Options Greeks

What do the options Greeks actually measure?

An option price depends on a short list of inputs: the underlying price, the strike, the time remaining, implied volatility, the interest rate, and expected dividends. The option Greeks are the derivatives of that price with respect to each input. Each Greek answers one narrow question: if this single variable moves by one unit while everything else freezes, how much does the theoretical value move?

Two consequences follow before any arithmetic. The Greeks price a scenario, so choose the scenario from the chart, the catalyst and the flow, then let the Greeks tell you what it does to the position. And because each Greek holds the other inputs still, no single Greek describes a real session. A stock that rallies while implied volatility drops and a day burns off moves an option through delta, gamma, vega, and theta at once, and the observed price change is the sum of those partial effects plus whatever the first-order set leaves out.

The classification that organizes the rest of this page is the order of the derivative. First-order Greeks measure how option value responds to an input: delta to price, theta to time, vega to implied volatility, and rho to interest rates. Second-order Greeks measure how a first-order Greek responds to an input: gamma is delta with respect to price, charm is delta with respect to time, vanna is delta with respect to volatility, and vomma is vega with respect to volatility.

Where do the Greeks come from, and why are they estimates?

The Greeks on a chain come from a pricing model run on the live quote. The Black-Scholes model maps six inputs to a single theoretical value for a European-style contract. Delta, theta, vega, and rho are the first partial derivatives of that value with respect to spot, time, volatility, and the rate, and gamma is the second derivative with respect to spot. Change one input and both the price and its sensitivities update together, which is why the numbers on a chain stay internally consistent.

Most listed single-name equity options are American-style and can be exercised early, so the closed-form formula does not fit them exactly. Pricing engines reach for the binomial option pricing model, which builds a lattice of possible prices and tests at every node whether immediate exercise beats holding. That node-by-node test is what captures the early-exercise right, and it is why a dividend-paying American contract can carry Greeks that a European formula would misstate. With enough steps, the binomial value for a European contract converges to the Black-Scholes value.

Spot and premium are quoted, strike and expiration are contract terms, and rates and expected dividends are selected market inputs or assumptions. Future realized volatility is not observable, so traders invert the model against the traded price and solve for the volatility that makes the two agree. Every Greek on the screen therefore moves with that implied volatility mark, the dividend assumption, the rate and the model. Match those inputs before comparing Greeks across two platforms; a different volatility mark or dividend gives a different Greek for the same contract.

What are the units, and why does per share versus per contract matter?

Every Greek discussed here is quoted per share, matching the convention for option premiums. A standard US equity option covers 100 shares, so the dollar effect on one contract is the quoted Greek multiplied by 100. Losing track of that multiplier is the fastest way to size a position wrong by two orders of magnitude.

Delta is the clearest case. Position delta is the per-share delta multiplied by the contract count and the multiplier, with the sign of the position applied:

Position delta = option delta × contracts × 100

Ten calls at 0.40 delta carry 0.40 × 10 × 100 = 400 positive share-equivalents. Long calls and short puts are positive, short calls and long puts are negative, and one long share is exactly +1 delta. That shared unit is what lets a trader express a call spread, a short put, and a block of stock as one directional number.

Read volatility and rate changes in percentage points. Vega of 0.15 corresponds to about $0.15 per option unit when implied volatility moves from 30% to 31%, or $15 for a standard contract. Rho of +0.32 corresponds to about $0.32 per unit when the rate moves from 4% to 5%. A 25-basis-point move is 0.25 percentage points, giving 0.32 × 0.25 = $0.08 per unit or $8 per contract.

How do the first-order and second-order Greeks behave?

Matched European calls and puts share gamma, vega, vanna, and vomma under the same pricing inputs. Put-call parity makes their value difference linear in spot and independent of volatility. Long vanilla calls and puts carry positive gamma and, outside the deep in-the-money European case in the table, pay for it in daily theta: the curvature that profits from movement costs time value.

Near expiry, gamma concentrates around the strike, because delta swings across its whole range within a narrow price band. Vega peaks near the money and grows with time to expiration. A weekly at-the-money contract is a gamma instrument. A one-year at-the-money contract is a vega instrument. Both sit on the same chain and both are called calls. The theta versus vega comparison works through the same split from the seller’s side.

