Black-Scholes Model Explained: Formula, Assumptions, and Limits

The Black-Scholes model is a mathematical formula, published in 1973 by Fischer Black and Myron Scholes, that estimates the fair price of a European-style option from five inputs: the current stock price, the strike price, the time to expiration, the risk-free interest rate, and the stock’s volatility. Robert Merton independently strengthened its mathematical foundation, and Merton and Scholes received the 1997 Nobel Memorial Prize in Economic Sciences for the work; Black had died in 1995.¹ It remains a reference point for clearinghouses, trading desks, and risk systems across global derivatives markets, even though practitioners now know exactly where it breaks.

The Formulas for Calls and Puts

A European call option gives the holder the right to buy a stock at a set strike price on expiration day. Its Black-Scholes price is:

C = S × N(d₁) − K × e^(−rT) × N(d₂)

A European put, the right to sell at the strike, is priced by:

P = K × e^(−rT) × N(−d₂) − S × N(−d₁)²

Both share the same two building blocks:

d₁ = [ln(S/K) + (r + σ²/2) × T] / (σ × √T)

d₂ = d₁ − σ × √T³

The symbols:

S is the current stock price.
K is the strike price written into the contract.
T is time to expiration, as a fraction of a year.
r is the risk-free interest rate, annualized and continuously compounded.
σ is the volatility of the stock’s returns, as an annualized standard deviation.
N(·) is the cumulative standard normal distribution, which converts a value into a probability between 0 and 1.
ln is the natural logarithm, and e is Euler’s number (about 2.71828).

The call formula weighs the current stock price against the present value of the strike, each adjusted by a probability factor. N(d₁) and N(d₂) represent the probabilities that the option finishes with value at expiration, filtered through the normal distribution.³ The put formula uses the mirror-image probabilities N(−d₁) and N(−d₂). Calls and puts are also linked by put-call parity, which says the difference between a call and a put at the same strike equals the stock price minus the present value of the strike.⁴

Where the Five Inputs Come From

Four of the inputs are easy to pin down. Volatility is where the judgment sits.

Price, Strike, and Time

The current stock price is quoted on any brokerage platform in real time. The strike is a fixed term of the contract. Time to expiration is converted into a decimal fraction of a year, so an option expiring in 182 days is entered as roughly 0.50 (182 ÷ 365).

Risk-Free Rate

The risk-free rate is typically pulled from U.S. Treasury bill yields, which carry virtually no default risk.⁵ Most practitioners match the Treasury maturity to the option’s life. For a three-month option, the 13-week T-bill yield is a common choice. Yields are published daily on the Department of the Treasury’s website. A 5% yield is entered as 0.05, and the formula treats it as continuously compounded.

Volatility

Volatility measures how much a stock’s price swings, expressed as the annualized standard deviation of its returns. Two ways to get the number exist, and the distinction matters.

Historical volatility looks backward. You take past closing prices over some window (commonly 20 to 90 trading days), calculate the daily logarithmic returns, find their standard deviation, and annualize. That tells you how much the stock actually moved.

Implied volatility looks forward. You take the option’s current market price and work the Black-Scholes formula in reverse to find the volatility level that would reproduce that price. Volatility cannot be isolated algebraically, so numerical methods such as the Newton-Raphson algorithm converge on the answer by successive approximation.⁶ The result reflects what the market collectively expects volatility to be between now and expiration.

Active traders usually watch implied volatility, since it captures the market’s real-time expectations. Historical volatility works as a sanity check and as a starting point when no liquid options market exists for a particular stock.

A Worked Example

The math is easier to follow with real numbers. Suppose a stock trades at $100 and you want to price a European call with a $95 strike, 3 months (0.25 years) to expiration, a 1% risk-free rate, and 50% annualized volatility.

Start with d₁:

d₁ = [ln(100/95) + (0.01 + 0.50²/2) × 0.25] / (0.50 × √0.25)

The natural log of 100/95 is about 0.0513. The term (0.01 + 0.125) × 0.25 equals 0.03375. Added, that gives 0.08505 in the numerator. The denominator is 0.50 × 0.50 = 0.25. So d₁ ≈ 0.3402.

Then d₂ = 0.3402 − 0.25 = 0.0902.

