Option Properties, Greeks, and Delta Hedging
An option has a payoff at expiry, but it also has a market value before expiry. That value changes when the underlying price moves, when expected volatility changes, and when time passes. The Greeks describe these local changes.
They are useful for two related tasks:
- explaining what drives an option’s price;
- estimating how a position’s risk changes after a market move.
The earlier options article explains calls, puts, premiums, payoffs, and profit. This article looks at the risk of holding those positions before expiry. The calculations use one hypothetical European call to keep the example consistent.
Option value is more than immediate payoff
Suppose a call has a strike of $100 and the underlying is worth $100. Exercising it immediately would produce no benefit, so its immediate payoff is zero. The option can still have a positive market value because the underlying may rise before expiry.
That extra value is often called time value. It reflects the remaining opportunity to receive a favourable payoff. The market price of the option is therefore different from its payoff at one particular moment. At expiry, there is no time left and the option is worth its expiry payoff, if any.
Several inputs affect the value of a stock option. The following statements hold with the other inputs unchanged, and they describe the usual direction rather than an exact amount.
| Input increases | Call value | Put value | Reason |
|---|---|---|---|
| Underlying price | Usually rises | Usually falls | A call benefits from a higher sale price; a put benefits from a lower one |
| Strike price | Usually falls | Usually rises | The strike is the price used when the option is exercised |
| Volatility | Rises for a long option | Rises for a long option | More possible movement helps a payoff with limited downside |
| Interest rate | Usually rises | Usually falls | The financing and discounting effects differ between calls and puts |
| Expected dividends | Usually falls | Usually rises | A dividend normally reduces the underlying price on its ex-dividend date |
Time requires a qualification. An American option with more time has at least as many exercise choices as a shorter option with the same other terms. European options can behave differently when fixed dividends or other cash flows occur during the extra period. In practice, the result also depends on the contract and the assumptions used by the pricing model.
A pricing or risk system should therefore store more than an option’s latest premium. It needs the contract terms, underlying price, expiry, strike, rates, dividends, volatility convention, contract multiplier, and the as-of time of the market data. The market-data article explains why these inputs must describe one consistent market snapshot.
The five Greeks
A Greek is a sensitivity: it estimates how much the option value changes when one input changes and the other inputs remain fixed. It is a local estimate, not a guarantee. The sign also depends on the position. A short option has the opposite Greek from the corresponding long option.
For a position, a useful implementation rule is:
The multiplier is important. A sensitivity quoted as $0.50 per share is $50 per standard 100-share contract and $5,000 for 100 such contracts.
Delta: sensitivity to the underlying
Delta is the change in option value for a small change in the underlying price:
Here, is the option value and is the underlying price. If a call has delta , a $1 rise in the underlying suggests an option-value increase of about $0.56 per underlying unit, assuming the move is small and the other inputs do not change. A long call normally has positive delta. A long put normally has negative delta. A share has delta ; a short share has delta .
Delta is sometimes described as a probability-like number, but it is primarily a price sensitivity. It is not generally the exact probability that an option will finish in the money. The two ideas can be related under particular models, but they answer different questions.
Gamma: how delta changes
Gamma measures how quickly delta changes as the underlying moves:
A long vanilla call or put normally has positive gamma. A short option normally has negative gamma. A share has zero gamma because its delta remains or .
Gamma is why a delta hedge is temporary. If a call’s delta is 0.60 or higher. The number of shares that offsets the position has changed, even though the option quantity has not.
For a small move, the main effects can be written as:
The first term is the linear delta effect. The second term is the curvature effect from gamma. The last term is the effect of time passing, where is measured per unit of time and is the elapsed time. This approximation is useful for explaining risk, but it becomes less accurate for large moves or changing market inputs.
Vega: sensitivity to implied volatility
Vega measures sensitivity to the option’s implied volatility:
Implied volatility is the volatility input that makes a pricing model agree with an observed option price. It is not the same as a simple average of past returns. A long vanilla option normally has positive vega: a higher implied volatility increases the value of the right to benefit from a large move. The underlying share has zero vega.
Market systems often quote vega for a one-percentage-point volatility move. A move from 20% to 21% is one percentage point, not a one-percent relative increase from 20% to 20.2%. A portfolio can also have separate vega exposures for different strikes and expiries because volatility is usually represented by a surface, not one number.
