LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
Black-Scholes, in this lecture, is a translator. It takes a traded option price and, under geometric Brownian motion, hands back an implied volatility and a delta. Those numbers let you compare contracts that are not the same stock. They describe what the market is charging now. They are not a forecast of the underlying, and a delta is not a frequency you will realize by holding that contract over and over.
Why it matters
Two at-the-money 15-day calls can both be $11, one on Nvidia near 214 and one on Google near 340, and still not be the same bet. Raw dollars cannot say which is richer. In that simulation Google's implied volatility is much lower, so Nvidia is the expensive contract in volatility terms.
The live screen later in the talk does not repeat that ranking. A 16-day at-the-money Google call was about $12, roughly $1,200 a contract, and the Nvidia call of similar tenor was about $690 to $700. At-the-money implied volatility was 40.8 percent for Google and 40.3 percent for Nvidia, so on that snapshot Google was slightly richer in volatility. The model is for the comparison in front of you. An hour later the ranking can flip.
The idea in plain words
Implied volatility is a now-cast. The lecture's Nvidia path ran from 208 to 216 to 218 and back toward 208 to 213, with volatility swinging beside it. People call the formula useless because it does not say where price will print. The lecture agrees, and says the use is a comparison and a filter.
Delta is the second number, often read as the chance of expiring in the money. In a stationary Black-Scholes world that reading is the cumulative normal, and a large sample converges. The lecture simulates a 15-day at-the-money Nvidia call with delta 0.53. Early bar charts are not 53 percent in the money and 47 percent out. By the end of the simulated sample the in-the-money share is 52.4 percent and the out-of-the-money share is 47.6 percent, moving toward 53. The lecture then says the law of large numbers and the central limit theorem do not hand you that convergence in listed markets. Drop the stock 3 percent in a couple of hours and that 0.53 delta can be near 0.1, because the contract is repriced. A 30 delta is not "this expires in the money 30 percent of the time." Buy a pile of them on that theory and you are treating a moving quote as a die.
On the live chain, with Nvidia near 210, a roughly 30 delta call sat about 12.5 points up, and the print ticked to 29 while they watched. The same point gap is not the same delta on another name. Google was near 346. Twelve points up was a 38 delta, not a 30. The 30 delta was the 365 strike, about 20 points up. Delta is how the lecture says you line contracts up across underlyings when a fixed number of points does not mean the same thing.
How to use it
Read implied volatility when you are asking which contract is rich, and delta when you are asking which strike is a comparable distance. Repeat the comparison after the spot moves. A rule that always sells the call 12 points up is a different option on Google than on Nvidia.
Use the output as a filter beside a decision you can explain. The model does not play the hand. Desks adjust a pure delta hedge because the continuous, costless hedge is not the market. The talk still wants the model, because without it there is no common scale.
What usually goes wrong
Equal premiums get called equal value. The classroom pair was both $11, and volatility said Nvidia was richer. The live pair said otherwise. Both can be true of their moment. A plan to sell every 30 delta and expect a 30 percent hit rate treats a moving quote as a die. Dropping the formula because prices are not geometric Brownian motion throws out the scale the lecture still wants for today.
Questions people ask
Which contract was more expensive, Nvidia or Google?
In the simulation, same $11 premium and a higher implied volatility on Nvidia, so Nvidia. On the later live 16-day snapshot, Google's at-the-money volatility was 40.8 percent and Nvidia's was 40.3 percent. The ranking is the current quote.
Does a 0.53 delta converge in real life?
In the lecture's stationary simulation the in-the-money share moved to 52.4 percent. The talk says that convergence is a classroom feature. A 3 percent drop can reprice the same contract near a 0.1 delta.
Why not compare strikes a fixed number of points apart?
Because 12 points was about a 30 delta on Nvidia near 210, and 12 points on Google near 346 was a 38 delta. The 30 delta there was about 20 points up, at 365.