N(0.3402) is roughly 0.6331 from a standard normal table. N(0.0902) is roughly 0.5359. The discounted strike is 95 × e^(−0.01 × 0.25) ≈ 94.76.

Plugging in: C = 100 × 0.6331 − 94.76 × 0.5359 ≈ 63.31 − 50.78 ≈ $12.53.

That $12.53 is the model’s theoretical fair price. If the market quotes $14, the option looks rich. If it quotes $11, cheap. In practice nobody runs this by hand. Spreadsheets and trading platforms have the normal distribution function built in, and the result is instant. Working through it once shows why nudging any input shifts the price.

What the Model Assumes

Every output rests on a set of idealized conditions. When the market diverges from them, the price diverges too.

Lognormal Stock Prices

The underlying stock is assumed to follow geometric Brownian motion, so the logarithm of the future price is normally distributed.⁷ Daily percentage returns are treated as random and bell-shaped, and the price itself can never fall below zero. It’s a reasonable first approximation that badly underestimates the frequency of extreme moves.

European Exercise Only

The standard formula assumes exercise only on the expiration date. American-style options, which allow exercise at any time, need different pricing methods because early exercise can add value the equation cannot see. The Options Clearing Corporation uses the Black-Scholes framework specifically for pricing European-style contracts such as binary options.⁸

No Dividends

The basic formula assumes no dividend payments before expiration. A dividend knocks the stock price down by roughly the dividend amount on the ex-dividend date, which shifts option values in ways the plain formula misses. Dividend-paying stocks need the modified version described further down.

Constant Volatility and Interest Rates

Both the risk-free rate and the stock’s volatility are assumed to hold steady for the option’s entire life. Neither actually does. Volatility in particular spikes during stress and compresses during calm. This is probably the assumption that fails most visibly.

No Frictions and Perfect Liquidity

The formula imagines free trading, no taxes, and arbitrageurs who instantly erase mispricings. That no-arbitrage condition is actually the mathematical backbone of the model, which is what lets it produce a single fair price. In the real world, bid-ask spreads, commissions, and taxes all create friction. The model also assumes you can trade any size at the quoted price without moving it, which holds for large-cap names and fails for thin ones.

The Greeks: How the Price Responds to Each Input

Because the formula expresses an option’s value as a function of its five inputs, you can take a derivative with respect to each input to measure sensitivity. Those sensitivities are the Greeks, and they turn the model from a price into a risk-management toolkit.

Delta is how much the option price moves per $1 change in the stock. For a non-dividend-paying stock, call delta equals N(d₁); with dividends, it becomes e^(−qT) × N(d₁), where q is the dividend yield. A 0.60 delta means the option gains about $0.60 per $1 rise in the stock.⁹
Gamma is how fast delta itself changes per $1 move. Think of it as the acceleration. Gamma peaks near the strike and shrinks as the option moves deep in or out of the money.
Theta is how much value the option bleeds each day as expiration approaches, all else equal. Theta accelerates near expiration, which is why the last few weeks can feel punishing for holders.
Vega is how much the price moves per one-point change in implied volatility. Both calls and puts gain when volatility rises and lose when it falls. Ignoring vega is one of the more expensive mistakes new traders make, because a position can lose money even when the stock moves the right way if volatility drops at the same time.
Rho is how much the price moves per one-point change in the risk-free rate. It’s small for short-dated options and becomes meaningful for long-dated ones such as LEAPS.

Working traders rarely think about a position as a single price. They manage delta, watch gamma around key levels, track theta bleed, and monitor vega constantly.

Where the Model Breaks Down

The Volatility Smile

If the assumptions held perfectly, every option on the same stock with the same expiration would show the same implied volatility across strikes. They don’t. Plotting implied volatility against strike produces a curve that’s higher on the ends and lower in the middle, which traders call the volatility smile.¹⁰

The smile exists because the normal distribution assigns too little probability to extreme moves. Deep out-of-the-money puts protect against crashes that “shouldn’t” happen on a bell curve but do happen in real markets, so the market prices in tail risk by paying higher implied vols for those strikes.¹⁰ The smile is often asymmetric, with out-of-the-money puts carrying higher vols than equidistant calls; that asymmetric version is called a skew or smirk. Traders don’t abandon the formula because of it. They adjust the volatility input by strike, patching the blind spot with market-observed data.

