Options Greeks Explained: Delta, Gamma, Theta, Vega, and Rho

Editorial responsibility: Oliver Sinz · Last verified: October 26, 2026

The option Greeks measure how much an option’s theoretical value changes when one input to a pricing model changes while every other input stays fixed. Delta measures the effect of the underlying price, gamma the effect of that price on delta itself, theta the passage of time, vega implied volatility, and rho interest rates. Each one is the slope of a model value at the current point. None of them is a forecast of what the option will do.

Rules of thumb such as “delta is the probability of expiring in the money” hold only under specific conditions. This guide defines each Greek with consistent units, shows on one option chain how the Greeks change with moneyness, time, and option type, adds them up across a position, and tests them against full repricing.

All prices and Greeks in this article are hypothetical model values, not market quotes. Nothing here is a recommendation to trade.

What the option Greeks measure

Mathematically, each Greek is a partial derivative: the response of the model value to a small change in one input with the others frozen. The Options Industry Council (OIC) defines them the same way, as theoretical changes with all other pricing factors held constant (OIC: Delta; OIC: Theta).

Three conventions matter before any number makes sense:

• Per share or per contract. Option chains usually quote Greeks per share. A standard U.S. equity option covers 100 shares, so a delta of 0.54 is about 54 shares of exposure per contract.
• Signs flip for short positions. A short option is a negative quantity, so a short call has negative delta and a short put positive delta (OIC: Delta). Short options have negative gamma, and stock has delta but no gamma (OIC: Gamma). Short options also have negative vega and usually positive theta.
• Greeks belong to a model. They depend on the pricing model and every input: price, strike, time, implied volatility, interest rate, dividends, and exercise style.

The example: one option chain, five Greeks

Hypothetical assumptions:

• Stock XYZ at $100.00, 45 calendar days to expiration
• 25% implied volatility at every strike, 4.0% continuously compounded risk-free rate, no dividends
• European-style Black–Scholes–Merton values (Black and Scholes, 1973; Merton, 1973)
• Theta per calendar day (annual theta ÷ 365), vega per volatility point, rho per percentage point, all per share

Three patterns stand out:

1. A call and a put with the same strike share gamma and vega, and their deltas differ by exactly 1. For European options on a stock without dividends, a long call plus a short put on the same strike behaves like stock bought with borrowed money: delta 1, no gamma, no vega. This is put-call parity at work.
2. Their thetas differ by the interest on the strike. The 100 put loses about 1.1 cents a day less than the 100 call: one day’s interest on the present value of the strike, 4% × $99.51 ÷ 365 ≈ $0.011.
3. Gamma, theta, and vega concentrate near the money. In-the-money options behave more like stock, with deltas nearer ±1 and the largest rho.

Standard U.S. stock options are American-style (OIC: Options Exercise FAQ). That changes nothing for these calls, because XYZ pays no dividend, but it changes the puts slightly, so the parity relationships above become approximate: a binomial tree (Cox, Ross, and Rubinstein, 1979) values the American 100 put at $3.29 instead of $3.25, with a delta of −0.47 instead of −0.46.

Delta: exposure to the underlying price

Delta is the estimated change in an option’s value for a $1 move in the underlying (OIC: Delta). The 100 call’s delta of 0.54 means a $1 rise adds about $0.54 per share, or $54 per contract. Delta is also a hedge ratio: shorting 54 shares against one long 100 call neutralizes the position for small moves.

