luger_04

This paper surveys recent developments in the theory of option pricing. The emphasis is on the interplay between option prices and investors' impatience and their aversion to risk. The traditional view, steeped in the risk-neutral approach to derivative pricing, has been that these preferences play no role in the determination of option prices. However, the usual
lognormality assumption required to obtain preference-free option pricing formulas is at odds with the empirical properties of financial assets. The lognormality assumption is easily reconcilable with those properties by the
introduction of a latent state variable whose values can be interpreted as the states of the economy. The presence of a covariance risk with the state variable makes option prices depend explicitly on preferences. Generalized option pricing formulas, in which preferences matter, can explain several well-known empirical biases associated with preference-free models such as that of Black and Scholes (1973) and the stochastic volatility extensions of Hull and White (1987) and Heston (1993).

Title:	Option Prices, Preferences, and State Variables *
Number:	04-18
Author:	René Garcia, Richard Luger and Eric Renault
Issue Date:	November 2004
Abstract:	This paper surveys recent developments in the theory of option pricing. The emphasis is on the interplay between option prices and investors' impatience and their aversion to risk. The traditional view, steeped in the risk-neutral approach to derivative pricing, has been that these preferences play no role in the determination of option prices. However, the usual lognormality assumption required to obtain preference-free option pricing formulas is at odds with the empirical properties of financial assets. The lognormality assumption is easily reconcilable with those properties by the introduction of a latent state variable whose values can be interpreted as the states of the economy. The presence of a covariance risk with the state variable makes option prices depend explicitly on preferences. Generalized option pricing formulas, in which preferences matter, can explain several well-known empirical biases associated with preference-free models such as that of Black and Scholes (1973) and the stochastic volatility extensions of Hull and White (1987) and Heston (1993).