Common mistake 1 with Black–Scholes option pricing

Feeding it a single volatility and trusting the price of a far out-of-the-money option - that is exactly where the smile bites hardest.

Common mistake 2 with Black–Scholes option pricing

Using it for an American option on a dividend-paying stock, where early exercise can be optimal.

Common mistake 3 with Black–Scholes option pricing

Reading N(d₂) as the real-world probability of exercise. It is the risk-neutral probability; the real-world number is higher for a stock with positive expected return.

Common mistake 4 with Black–Scholes option pricing

Quoting a price to the paisa from an implied volatility guessed to the nearest whole number. The price inherits the input's precision, and volatility is the input nobody observes.

EconStatsCraft › Black–Scholes option pricing

Black–Scholes option pricing

Prices a European option from spot, strike, time, rate and volatility, and reports every Greek scaled to the move people actually quote. Give it a market price and it backs out the implied volatility instead.

The intuition

An option is a right, not an obligation, so its payoff is bent: you take the upside and walk away from the downside. Black and Scholes showed that if you can trade the underlying continuously, you can build a portfolio of stock and cash that exactly reproduces that bent payoff. If two things pay the same in every state, they cost the same today - and that argument, not any forecast of where the stock is going, is what prices the option.

When to use it

Pricing or hedging European calls and puts on a liquid underlying; converting a quoted option price into an implied volatility; teaching how the Greeks behave. For American options, dividends paid discretely, or path-dependent payoffs you need a lattice, a PDE solver or Monte Carlo instead.

Where the estimator comes from

Assume dS=\mu S\,dt+\sigma S\,dW. Form \Pi=V-\Delta S and choose \Delta=\partial V/\partial S. Ito\'s lemma gives d\Pi=(\partial_t V+\tfrac12\sigma^2S^2\partial_{SS}V)dt, with the dW term cancelled by construction. A portfolio with no randomness must earn the risk-free rate or there is an arbitrage, so d\Pi=r\Pi\,dt, which rearranges to the Black-Scholes PDE \partial_t V+\tfrac12\sigma^2S^2\partial_{SS}V+rS\partial_S V-rV=0. Note what vanished: \mu, the expected return on the stock. Solving with the call payoff gives C=S\Phi(d_1)-Ke^{-rT}\Phi(d_2), where d_1=[\ln(S/K)+(r+\sigma^2/2)T]/(\sigma\sqrt{T}) and d_2=d_1-\sigma\sqrt{T}.

The model

C = S e^{-qT}N(d_1) - K e^{-rT}N(d_2),\qquad d_1 = \frac{\ln(S/K) + (r - q + \tfrac12\sigma^2)T}{\sigma\sqrt{T}},\qquad d_2 = d_1 - \sigma\sqrt{T}

What it assumes

The underlying price is a geometric Brownian motion - lognormal, no jumps.
Volatility and the interest rate are constant over the life of the option.
Frictionless, continuous trading: no transaction costs, any size, any time.
European exercise - the option can only be exercised at expiry.
Volatility is constant over the life of the option. It is the one input you cannot observe, and the fact that implied volatility differs across strikes - the smile - is the market telling you this assumption is false.
The stock pays no dividend over the life of the option, or the dividend is entered explicitly. An unmodelled dividend makes a call worth less than the formula says and an early exercise potentially rational.
Trading is continuous and frictionless, so the hedge can be rebalanced at no cost. The delta the formula reports is only a hedge ratio if you can actually trade on it.
Returns are lognormal with no jumps. A single overnight gap does more damage to the hedging argument than a whole month of ordinary volatility.

Statistical properties

The expected return on the stock does not appear in the formula. Two investors who disagree completely about where the stock is heading must still agree on the option price, provided they agree on volatility.
Delta, the hedge ratio, is Phi(d1). Phi(d2) is a different quantity: the risk-neutral probability that the option finishes in the money.
Volatility is the only input that cannot be observed. In practice the formula is run backwards - take the market price, solve for the volatility that reproduces it, and quote that.
Implied volatility is not constant across strikes. The smile or skew that appears when you plot it is the market pricing in fat tails and crash risk that the lognormal assumption denies.
The model assumes continuous costless trading, constant volatility and a constant rate, no jumps, and no dividends unless added explicitly. Every one of these is false; the formula survives because it is used as a quoting convention for implied volatility rather than as a belief.
Put-call parity, C - P = S - K exp(-rT), is model-free. It follows from arbitrage alone, so it holds even where Black-Scholes does not, and it is the first thing to check.