Delta moves with every input:

• Moneyness and time. Deep in the money, deltas approach 1 for calls and −1 for puts; far out of the money, they approach 0. Both pulls strengthen as expiration nears. With a positive rate and no dividends, as here, at-the-money calls sit above 0.50, increasingly so with more time: 0.52 with 7 days left, 0.54 with 45, 0.61 with one year, and 0.66 with two years. In this model the 45-day strike with a 0.50 delta is about $100.88, not $100. A large enough dividend yield reverses this: at the same rate and volatility, a one-year at-the-money call on a stock yielding 6% has a delta of about 0.49.
• Implied volatility. Higher implied volatility tends to pull deltas toward 0.50 (OIC: Delta), as it does here: at 30% instead of 25%, the 110 call’s delta would rise from 0.16 to 0.21 and the 90 call’s would fall from 0.90 to 0.86. That is a tendency, not a rule; the direction depends on moneyness, time, rates, dividends, and the volatility level. From 25% to 50%, for example, the one-year at-the-money call’s delta rises from 0.61 to 0.63, away from 0.50. This cross-sensitivity is called vanna.
• Time alone. Over a weekend (three calendar days, Friday to Monday) with XYZ unchanged, the 110 call’s delta drifts from 0.162 to 0.152 and the 90 call’s from 0.903 to 0.910, so a delta-hedged position can drift out of balance without any price move. This is called charm.
• Dividends. A continuous dividend yield q multiplies delta by e^(−qT) and also lowers d1 in the delta formula (Merton, 1973). A 2% yield would lower the 100 call’s delta from 0.540 to 0.527.

Delta is not the probability of finishing in the money

OIC notes that some traders treat delta as the probability that an option expires in the money. In the Black–Scholes–Merton model, a call’s delta is N(d1), while the model’s own probability of finishing in the money is N(d2), where d2 = d1 − σ√T. The two are close only when volatility times the square root of time is small:

For short-dated options at moderate volatility, the shortcut is off by a few points; for a two-year at-the-money call at 50% volatility, it is off by more than 27 points. Both figures also come from the pricing model, which assumes the stock grows at the risk-free rate, so neither is a forecast of real-world odds.

Gamma: how fast delta changes

Gamma is the change in delta for a $1 move in the underlying (OIC: Gamma). The 100 call’s gamma of 0.045 predicts that its delta rises from 0.540 to about 0.585 if XYZ moves to $101; full repricing gives 0.5845.

Gamma is what makes option values curved. For a move of ΔS, it adds about ½ × Γ × ΔS² to the delta estimate, a term that is positive for a long option whichever way the stock moves. Gains accelerate and losses slow down; short options have the mirror image. For the 100 call, per share, with time and implied volatility unchanged:

On a $5 move, delta alone misses by more than $0.50 and gamma fixes almost all of it. On a $10 move, about 1.1 times the model’s one-standard-deviation move over 45 days ($8.78), the estimate is off by $0.33 up and $0.08 down, because gamma itself changes along the way.

Gamma also changes with time, in opposite directions for different strikes. With XYZ held at $100:

At-the-money gamma climbs steeply into expiration, when delta can swing between 0 and 1 as the stock crosses the strike (OIC: Gamma). The out-of-the-money 110 call’s gamma peaks weeks before expiration and then fades, because a $10 gap becomes ever less likely to close.

Theta: options time decay, and why it is the price of gamma

Theta is the instantaneous rate at which value changes as time passes with nothing else changing, usually quoted per day (OIC: Theta). The 100 call’s theta of −0.044 means about $4.42 per contract per calendar day. Three corrections to the usual rules follow.

Decay accelerates for at-the-money options, not for all options. In the table above, the 100 call’s theta is −$0.044 per day with 45 days left and −$0.266 per day with one day left. That last figure is a local rate, not the final day’s loss: the call is then worth $0.53, all of it gone at expiration if XYZ stays at $100. The loss is about twice the quoted theta because an at-the-money option’s value shrinks roughly with the square root of the time left, so the decay rate keeps rising through the last day. The 110 call’s dollar theta peaks about 60 days before expiration, at roughly 2.6 cents a day, and falls to about 1.2 cents with 14 days left, because little value remains. Yet that 1.2 cents is an instantaneous rate of 22% of the option’s value per day. Decay accelerates in dollars for at-the-money options and in percentage terms for out-of-the-money ones.

