Abstract:In this note, we consider European options of type $h(X^1_T, X^2_T,\ldots, X^n_T)$ depending on several underlying assets. We give a multidimensional version of the result of Breeden and Litzenberger \cite{Breeden} on the relation between derivatives of the call price and the risk-neutral density of the underlying asset. The pricing measure is assumed to be absolutely continuous with respect to the Lebesgue measure on the state space.
What problem does this paper attempt to address?
The paper primarily aims to address the issue of state price density representation in multi-asset European option pricing. Specifically, the authors Jarno Talponen and Lauri Viitasaari extend the classic results of Breeden and Litzenberger on single-asset European options to the multi-asset case.
In financial mathematics, option pricing is a core problem. However, in many practical models, option pricing often cannot be expressed in closed form, thus requiring different methods to solve it, such as partial differential equations (PDE), stochastic calculus methods, Monte Carlo simulations, etc. Breeden and Litzenberger's work shows that by calculating the second derivative of the European call option price with respect to the strike price, the risk-neutral density of the underlying asset can be obtained. This finding is of great significance for static hedging.
The main contribution of this paper is to propose a multi-dimensional version of the Breeden-Litzenberger representation theorem, that is, if the price distribution of the underlying assets is absolutely continuous with respect to the Lebesgue measure, the state price density can be derived through the partial derivatives of rainbow options prices. Here, rainbow options are options that depend on multiple underlying assets, and their payoff function is:
\[ h_p(\mathbf{x}, K, K) = \left( \sum_{i=1}^n \left( (x_i - K_i)^+ \right)^p \right)^{1/p} - K \]
By introducing this multi-dimensional representation, the paper covers a wide class of models and does not require the assumption that the pricing measure is unique, making the theory applicable to both complete and incomplete market models as well as discrete or continuous time scenarios. Additionally, the paper provides specific methods on how to extract state price density from different types of option prices, thereby offering new tools for financial derivatives pricing.