Valuation guides

Basket, spread and rainbow options explained

What options on several assets are, where they are used, worked examples and how they are valued by Monte Carlo and approximations.

Definition

A basket option pays on the weighted sum of several assets. A spread option pays on the difference between two. A rainbow option pays on the best or worst of several assets. The correlation between the assets is a key driver: a basket of correlated assets is more volatile, and a worst-of option is cheaper the less correlated the assets are.

History

Margrabe (1978) priced the option to exchange one asset for another, and Stulz (1982) the option on the best or worst of two. Basket options grew in the 1990s, and worst-of options became core to structured products. Kirk's (1995) spread approximation and Ju's (2002) basket expansion gave fast alternatives to simulation.

Examples

A worst-of call

A call on the worst of three shares, struck at 100% of their starting levels, pays only if all three rise. It is much cheaper than a call on any one of them, and its price rises with correlation.

Variations

Basket option
On a weighted sum, with Ju's expansion alongside.
Spread option
On the difference, with Kirk's approximation alongside.
Rainbow option
Best of or worst of, with Stulz for two assets.
Quanto baskets
With a quanto adjustment for foreign assets.

Valuation methods

The assets are simulated as correlated lognormals, with antithetic paths and a geometric basket as a control variate to cut the error. Greeks use common random numbers. The moment-matched, Kirk, Ju and Stulz approximations are shown alongside.

Models on RiskAtoms

InstrumentPricing models
Rainbow option (best of, worst of)
Monte Carlo (correlated lognormal)
Stulz two-asset rainbow
Basket option
Monte Carlo (correlated lognormal)
Ju (2002) Taylor expansion (basket)
Spread option (quanto too)
Monte Carlo (correlated lognormal)