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-Merton partial differential equation is what remains after a hedged option portfolio is required to earn the riskless rate. The lesson builds a book that is long one derivative and short a chosen number of shares. Ito's lemma expands the derivative. Choosing the share count equal to the derivative's spot sensitivity cancels the random term. A riskless book is then not allowed to earn more than the riskless rate, the time increments cancel, and the equation left on the page is the riskless rate times the portfolio, equal to the time derivative plus one half sigma squared S squared times the second spot derivative.
Why it matters
The hedging step removes the Brownian shock and leaves a partial differential equation. The lesson says this one has a known analytical solution, and that other stochastic equations need numerical methods. The session stops at the differential equation. It mentions later deep-learning work only as something built on this foundation.
The idea in plain words
The portfolio value, pi, starts as the derivative value d minus the share price S. The derivative is a function of the share price and of time. The share follows geometric Brownian motion: the relative change is a drift times d t plus volatility times a Brownian increment. Randomness enters the book only through that Brownian piece, once through the short shares and once through the derivative.
Ito's lemma expands the derivative into a time partial times d t, a spot partial times dS, and one half the second spot partial times dS squared. The squared price increment reduces to volatility squared times S squared times d t, because a squared time increment vanishes, a time increment times a Brownian increment vanishes, and a squared Brownian increment equals the time increment.
After that substitution, the only random piece inside the derivative is the spot partial times dS. The short share contributes a random piece as well. The lesson rewrites the share holding as p shares, with p chosen by the hedger. Set p equal to the spot partial of d. That quantity is delta. The two dS terms are identical and drop. The change in the portfolio is then the time partial times d t, plus one half the second spot partial times volatility squared times S squared times d t. Nothing random remains. The lesson calls this a delta hedge and calls the resulting book riskless.
A riskless change, the lesson argues, cannot earn more than the riskless rate. It sets the change in the portfolio equal to the riskless rate times the portfolio value times d t. Both sides contain d t. Cancel it. The equation is r times pi equals the time partial of d, plus one half sigma squared S squared times the second spot partial of d. That is the form the lesson names as the Black-Scholes-Merton equation. pi is still the derivative minus delta shares.
How to use it
Write the book as long the derivative and short p shares. Expand the derivative with Ito's lemma. Reduce dS squared with the three multiplication rules. Choose p as the spot partial. Watch the dS terms cancel. Set the remaining change equal to r times pi times d t, and divide by d t. The two partials of d and the portfolio value are what the equation relates.
The lesson does not solve the equation in this session. Any share count other than the spot partial leaves a random term, and the riskless comparison does not apply.
What usually goes wrong
Hedging with a constant share count that is not delta leaves the Brownian term in the portfolio change. The riskless step is then unavailable. The lesson's p is specifically the spot partial.
Using the wrong multiplication rule for dS squared drops or doubles the volatility term. The term that should survive is one half the second spot partial times sigma squared S squared. That is the only place volatility appears after the hedge.
The lesson sets the sure change equal to the riskless change, because a sure gain above that rate should not be available. The displayed equation keeps the portfolio value on the left.
An analytical solution is a property of this equation, in the lesson's words, and not a property of every pricing equation. Reaching for a closed form after a different hedge fails is the case they flag for numerical methods.
Questions people ask
What share count cancels the randomness?
Delta, the partial derivative of the option value with respect to the share price. The lesson holds that many shares short against one long derivative.
What equation is left after the time increment cancels?
The riskless rate times the portfolio equals the time partial of the derivative, plus one half sigma squared S squared times the second spot partial.
Does the session also produce the call formula?
It says an analytical solution of this equation is known. It does not derive that formula here.