Black-Scholes-Merton Model: Contract, Inputs, Implied Volatility, and Greeks

Apply the Black-Scholes-Merton model only after matching the option claim, valuation timeline, discounting, dividends, volatility quote, settlement terms, and Greek units.

For educational purposes only; not investment advice. Investing may result in loss.

Direct answer

The Black-Scholes-Merton model is a no-arbitrage framework, not a legally binding quote or a forecast of the underlying. Its familiar closed form values a plain-vanilla European option under a specified diffusion, funding, dividend, trading and market-completeness setup. Before applying it, identify the actual series, exercise right, deliverable, multiplier, official settlement value and currency: an American physically settled equity option and a European cash-settled index option are different claims even when their screens show the same strike and expiration.

For spot S₀, strike K, time in years T, continuously compounded zero rate r, continuous proportional dividend yield q and annualized volatility σ, the European call and put formulas are C = S₀e^(−qT)N(d₁) − Ke^(−rT)N(d₂) and P = Ke^(−rT)N(−d₂) − S₀e^(−qT)N(−d₁), where d₁ = [ln(S₀/K) + (r − q + σ²/2)T] ÷ (σ√T), d₂ = d₁ − σ√T, and N(x) is the standard normal cumulative distribution function. With matched inputs, European put-call parity is C − P = S₀e^(−qT) − Ke^(−rT).

Market participants also invert the model: implied volatility is the σ that makes a chosen model reproduce a chosen option price. It therefore depends on the contract, timestamp, underlying input, curve, dividends, quote side, numerical method and model. A price, an implied volatility and a Greek are model outputs with conventions and units; none by itself is an executable trade, a physical probability or a guarantee that a hedge will work.

Seven-step contract, data, valuation, and control process

Classify the claim before choosing a formula. Record issuer or clearing claim, underlying, option type, strike, expiration, last trade, exercise style, official settlement process, cash or physical settlement, multiplier, deliverable, currency and any adjustment. The closed form above is for a European vanilla payoff; American early exercise, adjusted deliverables, barriers and other path-dependent rights require a model that represents those terms.
Lock the valuation clock and units. Set valuation timestamp t₀, expiry or settlement timestamp, timezone, business-day calendar and year fraction T; distinguish last trading time from exercise cutoff and official settlement determination. State whether premium and Greeks are per underlying unit, per quoted point, per contract or for the whole position.
Build synchronized market inputs. Use contemporaneous spot or forward data, strike, a maturity-matched discount factor D(0,T), dividend amounts and dates or a justified continuous yield, stock-borrow and funding assumptions, and option bid, ask and size. A published Treasury CMT is a par yield, not automatically the continuously compounded zero rate r; bootstrap or otherwise construct a consistent discount curve before using it.
Check bounds and matched parity. For the continuous-yield European setup, test max(S₀e^(−qT) − Ke^(−rT), 0) ≤ C ≤ S₀e^(−qT) and max(Ke^(−rT) − S₀e^(−qT), 0) ≤ P ≤ Ke^(−rT). Compare only options with the same exercise, strike, maturity, deliverable and settlement. With known deterministic cash dividends, replace the yield approximation with their discounted cash flows and use the corresponding parity bridge.
Select and run the model explicitly. Black-Scholes-Merton assumes continuous diffusion, constant volatility and rates, frictionless continuous trading, no jumps and sufficient borrowing, lending, shorting and replication. A binomial tree or another numerical method can represent American exercise and discrete events; stochastic-volatility, local-volatility or jump models address different risks but introduce new parameters and calibration choices.
Invert prices into implied volatility carefully. Reject or investigate prices outside the model’s no-arbitrage range; solve bid, midpoint and ask separately with a documented root finder, tolerance and maximum volatility. A finite vanilla European IV is normally unique in the interior because option value rises with σ, but a boundary price can imply zero, infinite or no finite solution. Validate strike-expiry surfaces for stale quotes, crossed inputs and calendar or butterfly arbitrage rather than treating one ATM IV as universal.
Report local risk and reconcile to the market. Calculate delta, gamma, vega, theta and rho with sign, denominator and time convention; preserve the exact market snapshot and model version. Stress jumps, dividend revisions, skew and term-structure moves, borrow, discrete hedging, transaction costs and liquidity. Reconcile theoretical value, executable bid or ask, fills, official settlement and realized cash as separate layers, then independently reproduce material calculations.