What the second-order set adds

Charm measures delta changes as time passes. A call at 0.52 delta with charm of −0.018 per elapsed calendar day has a one-day delta estimate of 0.502. Across 80 standard contracts, that changes exposure from 4,160 to 4,016 share-equivalents, a reduction of 144. Use the derivative with respect to elapsed time for this projection; a derivative with respect to remaining time carries the opposite sign.

Vanna connects delta with implied volatility and vega with the underlying price. It flips sign near the money, at a point that volatility, time and rates set. Rising volatility lifts the delta of a far out-of-the-money call and lowers the delta of a deep in-the-money call. Read the sign at each contract, then multiply vanna by the volatility-point change to estimate the resulting delta adjustment.

Vomma, also called volga, measures how vega changes with implied volatility. In Black-Scholes, vomma = vega × d1 × d2 ÷ volatility when volatility is measured as a decimal. Its sign follows d1 × d2. At spot equal to strike with zero rates and dividends, d1 is positive and d2 is negative, so vomma is negative. For a contract with positive vomma, rising volatility increases vega. If vega rises from $0.10 to $0.16 per option unit over a 10-point volatility increase, the average-vega estimate is $1.30 per unit, compared with $1.00 from the starting vega alone.

How does one day of profit and loss attribute across the Greeks?

Greeks P&L attribution is the second-order Taylor expansion of the pricing model, applied backward to explain a realized result:

Option P&L per share = (delta × change in spot) + (0.5 × gamma × change in spot squared) + (theta × days elapsed) + (vega × change in implied volatility)

A worked example

A trader holds 5 long XYZ 100-strike calls with 30 days to expiration. XYZ is $100.00 and the calls mark at $2.80 per share, so the premium at risk is $2.80 × 100 × 5 = $1,400. The opening Greeks per share are delta 0.52, gamma 0.055, theta −0.045 per day, vega 0.11 per volatility point, and rho 0.04.

Position delta at the open is 0.52 × 5 × 100 = 260 share-equivalents.

Over the session XYZ closes at $102.00, implied volatility falls from 28.0 percent to 26.5 percent, one day elapses, and the rate input is unchanged. The buckets, per share:

The four active buckets total $0.94 per option unit, or $470 across five standard contracts. The observed $0.95 mark change produces $475, leaving $5 to reconcile. That $5 is what the four buckets leave out: the cross sensitivities and volatility curvature. Reprice the full position at the new spot, volatility and date to capture it.

For the price move alone, delta changes from 0.52 to 0.52 + 0.055 × 2 = 0.63, taking five standard contracts from 260 to 315 share-equivalents. To include volatility and time, apply vanna and charm at the same starting state. For example, vanna of −0.010 per volatility point and a 1.5-point IV decline add 0.015 delta, taking the estimate before charm to 0.645, or 322.5 share-equivalents. Reprice the full position for the final hedge target.

How do the Greeks read on a live tape?

On an option chain, the Greeks are columns, and the useful skill is scanning them as a profile instead of reading one at a time. Delta identifies which strikes are effectively stock-like and which are lottery-shaped, faster than comparing dollar premiums. Gamma marks the band where hedging demand concentrates, usually within a strike or two of spot on the front expiration. Theta and vega on the same row show what the contract costs to hold and how much of its price is a volatility bet rather than a directional one.

Flow prints add the size and the side that the chain alone cannot show. A large print in a weekly at-the-money contract is a gamma and theta statement: high curvature, high decay, a short window for the thesis to work. The same premium spent in a six-month contract is a vega statement, since the longer-dated contract carries multiples of the weekly’s volatility sensitivity and a fraction of its daily decay. When a stock print or a futures print arrives alongside an option print, it often signals a delta hedge rather than a directional view, which changes how the tape should be read.

Aggregating contract-level Greeks across the chain is what produces the per-strike gamma ladders and dealer-positioning panes traders watch for levels. Cboe describes the mechanism directly for market makers hedging their books: gamma is how much the hedge has to be adjusted as the underlying moves. Long-gamma hedging sells strength and buys weakness, which dampens ranges. Short-gamma hedging does the reverse and amplifies them. Those aggregates assign each side of the open interest to dealers or customers, which makes a heavy gamma strike the level to watch. Price action at that strike confirms whether the hedging is there.

