Black-Scholes in Plain English
Black-Scholes is the standard option pricing model. What it takes as inputs, what it gives back, the assumption that mostly holds, the one that doesn't, and where it specifically struggles for 0DTE.
Black-Scholes is the shared language of option pricing. Every IV number on every broker's options chain, every greek displayed in every analytics tool, every model in every trading book — they all reference it. Knowing what the model actually does, and where it stops working, is the difference between using its outputs intelligently and treating them like ground truth they're not.
This article explains what Black-Scholes does in plain English. What inputs it takes, what it gives back, the core assumption that mostly holds, the one that doesn't, and where it specifically struggles for 0DTE. No derivation, no integrals.
What it is
Black-Scholes — sometimes called Black-Scholes-Merton, or BSM — is a formula that converts a small set of inputs into a theoretical option price. It was developed by Fischer Black, Myron Scholes, and (independently) Robert Merton in the early 1970s. The 1997 Nobel Prize in Economics went to Scholes and Merton; Black had passed away before the award.
The standard option pricing model used across modern derivatives markets. Takes underlying price, strike, time to expiration, volatility, and risk-free rate (optionally dividend yield) as inputs; outputs a theoretical option price. The greeks fall out as partial derivatives of that price with respect to its inputs.
The reason BSM matters today isn't that it's perfectly accurate — it isn't. It matters because market participants converged on it. Brokers display BSM-derived greeks. Exchanges compute reference IVs from it. Trading software speaks its vocabulary. Even traders who use more sophisticated models (Heston for stochastic volatility, Merton/Bates for jump-diffusion, local-volatility models) communicate in BSM-equivalent terms because that's what everyone else does.
The inputs
Five (or six, if you model dividends) numbers go into Black-Scholes:
- Underlying price (S) — the current level of SPX, or the price of SPY, or the price of a stock. Directly observable from the market.
- Strike price (K) — the strike of the option being priced. Fixed when the option is listed.
- Time to expiration (T) — how long until the option expires, usually expressed in years. Known exactly at any moment.
- Risk-free rate (r) — the rate at which money compounds without risk over the option's life. Usually a short Treasury yield matching the option's maturity. Observable.
- Implied volatility (σ) — the volatility number that, when plugged into BSM, produces the option's observed market price. Not directly observable from anything — inferred by inverting the formula against the market price.
- Dividend yield (q) — the continuous dividend rate on the underlying. For SPX, modeled from the S&P 500's dividend yield. For SPY, the ETF's actual distribution. Often omitted in simplified examples.
The interesting one is volatility. Four of the inputs are essentially given: S, K, T, r. Volatility is the one the market doesn't directly hand you. Instead, the market quotes the option's price, and BSM is used in reverse: given the price and the other four inputs, what σ would BSM need to reproduce this price? That number is the implied volatility.
What it gives you
BSM produces two things:
A theoretical option price. Given inputs, BSM returns a fair price. In practice, that price is compared against the market price to determine implied volatility (as above), rather than treated as the answer to "what should this option cost?"
The greeks. Each greek is a partial derivative of the BSM price with respect to one input:
- Delta = how price changes when the underlying changes. Full coverage in Delta.
- Gamma = how delta itself changes when the underlying changes (second derivative). Full coverage in Gamma.
- Theta = how price changes as time passes. Full coverage in Theta.
- Vega = how price changes when implied volatility changes. Full coverage in Vega.
- Rho = how price changes when the risk-free rate changes. Smallest of the main greeks for short-dated options; mostly negligible for 0DTE.
There are higher-order greeks too — charm (the rate at which delta drifts over time, with spot held constant), vanna (how vega responds to underlying moves), volga (how vega responds to changes in vol). These second-order greeks become significant near expiration, which is why they matter for 0DTE specifically.
Every greek on every broker chain is computed by feeding the current inputs into BSM (or a BSM-equivalent) and reading off the derivatives.
The assumption that mostly holds
BSM rests on a small set of mathematical assumptions. The biggest one is the model of how the underlying moves: geometric Brownian motion.
A model for asset prices in which log returns are normally distributed with constant volatility, and the price evolves continuously through small random changes. The underlying never jumps; it walks.