Theta: sensitivity to time
Theta measures the change in value as time passes with the other inputs held fixed. It is commonly reported per calendar day or per trading day. Long options often have negative theta, which is why this effect is called time decay.
Time decay is not linear. It often becomes faster as expiry approaches, especially for an at-the-money option. A long option can lose value during a quiet day even when the underlying price does not move. A rise in implied volatility or another market input can offset that loss, so an observed daily P&L is not determined by theta alone.
Theta is different from delta. A trader can trade the underlying to offset delta, but time passes regardless of the underlying position. A position with positive gamma often has negative theta, and a position with negative gamma often has positive theta, although the exact relationship depends on the portfolio and model.
Rho: sensitivity to interest rates
Rho measures sensitivity to the interest-rate input:
For a standard stock option, a long call usually has positive rho and a long put usually has negative rho. The effect is often larger for longer-dated options because there is more time for financing and discounting to matter.
Production systems use curves and conventions rather than one universal interest rate. A portfolio with options at different expiries can have exposure to several parts of the curve. Rho is therefore a useful summary, but a risk report may need rate buckets as well.
Delta as a hedge ratio
Assume a trader has written options. The option position has negative delta: if the underlying rises, the short options lose value. The trader can buy shares to create positive delta and reduce the combined sensitivity.
If the option position has total delta , buying 559.6 shares gives a combined delta of approximately zero. Since shares are normally traded in whole units, the trader might buy 560 shares. The hedge is called delta-neutral at that valuation point.
This does not mean that the position is risk-free. It means that the first-order effect of a small underlying move is approximately offset. The position can still have gamma, vega, theta, rho, funding, liquidity, jump, and model risk.
A small delta-hedging example
Consider a hypothetical European call with these assumptions:
- underlying price: $100;
- strike: $100;
- time to expiry: three months;
- continuously compounded interest rate: 4% per year;
- assumed volatility: 20% per year;
- no dividends;
- 10 contracts, with a 100-share multiplier.
A simple Black–Scholes calculation gives an illustrative call value of $4.4852 per share and these sensitivities:
| Quantity | Per share or option unit |
|---|---|
| Delta | |
| Gamma | per $1 move in the underlying |
| Vega | per one volatility percentage point |
| Theta | per calendar day |
| Rho | per one interest-rate percentage point |
These are model outputs, not market guarantees. The model assumptions are deliberately simple so that the hedge mechanics are visible.
The position contains 1,000 option units: 10 contracts multiplied by 100 shares. Because the trader is short, its total delta is:
The trader buys 560 shares. The stock position offsets almost all of the option position’s local delta.
Now suppose the underlying moves immediately from $100 to $102. Holding the other model inputs fixed, the call value rises from about $4.4852 to $5.6818. The short option position therefore loses:
The 560 shares gain:
The combined result is approximately -$76.59 before fees, funding, dividends, and other effects. The hedge reduced the loss, but it did not make it zero. The move was large enough for gamma to matter, and the initial hedge was rounded to whole shares.
The gamma estimate explains much of the residual. The short position’s gamma is approximately . With a $2 move:
This is a second-order approximation. The exact result differs because gamma changes during the move and the hedge used 560 rather than 559.6 shares.
After the move, the model delta of one call is about 0.6361. The target hedge is therefore about 636 shares. The trader buys 76 more shares. At the new price, that trade exchanges \7,752 of gross cash. It is not itself a $7,752 loss: the shares become part of the hedge. The actual cost of rebalancing comes from the bid-ask spread, commissions, market impact, financing, and the price at which the trade is executed.
What the hedge does and does not achieve
Delta hedging controls the local linear exposure to the underlying. It does not remove the position’s curvature. A short-gamma position can lose when the underlying moves sharply in either direction, even if it started delta-neutral. A long-gamma position has the opposite curvature exposure, but it commonly pays negative theta while the market remains quiet.
Rebalancing more often can keep delta closer to zero, but it also creates more trades and costs. No schedule removes the risk of a price jump between observations. The hedge result also depends on the volatility surface, rates, dividends, contract terms, market-data timestamp, and pricing model used by the system.
For a risk engine, a Greek is a local sensitivity at a defined valuation state. It is an input to hedging and scenario analysis, not a replacement for either one.