The other live read is realized against implied. A delta-neutral long-option position rehedged through the day is running gamma scalping: under a local constant-gamma approximation, each price leg contributes one half times position gamma times the square of that leg, so a full round trip out and back books that amount twice, and the accumulated scalps have to out-earn the theta paid. In a $100 name with 10 long calls at 0.50 delta and 0.05 gamma, hedged with 500 short shares, position gamma is 0.05 × 10 × 100 = 50 share-deltas per $1. A move to $102 lifts delta to 0.60 and the trader sells 100 shares at $102, a leg worth 0.5 × 50 × 2.00 × 2.00 = $100; a return to $100 pulls delta back to 0.50 and the trader buys 100 shares back for a second $100 leg, so the shares capture 100 × $2.00 = $200 on the round trip. Whether that beats the decay is the implied versus realized volatility question, and the delta hedging versus gamma scalping comparison separates the one-time hedge from the ongoing harvest. Count the costs before calling the scalp a win: the $200 round trip has to cover the calls’ decay and any IV change, plus financing and execution on the share hedge.

What do traders get wrong about the Greeks?

Reading delta as the probability of a profitable trade. Read delta as a rough proxy for the risk-neutral chance of finishing in the money, and judge profit from the breakeven. Delta and that probability coincide only under specific model assumptions; skew, dividends, and the gap between risk-neutral and real-world probabilities pull them apart. A contract can finish in the money and still lose money after premium, so a 0.30 delta is not a 30 percent chance of winning.

Multiplying today’s theta by the days remaining. Project decay from the repriced contract day by day, because theta is an instantaneous rate that changes as the position ages. For an at-the-money contract it accelerates into the final days, so a five-day projection built from today’s figure understates the decay. The intrinsic versus extrinsic value split is why: only the extrinsic portion decays, and how much of the premium is extrinsic depends on moneyness.

Treating delta neutral as risk free. Manage the exposures that remain after the hedge. Zero net delta cancels first-order directional risk for small moves only; the position still carries gamma that reshapes delta before the next rehedge, vega that reacts to a volatility reprice, theta that erodes long premium daily, financing costs on the share hedge, and early-assignment risk on short options.

Mixing per-share and per-contract figures. Convert every leg to the same unit before you aggregate. A theta of −0.08 is $8 per contract per day, not $0.08, and a multi-leg position that mixes per-share and per-contract legs produces a hedge that is wrong by a factor of 100.

Comparing second-order Greeks across platforms without checking units. Confirm the unit and the sign convention first. A vanna quoted per one volatility point is 100 times smaller than the same sensitivity quoted per full 1.00 decimal move in volatility, and charm conventions differ on whether time advances or remaining life declines. A correct magnitude with a reversed sign is worse than no number.

Extending one Greek across a large move. Reprice the position at the new spot, volatility, and date. Every Greek is a local slope at a single point: delta drifts as gamma acts on it, gamma changes as the underlying and time change, and vega itself moves through vomma.

Applying a European calculation to an American contract with a dividend. Price it on a lattice. The early-exercise right and a pending ex-dividend date can both move the correct value and its derivatives; the lattice checks both at every node, and a closed-form European formula cannot see them.

What to learn next

The efficient order is model first, then units, then interactions. Start with the Black-Scholes model to see where the numbers originate and why implied volatility is the one input that has to be solved for, then the binomial option pricing model for the American-style and dividend cases that dominate single-name equities.

From there, work the everyday set in the order it usually bites: delta for exposure and the hedge ratio, gamma for how fast that exposure moves, theta and time decay for what holding costs, vega for the volatility axis, and rho for the long-dated and financing-heavy cases where rates stop being background noise.

The second-order set is best read after a position has surprised you once. Charm explains overnight delta drift, vanna explains a hedge that breaks on a volatility move, and vomma explains a vega estimate that missed on a large repricing. Then close the loop with Greeks P&L attribution, which turns the whole set into a daily ledger, and with the applied pages: delta neutral for isolating an exposure, gamma scalping for monetizing realized movement, and hedging for the broader practice of paying a known cost to bound an unknown one.