A Family of Maximum Entropy Densities Matching Call Option Prices

TL;DR

This paper explores the Buchen-Kelly maximum entropy density matching market call prices, introduces a fast algorithm for its computation, and demonstrates convergence properties as more call prices are incorporated.

Contribution

It presents a novel root-finding algorithm for the Buchen-Kelly density and analyzes its uniqueness and convergence in a family of entropy-maximizing densities.

Findings

The Buchen-Kelly density is the unique continuous maximum entropy density matching call prices.

The algorithm efficiently computes the density with convergence guarantees.

As call prices increase, the density converges to the true market density in relative entropy.

Abstract

We investigate the position of the Buchen-Kelly density in a family of entropy maximising densities which all match European call option prices for a given maturity observed in the market. Using the Legendre transform which links the entropy function and the cumulant generating function, we show that it is both the unique continuous density in this family and the one with the greatest entropy. We present a fast root-finding algorithm that can be used to calculate the Buchen-Kelly density, and give upper boundaries for three different discrepancies that can be used as convergence criteria. Given the call prices, arbitrage-free digital prices at the same strikes can only move within upper and lower boundaries given by left and right call spreads. As the number of call prices increases, these bounds become tighter, and we give two examples where the densities converge to the Buchen-Kelly density in the sense of relative entropy when we use centered call spreads as proxies for digital prices. As pointed out by Breeden and Litzenberger, in the limit a continuous set of call prices completely determines the density.