How to read the output

Price splits into intrinsic value (payoff if exercised now) and time value (everything else). Time value is largest at the money and decays to zero at expiry.
Delta is the hedge ratio: hold −delta units of the underlying per option to be first-order neutral. It also approximates the risk-neutral probability of finishing in the money.
Gamma is how fast delta moves; high gamma near the strike near expiry is why hedges have to be rebalanced most there.
Vega, theta and rho are quoted per 1 vol point, per calendar day, and per 1 rate point respectively - the sizes traders actually think in.
Given a market price, the implied volatility is the σ that reproduces it. Plotting implied vol against strike gives the volatility smile, the market's admission that the lognormal assumption is wrong.
Time value is the whole price of an out-of-the-money option and decays fastest near expiry, which is what theta measures. A position that is right about direction and wrong about timing loses money.
Vega is largest for at-the-money options with time left, so that is where a volatility view is most cheaply expressed - and where a volatility error costs most.
Gamma is the rate at which delta changes, so it is the cost of hedging: high gamma means the hedge must be rebalanced often, and the rebalancing is where the theoretical profit goes.

A worked reading

S = 100, K = 105, six months to expiry, r = 4%, σ = 25%. The call is worth 5.7799, all of it time value: the intrinsic value is zero because the option is out of the money, so every penny of the price is paying for the chance that it will not stay that way. Delta is 0.4703, meaning a short position in one call is hedged by holding 47 shares per 100 options - and gamma of 0.022505 says that hedge ratio moves by roughly two shares per hundred for each one-point move in the spot, which is why the hedge has to be rebalanced rather than set once. Vega is 0.2813 per volatility point and theta is -0.0238 a day, so roughly one day of decay is paid for by a tenth of a volatility point. Note d₂ = -0.2513: Φ(d₂) is the RISK-NEUTRAL probability of finishing in the money, which is a pricing device and deliberately not a forecast, because it is computed in the measure where the stock drifts at the risk-free rate rather than at whatever it actually earns.

Symptom → diagnosis → what to do

If you see this	It means	Do this
Implied volatility differs across strikes	The volatility smile	Expected. The lognormal assumption is the thing failing
Price below intrinsic value	An arbitrage, or an input error	Check units on rate and maturity: both must be annual
Deep in-the-money call worth less than S - K exp(-rT)	Parity violated	An input is wrong; parity needs no model
Vega near zero	Deep in or out of the money	Implied volatility cannot be recovered from such an option
Very short maturity, tiny price	Correct behaviour	Time value goes to zero; only intrinsic value remains
Volatility entered as 25 rather than 0.25	Percent versus decimal	The commonest input error by a wide margin

Common mistakes

Feeding it a single volatility and trusting the price of a far out-of-the-money option - that is exactly where the smile bites hardest.
Using it for an American option on a dividend-paying stock, where early exercise can be optimal.
Reading N(d₂) as the real-world probability of exercise. It is the risk-neutral probability; the real-world number is higher for a stock with positive expected return.
Quoting a price to the paisa from an implied volatility guessed to the nearest whole number. The price inherits the input's precision, and volatility is the input nobody observes.

Where to read this properly

Black and Scholes (1973), "The Pricing of Options and Corporate Liabilities", Journal of Political Economy 81(3).
Merton (1973), "Theory of Rational Option Pricing", Bell Journal of Economics 4(1).
Hull, Options, Futures and Other Derivatives (2021), ch. 15 and 19.
Derman and Miller, The Volatility Smile (2016), for why the constant-volatility assumption fails.

Related Finance & risk

Bond price, duration & convexity · Merton distance to default · EMI / home loan planner · CAPM / beta estimation · Multi-factor model · Portfolio optimization · Value at Risk & Expected Shortfall · Performance & drawdown · Time value of money & NPV/IRR · Equity valuation (DDM, FCF, residual income) · Cost of capital (WACC) · Binomial option pricing · Option strategies (payoff & profit) · Forwards, futures & swaps · Credit risk & Basel capital · Extreme value theory (tail risk)

All analyses

Data & variables

Each example is simulated from a known model, so you can check whether the estimator recovers the truth - and nothing here is anyone's private data.

Filter variables

Drag a variable into a slot on the right - or just click it.

Choose an example dataset above to begin.

Choose a transformation - it opens in the data editor, with a live preview.

Apply to

Multiply by

Columns ctrl-click for several

Group by ctrl-click for several

Within groups

Statistic

Bands

How many

Cut points

Keep rows where

Value

and

Fill with

Value

Also mark which rows were filled

Rules

Base period = 100

Which part

Percent at each tail

Threshold

Survey weight optional

Direction

Names column

Values column

Largest first

Delete them, not just count them

Read this column as

Tidy the text

New name optional

EconStatsCraft rigorous, plain-spoken econometrics

Econometrics that explains itself.

Regression, panel data, time series and the whole battery of tests. Every result is followed by a verdict you can defend.

Load a dataset to see it here.

Black–Scholes option pricing

The intuition

When to use it

Where the estimator comes from

The model

What it assumes

Statistical properties

How to read the output

A worked reading

Symptom → diagnosis → what to do

Common mistakes

Where to read this properly

See also

Related Finance & risk

All analyses

Charts

Maps

Explore

Regression

Time series - one series

Time series - several series

Panel data

Multivariate statistics

Marketing research

Causal

Finance & risk

Actuarial

Tests & breaks

Machine learning

Econometrics that explains itself.