In plain terms: BSM assumes the underlying moves in many tiny, normally-distributed steps over time, with the volatility of those steps held constant. For most equity-like underlyings on most days, this is a workable approximation. SPX daily log returns at the bulk of the distribution look roughly normal. The continuous-walk picture breaks at the edges — earnings, Fed announcements, macro headlines — but across the broad middle of typical sessions, it's close enough that BSM's machinery is useful.
The assumption that doesn't
The assumption that visibly doesn't hold is constant volatility. Real markets price different strikes at different IVs (the smile and skew), real volatility changes intraday and across weeks (the IV term structure), and real volatility responds to events.
The market's response to this failure is to use BSM strike-by-strike with a different IV at each strike. The IV skew is the market's way of patching BSM. If BSM with a single vol assumption matched reality, every strike at the same expiration would share one IV — the way they share one risk-free rate. They don't. OTM puts trade at higher IV than ATM, OTM calls (on SPX) trade at lower IV than ATM, and the shape changes with the market regime.
This is why an options chain shows IV as a column with a different value in every row. If the constant-vol assumption were correct, IV would be a single number per expiration. The fact that it isn't — and that the skew has a recognizable shape — is the visible evidence that BSM is an approximation traders work around rather than a true description of how options are priced. The shape itself is covered in The Volatility Smile and SPX Skew.
Where BSM specifically struggles for 0DTE
On 0DTE, the gap between the model and reality is widest. Four issues:
Discrete events dominate when time is short. BSM assumes continuous trading and continuous price evolution. With minutes left to expiration, single ticks, single news items, and single market-maker hedging waves move the price in ways that don't look like a smooth random walk. The model's continuous-process picture is least accurate when there's the least time for many small steps to average out.
Lognormal tails underweight large moves. The normal distribution at the core of BSM assigns small probabilities to large moves. Real markets show larger moves more often than the math predicts. On 0DTE, the option's payoff is decided by what happens in the next few hours — exactly the window where tail moves are most consequential.
Second-order greeks grow. Charm, vanna, and other second-order effects become significant near expiration. BSM still gives you delta, gamma, theta, vega — but those first-order numbers can shift substantially within minutes. Treating a delta you read at 1:00pm as still valid at 3:30pm is a mistake on 0DTE in a way it isn't on a 30-day option.
IV itself becomes unstable. With very little time-value left to revalue, the IV that BSM produces from a market price becomes sensitive to tiny price changes. A penny move in an option's price can produce a meaningful move in the computed IV. The IV number is still useful in aggregate, but trusting any single reading to many decimal places is a mistake.
Why we still use it
Despite all of the above, BSM remains the shared coordinate system of option markets. Every IV quote is a BSM-derived number. Every greek you see on any retail platform is from BSM or a BSM-equivalent. Brokers, exchanges, analytics tools, and the bulk of trading software speak BSM.
Alternative models exist — Heston for stochastic volatility, Bates/Merton jump-diffusion for jumps, local-volatility models for fitting the smile exactly — and professional desks do use them. For the typical retail 0DTE trader, those models live in the background. The numbers in your trading interface are BSM-derived.
Knowing what BSM does and where it's approximate lets you read those numbers with the right level of trust. They're the market's best shared estimate, not ground truth.
Key takeaways
- Black-Scholes is a formula that converts a few inputs into a theoretical option price. The greeks fall out as partial derivatives of that price with respect to its inputs.
- Five inputs go in: underlying, strike, time to expiry, risk-free rate, implied volatility. Implied volatility is the one not directly observable — it's derived by inverting BSM against the market price.
- The big assumption that mostly holds: the underlying moves in small normally-distributed steps over time (geometric Brownian motion). Reasonable for typical equity-like underlyings.
- The big assumption that doesn't hold: constant volatility. The IV smile and skew exist because traders use BSM with a different IV at each strike to make it produce realistic prices.
- For 0DTE specifically, BSM struggles with short-time discreteness, fat tails, growing second-order greeks, and IV instability near expiration. Use its outputs, but don't trust them to a precision they don't have.