Pull up an NVDA call three weeks out and your platform shows a price, a delta, a theta and an implied volatility number. All of that comes from a pricing model, and in the vast majority of cases that model is some version of Black-Scholes. The Black Scholes model — you'll see it written with the hyphen and without — is a formula that takes five inputs: the stock price, the strike, the time left, the interest rate and volatility, and returns a theoretical price for a European-style option.
That's the whole idea. Everything else is plumbing. The formula looks intimidating because it uses logarithms and a normal distribution function, but what it's doing is simple to describe: it works out the average payoff of an option across every price the stock might plausibly reach by expiration, then discounts that average back to today's dollars.
Fischer Black and Myron Scholes published the formula in 1973, the same year the Cboe opened its trading floor, and Robert Merton extended it almost immediately. Scholes and Merton received the Nobel Memorial Prize in Economic Sciences in 1997; Black had died two years earlier. The timing mattered. Listed options needed a common language for value, and the formula gave traders one. It still shapes most of what you'll read about options today, from quote screens to risk reports.
What the model is actually calculating
Think about a $400 call on a stock trading at $395 with a month to go. Nobody knows where the stock lands. But if you assume returns are randomly scattered in a particular way, you can describe the whole range of outcomes as a probability distribution, with most of the weight near today's price and thinner tails stretching out either side.
The option only pays when the stock finishes above $400, and it pays more the further above it goes. So the fair value is the probability-weighted average of all those payoffs, discounted for the time you'll wait to collect. Black-Scholes does exactly that, assuming stock prices follow geometric Brownian motion — a random walk with a constant drift and a constant volatility.
That one assumption is why the formula is both elegant and imperfect. It lets the math resolve into a closed-form answer instead of requiring a simulation. It also means the model thinks markets move in a smooth, well-behaved way, which they do right up until they don't.
The five inputs
Underlying price
Where the stock or index trades right now. Move AMD from $160 to $165 and every call on it is worth more, every put worth less. This is the input that changes fastest during the session, and the one delta measures sensitivity to.
Strike price
Fixed for the life of the contract. The relationship between strike and underlying price determines how much of an option's value is intrinsic (real, exercisable value) and how much is extrinsic (pure time and volatility premium).
Time to expiration
Measured in years in the formula, so a 30-day option goes in as roughly 0.082. More time means more opportunity for the stock to move, which means more premium. Notice the math: the time term enters as a square root, so doubling the days left does not double the extrinsic value. Four times the time gives you twice the volatility-driven premium.
Risk-free interest rate
Usually a short-term Treasury yield matching the option's life. It's the least dramatic input. Rates affect calls and puts in opposite directions through the discounting term, and on a two-week contract the effect is small enough that most directional traders never think about it. On a one-year LEAPS, it's real.
Volatility
The only input you can't look up. Everything else is observable; volatility is a forecast of how much the underlying will move, annualized, over the life of the contract. It's also the input the price is most sensitive to, which is why so much of options trading becomes an argument about volatility rather than direction. If you want to see how much each input moves the number, the Cboe options calculator lets you change one at a time and watch the theoretical price and Greeks respond.
Dividends are the sixth input in practice. The original 1973 paper assumed none; Merton's extension handles a continuous dividend yield, and that's the version most platforms run.
The Black-Scholes formula, piece by piece
For a European call, the formula is:
C = S × N(d1) − K × e^(−rT) × N(d2)
Where S is the stock price, K is the strike, r is the risk-free rate, T is time to expiration in years, and N() is the cumulative standard normal distribution — a function that converts a number into a probability between 0 and 1.
The two inner terms are:
d1 = [ln(S/K) + (r + σ²/2) × T] ÷ (σ × √T)
d2 = d1 − σ × √T
Here's how to read it without doing the algebra. The second half, K × e^(−rT) × N(d2), is what you expect to pay to exercise: the strike, discounted back to today, multiplied by the risk-neutral probability the option finishes in the money. N(d2) is that probability. The first half, S × N(d1), is what you expect to receive: the value of the stock you'd end up holding, weighted by the chance you end up holding it. Subtract the cost from the benefit and you have the call's value.
Puts come out of the same machinery, and you can also get there through put-call parity: P = C − S + K × e^(−rT). If a put and call on the same strike and expiration don't satisfy that relationship, there's an arbitrage, which is precisely the condition the model was built on.
Where the Greeks come from
Every Greek on your screen is a derivative of the Black-Scholes price with respect to one input. Change the stock price and ask how fast the option price changes, and you get delta. Change it again and you get gamma. Differentiate with respect to time and you have theta; with respect to volatility, vega; with respect to the interest rate, rho. The whole family of options Greeks exists because someone wrote down a pricing function that can be differentiated.