A long option can have positive theta. A deep in-the-money European put can be worth less than its intrinsic value, because the holder must wait to receive the strike. With XYZ at $100, the 45-day 120 put is worth $19.49 against $20.00 of intrinsic value, and its theta is +$0.007 a day. An American-style put with the same terms is different: its model exercise boundary sits near $102.60, so at $100 immediate exercise is optimal, and the put is worth exactly $20.00 with a delta of −1 and zero gamma and theta. Exercise style changes the Greeks, not only the price.

Calls and puts with the same strike decay at different rates, by the interest on the strike, as the chain showed.

Theta pays for gamma

In the Black–Scholes–Merton model, theta, gamma, delta, and value are tied together by the equation Black and Scholes derived (1973): Θ + ½σ²S²Γ + rSΔ − rV = 0, with theta per year and no dividends (with a yield q, rSΔ becomes (r − q)SΔ). For the 100 call, the daily theta of −$0.0442 splits into −$0.0387 from the gamma term (−½σ²S²Γ ÷ 365) and −$0.0055 from the interest terms ((rV − rSΔ) ÷ 365).

That identity implies a breakeven move. A delta-hedged long option gains about ½ × Γ × ΔS² from a move of ΔS. Within the model’s local approximation, and including interest on the premium and the stock hedge, that gain exactly offsets the day’s decay when the daily move equals σ × S ÷ √365: $1.31, or 1.31%, for XYZ at 25% volatility. Larger moves leave the hedged long option ahead before costs; smaller moves leave it behind. Because the gamma part of annual theta is always −½σ²S² × gamma, this breakeven is the same for every option on the same stock at the same volatility.

For a delta-hedged option seller, theta is therefore compensation for bearing negative gamma, earned only if the stock’s actual moves turn out smaller than those priced in. Studying S&P 500 index options, Bakshi and Kapadia (2003) found that delta-hedged long option positions underperformed zero on average, which they interpret as evidence of a negative volatility risk premium. That is evidence about one market and sample, not a forecast.

Vega: exposure to implied volatility

Vega is the change in value for a 1-percentage-point change in implied volatility, which OIC describes as a 1% or 100-basis-point change (OIC: Vega). The 100 call’s vega of 0.139 means a rise from 25% to 26% adds about $13.94 per contract. Long calls and puts both have positive vega.

Vega grows with time to expiration, roughly with its square root: the at-the-money call’s vega is 0.055 with 7 days left, 0.139 with 45, 0.383 with one year, and 0.520 with two years. As a share of the option’s value, it moves the other way, from about 3.9% per point at 7 days to 2.9% at two years.

Vega is itself nonlinear. For the at-the-money call it is nearly constant: a 5-point rise adds $0.697, exactly as predicted. For the out-of-the-money 110 call, vega rises with volatility: a 5-point increase adds $0.47 against an estimate of $0.43, and a 10-point increase adds $1.00 against $0.86. The sensitivity of vega to volatility is called volga, or vomma.

Within Black–Scholes–Merton, vega is a well-defined derivative: it compares the option’s value under two different constant-volatility assumptions. What the standard model does not describe is how volatility itself evolves, whether as stochastic volatility or as shifts in the implied-volatility surface. Implied volatility does change, even with no move in the underlying (OIC: Vega), and rarely by the same amount at every strike and expiration, so a vega figure measures exposure to one assumed change, usually a uniform one.

Rho: exposure to interest rates

Rho is the change in value for a 1-percentage-point change in the risk-free rate: positive for long calls, negative for long puts (OIC: Rho). Parity shows why rates push calls up relative to puts: a European call minus a put equals the stock minus the present value of the strike, and a higher rate lowers that present value.

Rho is small for short-dated options and grows with time. The 45-day at-the-money call’s rho is $0.062 per point, 1.7% of its value. At one year it is $0.49, and at two years $0.96, or 5.4% of value; the two-year put’s rho is −$0.89. Repricing the two-year call at 5% instead of 4% adds $0.97, so rho is close to linear. OIC also points out that the rates used in pricing differ across clearing firms, brokers, and even accounts.

