Black-Scholes Model

TL;DR: Black-Scholes prices an option by asking what it would cost to manufacture its payoff by trading the stock. It is not predicting anything — it is ruling out free money.

The model, published by Fischer Black and Myron Scholes in 1973 with Robert Merton, starts from an unusual question: instead of guessing what an option is worth, it asks what it would cost to replicate the option's payoff by continuously buying and selling the underlying stock. If that replication is possible, the option cannot trade above or below its replication cost without handing someone a risk-free profit. The price falls out of the arithmetic, and no view about where the stock is going ever enters it.

Five inputs go in: the stock price, the strike, the time to expiration, the risk-free rate, and volatility. Four of those you can read off a screen. Volatility — how much the stock will move between now and expiration — you cannot, because it has not happened yet. This asymmetry is the model's most useful consequence: run the formula backward from a traded option price and you recover the volatility the market is assuming, which is exactly what implied volatility is. Traders quote options in volatility terms precisely because the model makes that translation possible.

The partial derivatives of the formula are the greeks: how the price moves with the stock (delta), how delta itself moves (gamma), how much value time removes (theta), and how sensitive the price is to the volatility assumption (vega). These are not add-ons — they are the same equation read from different angles, which is why they are internally consistent in a way that eyeballed rules of thumb are not.

The assumptions are where the model and reality part company. It assumes volatility is constant, returns are lognormal, trading is continuous and frictionless, and the option can only be exercised at expiration. Real option chains disagree visibly: different strikes trade at different implied volatilities, a pattern called volatility skew, which is the market's way of saying crashes are fatter and more sudden than a lognormal distribution allows. US single-stock options are also American — exercisable early — so they can be worth more than the model says, most notably in-the-money calls just before a dividend.

None of that has retired it. Feed each strike its own implied volatility rather than one number for the whole chain and Black-Scholes stops being a theory that is wrong and becomes a quoting convention that works — a shared language for translating between price and volatility. That is how the market actually uses it, and it is why every options platform, including this one, still computes it. Work through the numbers on the Black-Scholes calculator, which prices a call and a put side by side and prints the put-call parity check so you can verify the output by hand.