That connection explains a few things that otherwise look like trivia. N(d1) in the call formula is delta for a non-dividend-paying stock, which is why delta is sometimes described loosely as the probability of finishing in the money — it's close to N(d2), but not the same number. It's also why theta accelerates into expiration: the square root of time in the denominator gets small fast.
Our guides to delta, gamma, theta and vega go input by input, and they're worth reading in that order, because the Greeks are much easier to hold in your head once you see them as answers to "what happens to this formula if I nudge one variable?"
Where the model breaks
Black-Scholes assumes constant volatility, continuous trading, no transaction costs, no early exercise, and returns that are normally distributed in log terms. Real markets agree with none of that.
- Volatility isn't constant. It clusters, spikes on earnings and collapses afterward. A single sigma for the life of the contract is a convenient fiction.
- Prices gap. The model assumes you can hedge continuously at any price in between. A crypto-linked name like COIN opening 9% lower after a weekend headline says otherwise.
- Tails are fatter than the normal distribution allows. Large moves happen far more often than the math predicts, which is why far out-of-the-money puts trade richer than theory says they should.
- American options can be exercised early. Black-Scholes prices European exercise. For single-stock options that can be assigned any day, binomial models handle early exercise more honestly, though the difference is usually small outside of dividend situations.
- Very short-dated contracts strain the assumptions. The smooth decay curve the model implies turns jagged on expiration day, and anyone trading 0DTE options is working in a regime where gamma dominates and a hedging model built on continuous rebalancing has very little room to work.
The 1987 crash made the mismatch visible in market prices. Before it, implied volatility across strikes was roughly flat. After it, puts started trading at systematically higher implied volatilities than calls, producing the volatility skew that still shows up on every equity index today. The market effectively decided the model's distribution was wrong and priced the correction into the smile.
How traders actually use it
Almost nobody runs Black-Scholes forward to find a "correct" price and then compares it with the market. The market price is the market price. What traders do is run the model backwards: take the price being quoted, hold the other four inputs fixed, and solve for the volatility number that makes the formula match. That output is implied volatility, and it's the single most useful thing the model produces.
Once every contract has an implied volatility attached, you can compare a two-week QQQ call with a three-month SPX put and ask which one is expensive. You can see that an NVDA call into earnings carries a 70% implied volatility against a 35% baseline and understand why it might lose money even on a correct directional call. You can build a term structure and a skew and trade the shape of them.
This is also where strategy selection comes from. Choosing between a long call, a vertical spread and an iron condor is usually a volatility decision dressed up as a directional one, and the options strategies that fit a high-IV environment are different from the ones that work when premium is cheap. The model doesn't make that call for you. It just gives you the measuring stick.
FAQs
Is the Black-Scholes model still used today?
Yes, constantly. Brokers use it or a close variant to display theoretical values and Greeks, exchanges use it to disseminate implied volatility, and risk systems use it to aggregate exposure. More sophisticated desks layer stochastic volatility or local volatility models on top for exotic products, but Black-Scholes remains the common reference point for listed equity and index options.
What's the difference between the Black-Scholes model and the binomial model?
Black-Scholes gives a closed-form answer for European-style options in a single calculation. The binomial model builds a tree of possible prices step by step, which is slower but can handle early exercise, discrete dividends and American-style contracts. For most short-dated options on non-dividend-paying stocks, the two produce nearly identical numbers.
Does Black-Scholes work for American options?
Not exactly. It assumes exercise only at expiration. For American calls on stocks that don't pay dividends, early exercise is almost never optimal, so the difference is negligible. For American puts, and for calls on dividend-paying names near the ex-dividend date, a binomial or trinomial model is the better tool.
Why does my platform's theoretical price differ from the market price?
Because your platform is feeding the model its own assumptions — usually a volatility estimate, an interest rate and a dividend yield that may not match what market makers are using. It can also be a stale quote, a wide bid-ask spread, or genuine supply and demand pushing premium away from theory. Treat the theoretical price as a reference, not a verdict.
Do I need to understand the formula to trade options?
You need to understand what it responds to, not how to compute it. Knowing that extrinsic value scales with the square root of time, that volatility is the only guessed input, and that the Greeks are just sensitivity measures will change how you pick contracts. Deriving d1 by hand will not.
You don't need to memorize the Black Scholes model to trade well, but the traders who understand what it's doing make better choices about which strike and expiration to buy, because they can see where the premium is coming from before they pay it. That edge shows up in results over a few hundred trades, not one.
When you're ready to put pricing knowledge to work under a defined risk framework, our options trading plans give you a funded path with published rules and no hidden limits.
