LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
Delta is the slope of an option price with respect to the stock. In the lecture's chart, a call on a 100 stock costs about 10, and the delta is about 0.60, so a move from 100 to 101 is expected to lift the call by about 0.60. Delta is also used as a rough chance of finishing in the money. The Black-Scholes price is what you get, on average, if you sell the option and keep the position delta-neutral often enough. Daily hedging does not get all the way there. In one 10,000-path run the model price was 10.45 and the average daily-hedge gain was 5.72.
Why it matters
The Black-Scholes argument hedges in continuous time. Hold minus one call and plus delta shares, and an instantaneous stock move changes the option and the shares by offsetting amounts. The hedged portfolio then has no random piece left, so it should earn the riskless rate, or someone could arbitrage it. Continuous hedging is not something you can do. Each rebalance costs a trade. The lecture's numerical version hedges once a day, then more often, and watches the average gain move toward the model premium. That is the content of the formula: the premium you collect for a short option, if the hedge is fine enough and the assumed dynamics are true, not an extra gain on top.
The idea in plain words
The stock is assumed to follow geometric Brownian motion, a continuous path with a drift, a volatility, and a Brownian driver. The lecture simulates those paths and, alongside them, the call price and its delta. Far from expiry, delta is a gentle curve. Very near expiry it steepens around the strike, like a step, because a small stock move decides whether the call finishes in or out of the money. The price chart itself starts to look like the hockey-stick payoff. A short call has negative delta. If the stock rises, the short option loses. Buying delta shares offsets that, locally. Because both the price and delta are curved, the straight-line offset decays as the stock moves, so the hedge has to be reset. Resetting every instant is the theory. Resetting daily is the experiment. The parameter set is spot 100, strike 100, one year, riskless rate 5 percent, volatility 20 percent, and a drift of 8 percent.
How to use it
Read a daily hedge as a noisy version of the pricing argument, not as a gain engine. If your simulated mean sits well below the premium, you are seeing hedge error and path noise, which is what 5.72 versus 10.45 is. If you then change only the frequency and the mean approaches 10.45, you have checked the lecture's claim about continuous hedging. A separate experiment is an opinion about volatility. Price the option at the market's implied volatility, and simulate the stock at a lower realized volatility. The lecture's sketch uses 40 percent implied, as in a hypothetical where a name such as Nvidia is priced at that level, against paths with less volatility, on the order of 10 or 20 percent. One such run showed an average near 23.50, larger than the premium collected in the constant-volatility case they had just left. The interpretation they offer is that the seller collected a premium built for more movement than the paths delivered. That is a model statement about a volatility gap. It is not evidence that the gap will be there, or that a book of short options will keep the difference.
What usually goes wrong
Treating one path's gain of 44 or loss of 56 as the hedge result. The object the model talks about is the average. A second mistake is calling daily hedging "the Black-Scholes hedge." The formula assumes continuous rebalancing. The 5.72 mean is what a coarse hedge left on the table relative to 10.45. A third is smuggling in a volatility view and still describing the trade as model-neutral. The moment realized volatility is set below the volatility used to sell the option, you are betting on that gap. If the gap is wrong, the average moves the other way, and transaction costs, which the lecture flags as the reason continuous hedging is impossible, take another slice.
Questions people ask
What does a delta of 0.60 say in the chart?
Two things at once. A one-point stock rise should add about 0.60 to the call, and the lecture also reads delta as an approximate probability of expiring in the money. Both readings use the same curve.
Why was the average 5.72 and not 10.45?
Because the hedge was daily, not continuous, on that run of 10,000 paths. The lecture's claim is that a finer schedule pulls the mean up toward the 10.45 model price.
Does the 23.50 figure mean the short call earned that much for sure?
No. It is one simulation under a lower realized volatility than the volatility used to set the premium. Single paths in the matched-volatility hedge already swung from about plus 44 to about minus 56.