Fat Tails

The normal distribution treats large swings as extraordinarily rare. A five- or six-standard-deviation daily move should happen once every few thousand years, yet real markets produce them every few years. Distributions that accommodate fat tails reduce pricing biases compared to standard Black-Scholes.¹¹ Research has shown that accounting for fat tails can raise the model-implied equity premium by 80% and lower the implied risk-free rate by 20%.

Real-World Stress Tests

The most famous failure came in 1998. Long-Term Capital Management, a hedge fund whose principals included Myron Scholes himself, relied heavily on model-driven trading. When the Russian debt crisis produced correlations and price moves the models treated as near-impossible, LTCM lost $4 billion in six weeks and required a Federal Reserve-coordinated bank bailout. The 1987 crash told a similar story in a single day: the S&P 500 dropped more than 20% on Black Monday, an event the model would have assigned a probability indistinguishable from zero. Neither episode killed the formula, but both pushed practitioners to treat its output as a starting estimate rather than ground truth.

Volatility Clusters

Volatility stays high after it’s been high and low after it’s been low, which contradicts the constant-volatility assumption directly. Stochastic volatility models like Heston and SABR let it change over time and match real options markets more closely. The basic Black-Scholes formula still works as the common quoting language, though. When traders say “vol” to each other, they mean Black-Scholes implied volatility, even if a more sophisticated model is running the book.

Adapting the Model for Dividends and American Options

The Merton Dividend Adjustment

Robert Merton extended the formula for stocks that pay a continuous dividend yield (q). The adjusted call replaces S with S × e^(−qT) throughout:

C = S × e^(−qT) × N(d₁) − K × e^(−rT) × N(d₂)

The d₁ calculation also changes, with (r − q + σ²/2) replacing (r + σ²/2) in the numerator.⁹ This fits stock indexes well, where many component dividends are paid throughout the year and the continuous-yield approximation holds up. For individual stocks that pay on specific dates, the continuous model is rougher, and traders often calculate the present value of expected dividends and subtract it from the stock price before using the standard formula.

American Options and Binomial Trees

The standard formula cannot price American-style options because it has no mechanism to check whether early exercise beats holding to expiration. The binomial tree model fills that gap by breaking the option’s life into many small steps and checking at each step whether early exercise pays.¹² Monte Carlo simulation is another alternative, useful for options with complex payoffs or multiple sources of uncertainty. For American puts and calls on dividend-paying stocks, these alternatives are standard practice.

A Note on Taxes

The model ignores taxes, but anyone trading on it cannot. Section 1256 contracts, which include nonequity options such as broad-based index options and regulated futures, receive a blended tax treatment: 60% of the gain or loss is long-term and 40% is short-term, regardless of holding period, and open positions are marked to market at year-end.¹³ Equity options on individual stocks generally fall outside Section 1256 and follow standard capital gains rules: short-term if held one year or less, long-term if held longer. The wash sale rule also applies, so selling a stock at a loss and buying an option on the same stock within 30 days before or after can disallow the loss and roll it into the new position’s basis. These frictions are exactly what the frictionless-market assumption sets aside, and they can shift the net profitability of an options strategy meaningfully.

1
NobelPrize.org. The Prize in Economic Sciences 1997 – Press Release
2
Stanford University. Black-Scholes Equation and Its Solution for Call/Put Options
3
Stony Brook University Department of Computer Science. Lecture 12 – The Black-Scholes Model
4
Ohio State University Department of Mathematics. Derivatives and Their Portfolios, Put-Call Parity
5
arXiv. Mathematical Modeling of Option Pricing with an Extended Black-Scholes Framework
6
University of California, Los Angeles Department of Statistics. Implied Volatilities
7
Washington University in Saint Louis. Options and Futures Lecture 4 – The Black-Scholes Model
8
The Options Clearing Corporation. File No. SR-OCC-2026-003 – Proposed Rule Change
9
Columbia University. The Black-Scholes Model
10
Fordham University. Implied Volatility, Volatility Smile/Skew/Smirk, and Risk-Neutral Density
11
ScienceDirect. The Impact of Fat Tails on Equilibrium Rates of Return and Term Premia
12
BCP Business & Management. Options Pricing Comparison between the Black-Scholes Model and the Binomial Tree Model
13
Office of the Law Revision Counsel. 26 US Code 1256 – Section 1256 Contracts Marked to Market