Q as random inside the tiny step. And announcing a market price as "Black-Scholes" without the hedge argument that forced the riskless rate. The lecture also slips in real time, at one point writing the second derivative as if it belonged to a different function, then correcting it back to V. The corrected object is the option value.
Questions people ask
Why does the stock's expected return disappear?
Because it rides inside the random stock change. Once delta shares cancel that change, the drift has nothing left to multiply. The hedged book is financed at the riskless rate instead. That is the step that makes one option price rather than a different price for every investor's drift.
What is delta in this derivation?
It is the partial derivative of the option value with respect to the stock. Holding the negative of that many shares, per option, is the hedge. The lecture treats the sign as the choice that makes the two random terms subtract.
Does this prove geometric Brownian motion is realistic?
No. The lecture sets the objection aside on purpose. The equation is the consequence of the assumption plus the hedge. A different stock process produces a different equation. Realism is a question about which process you are willing to hedge.