How option Greeks are calculated

For a European-style option with a continuous dividend yield q, the Black–Scholes–Merton Greeks have closed forms (Merton, 1973). S is the stock price, K the strike, T the time to expiration in years, r the risk-free rate, σ the volatility, N the standard normal distribution, and φ its density, with d1 = [ln(S/K) + (r − q + σ²/2)T] ÷ (σ√T) and d2 = d1 − σ√T:

Divide vega and rho by 100 for per-point units, and theta by 365 for calendar days. For the 100 call, d1 ≈ 0.100, so N(d1) = 0.540 and φ(d1) = 0.397; gamma is 0.397 ÷ (100 × 0.25 × 0.351) = 0.045, and vega is 100 × 0.397 × 0.351 ÷ 100 = 0.139 per point. American-style puts, and calls facing dividends, have no general closed form; their Greeks come from numerical methods such as binomial trees.

Why two option Greeks calculators can disagree

The same 100 call yields different-looking numbers under different conventions:

The raw per-share vega and the per-contract, per-point vega share a number but not a meaning. Calculators also differ in:

• Day counting. OIC notes that pricing models account for weekends, so options decay seven days over five trading days, but that there is no industry-wide method for decaying options (OIC: Theta). Repricing the 100 call one calendar day later lowers it by $0.0444; a weekend of three calendar days lowers it by $0.135.
• Model. European formulas, binomial trees for American exercise, continuous dividend yields, and discrete cash dividends all produce different Greeks, especially for in-the-money puts and around ex-dividend dates.
• Inputs. The choice of implied volatility (bid, ask, midpoint, or a smoothed surface), interest rate, and time measurement all change the output.

Adding up the Greeks across a multi-leg position

A position’s Greek is the sum over legs of quantity × 100 × the per-share Greek, with short legs entering as negative quantities. In this model, stock contributes one unit of delta per share and nothing else. The sum is exact at the current point if every leg’s Greek is measured against the same risk factor, holding the same inputs constant; the legs need not share an implied volatility. Delta and gamma measure sensitivity to the common stock price with each leg’s implied volatility held fixed, while the sum of leg vegas measures a parallel shift that raises every leg’s implied volatility by the same number of points.

Hypothetical collar: own 100 shares of XYZ at $100, buy one 45-day 95 put, and sell one 45-day 105 call, for a net credit of about $0.43 per share. Per collar:

Totals are computed from unrounded values, so they can differ by a cent from the sum of the rounded legs. At $100 the collar looks simple: about 43 shares of delta and little else. The totals hide two large, offsetting option exposures, with two consequences.

Net Greeks change sign across the price range. With 45 days left:

Near the put strike the collar is long gamma and vega and pays time decay; near the call strike it is short both and collects it. The swings grow as expiration nears: with 15 days left, gamma is +6.95 at $95 and −6.53 at $105.

Net vega measures only a parallel volatility shift. A collar vega of −1.54 says a 2-point rise in implied volatility at both strikes costs about $3; repricing gives −$2.98. Independent moves in the legs’ volatilities are a different risk. If the 95 put’s implied volatility rises 2 points while the 105 call’s falls 2 points, a steepening skew, the collar gains $47.49, while net vega predicts nothing. Leg-by-leg vegas get it right (+$47.42). Summing vega across expirations likewise assumes a parallel shift across maturities.

Greeks versus realized changes in option prices

Combined, the Greeks give the standard approximation for a change in value:

ΔV ≈ Δ × ΔS + ½ × Γ × ΔS² + Θ × days + vega × ΔIV (points) + ρ × Δr (points)

Test it on the 100 call. Suppose one week passes, XYZ rises $5 to $105, and implied volatility falls from 25% to 22%:

The gap arises because the Greeks changed along the way. Once the call was $5 in the money with 38 days left, its vega had fallen from 0.139 to about 0.106 (at 25% implied volatility; about 0.100 at 22%), so the volatility drop cost less than the starting vega implied. Higher-order Greeks such as vanna, charm, and volga measure these effects. Per contract, the Greeks explain $254 of a $263 model change.