Worked examples

Closed-form valuation and parity. Let S₀ = $100, K = $100, T = 1, r = 5.00%, q = 2.00% and σ = 20.00%. Then d₁ = 0.250000, d₂ = 0.050000, N(d₁) = 0.598706326 and N(d₂) = 0.519938806. The call is C = $9.227006 and the put is P = $6.330081. The difference C − P = $2.896925 agrees with $100e^(−0.02) − $100e^(−0.05) = $2.896925; these are per-unit model values before multiplier and costs.
Greeks require units. Using the same inputs, call delta is 0.586851, put delta is −0.393348, and common gamma is 0.018951 delta per $1 underlying move. Vega is 37.901158 per 1.00 volatility, equivalent to $0.379012 per 1 volatility point. Call theta is −$5.089319 per year, or −$0.013943 per calendar day under the simple ÷ 365 display convention; call rho is 49.458109 per 1.00 decimal rate, or $0.494581 per 100 bp. These first- or second-order local sensitivities change when inputs move and must still be multiplied by the contract multiplier and position size.
Market price to an IV interval. Keep S₀ = $100, K = $100, T = 1, r = 5.00% and q = 2.00%. A call market price of $11.123762 inverts to σ = 25.00%, with d₁ = 0.245000 and d₂ = −0.005000. For reference, model values at 20.00% and 30.00% are $9.227006 and $13.020281. If the executable quote is $10.90 bid / $11.35 ask, matched inversions produce 24.410353% bid IV / 25.596190% ask IV; reporting only a midpoint IV hides that executable range.
Known discrete dividend and parity. Let S₀ = $100, K = $95, T = 0.5, continuously compounded r = 4.00%, and a known $2 cash dividend at t = 0.25. Its present value is PV(D) = $2e^(−0.04 × 0.25) = $1.980100, while PV(K) = $95e^(−0.04 × 0.5) = $93.118874. Matched European parity is C − P = S₀ − PV(D) − PV(K) = $4.901026; if C = $8.20, then P = $3.298974. Replacing this dated cash flow with an undocumented continuous q, or applying the equality directly to American options, changes the claim.

Risks and validation controls

Verify the exact series, legal claim and governing product specification before using a model.
Confirm American, European or other exercise rights; a label or underlying type is not conclusive.
Distinguish cash settlement, physical delivery and an adjusted basket, including multiplier and currency.
Lock the official settlement source, calculation time, rounding, correction and disruption rules.
Align spot, option, rate, dividend and borrow observations to the same defensible timestamp.
Distinguish last trade, expiration, exercise cutoff and settlement date when calculating T.
Apply consistent calendars, day counts, annualization and continuous or discrete compounding conventions.
Do not substitute a Treasury par yield mechanically for the required zero rate or discount factor.
Model known dividend amounts and dates; a continuous q can obscure early-exercise and event risk.
Include stock-borrow scarcity, funding, collateral and short-sale constraints when replication depends on them.
Check European bounds and parity only across genuinely matched claims and synchronized executable inputs.
Treat stale, crossed, locked, one-sided or size-deficient quotes as data-quality warnings.
Report bid and ask IVs when the spread is material; a midpoint is not an executable transaction.
Detect root-finder failure, boundary solutions, tolerance sensitivity and arbitrary volatility caps.
Validate surface interpolation and extrapolation for calendar, vertical-spread and butterfly arbitrage.
Stress jumps, earnings, distributions, trading halts and gaps that violate continuous-diffusion assumptions.
Stress stochastic volatility, skew dynamics and correlation between spot, volatility, rates and dividends.
Quantify discrete rebalancing, latency, market impact, fees and inability to trade continuously.
Label every Greek’s sign, scale and time convention, and do not extrapolate a local derivative too far.
Separate model value, executable price, accounting or tax measure, official settlement and realized P&L.

Common misconceptions

“Black-Scholes is the legally correct or executable price.” Contract rules determine the claim and markets determine available trades; the model supplies a conditional benchmark.
“N(d₂) is the real-world probability of expiring in the money.” It is a model-specific risk-neutral quantity under the stated numeraire and assumptions, not a physical forecast.
“Implied volatility is observed directly and predicts realized volatility.” It is inferred from a price through a selected model and can embed risk premia, supply, demand and data error.
“One volatility and one set of Greeks fit every strike and expiry.” Markets exhibit surfaces, and local sensitivities vary with state, time and model.
“Continuous delta hedging removes option risk.” Jumps, discrete trading, liquidity, costs, early exercise, dividend error and model misspecification remain.

Authoritative sources

The Pricing of Options and Corporate Liabilities - original Black-Scholes continuous-time option-pricing and corporate-liability framework.
Theory of Rational Option Pricing - Merton’s no-arbitrage development, assumptions and option-pricing results.
Option Pricing: A Simplified Approach - Cox-Ross-Rubinstein discrete-time model and convergence framework useful for exercise and event modeling.
A Closed-Form Solution for Options with Stochastic Volatility with Applications to Bond and Currency Options - stochastic-volatility extension illustrating model risk beyond constant σ.
Characteristics and Risks of Standardized Options - standardized-option rights, exercise, assignment, settlement and risk disclosures.
S&P 500 Index Options Product Specifications - product-specific evidence that exercise, settlement, multiplier and trading timelines must be identified before valuation.
Treasury Yield Curve Methodology - official explanation of Treasury’s par curve, inputs and construction; a par point is not automatically the model’s zero rate.
Search Filings - discovery portal for issuer filings and pricing supplements; substantive terms must be taken from the identified filing rather than inferred from the search page.