That test is generous, because it compares the model with itself. Real option prices add more:

• Inputs move together. For equity options, implied volatility tends to fall when the stock rises (Hull, Fields Institute talk abstract). Testing mainly on S&P 500 options, Hull and White (2017) show that because price and volatility moves are correlated, the standard delta does not minimize the variance of a hedged position. As an illustrative assumption, if implied volatility fell 0.5 points for every $1 rise, the 100 call’s effective delta would be about 0.54 + 0.139 × (−0.5) ≈ 0.47, and a $5 rise would add $2.94 instead of the $3.23 of a constant-volatility repricing.
• Market prices do things a one-factor model cannot. Using S&P 500 options, Bakshi, Cao, and Chen (2000) found that call prices often fell while the index rose, and calls and puts often rose or fell together, even after controlling for time decay and market microstructure. No one-factor model, in which an option’s value moves only with the stock price and time, predicts that.
• Frictions and jumps. Bid-ask spreads, price ticks, overnight gaps, discrete dividends, and early exercise all produce changes that a local slope does not capture. OIC itself cautions that delta changes constantly during trading and typically will not predict an option’s exact price change (OIC: Delta).

What the Greeks can and cannot tell you

In options trading, the Greeks are a compact snapshot of a position’s risk. They put different strategies on a common footing, show which input dominates a position’s risk, and size hedges against small moves.

They do not forecast: a positive theta is not an expected profit, and a delta is not a real-world probability. Large moves, or several inputs changing at once, call for full repricing across a grid of prices, dates, and volatilities rather than a sum of Greeks.

Key takeaways

• The Greeks are sensitivities of a model value to one input at a time, and they depend on the model, inputs, and units used.
• Delta is a hedge ratio and share equivalent. It approximates the model’s probability of finishing in the money only for short-dated options at moderate volatility, and neither figure is a real-world forecast.
• Gamma makes option values curved: delta alone misprices moderate moves, and delta plus gamma drifts on larger ones.
• Time decay accelerates in dollars for at-the-money options and in percentage terms for out-of-the-money ones. Quoted theta is a local rate, not a day’s actual loss near expiration. Deep in-the-money European puts can have positive theta. In the model, theta is the price of gamma: a delta-hedged long option breaks even before costs when the stock moves about σ × S ÷ √365 per calendar day.
• Vega grows with time and is nonlinear for out-of-the-money options. Rho is small for short-dated options but material for long-dated ones.
• Position Greeks add exactly at the current point when each is measured against the same risk factor. Net vega describes only a parallel volatility shift, and net figures can hide large offsetting exposures that change sign with the stock price.
• Realized option prices also reflect correlated inputs, microstructure, and jumps. Greeks describe risk; they do not predict returns.

Primary sources

• Options Industry Council: Delta · Gamma · Theta · Vega · Rho · Options Exercise FAQ
• Black, F., and Scholes, M. (1973). The Pricing of Options and Corporate Liabilities. Journal of Political Economy, 81(3), 637–654.
• Merton, R. C. (1973). Theory of Rational Option Pricing. Bell Journal of Economics and Management Science, 4(1), 141–183.
• Cox, J. C., Ross, S. A., and Rubinstein, M. (1979). Option Pricing: A Simplified Approach. Journal of Financial Economics, 7(3), 229–263.
• Bakshi, G., Cao, C., and Chen, Z. (2000). Do Call Prices and the Underlying Stock Always Move in the Same Direction? Review of Financial Studies, 13(3), 549–584.
• Bakshi, G., and Kapadia, N. (2003). Delta-Hedged Gains and the Negative Market Volatility Risk Premium. Review of Financial Studies, 16(2), 527–566.
• Hull, J., and White, A. (2017). Optimal Delta Hedging for Options. Journal of Banking & Finance, 82, 180–190. · Hull, J. Optimal Delta Hedging, Fields Institute talk abstract.