Risk-neutral pricing for arbitrage pricing theory (Q779871): Difference between revisions

In this paper the authors study the APT and they focus on some mathematical aspects of this popular financial model. The market model is a two-period model, the prices of assets in both periods are exogenously given and in the second period are random variables defined on a probability space \((\Omega, \mathcal{F}, P)\). Trading strategies are assumed to be members of the Hilbert space \(\ell^2\). It is well-known that in the APT a notion of arbitrage is well-defined and it is related to the behavior of limits of trading strategies and it is termed asymptotic arbitrage. In the third section the authors study carefully the relationship of this notion or of arbitrage with the standard one -- i.e. the existence of an EMM. Making a few reasonable technical assumptions they are able to show, by using well-known results, that the set of EMM \(\mathcal{M}_2\) is not empty. Moreover, by exploiting the properties of \(\ell^2\) and other results from functional analysis they are able to prove a quantitative version of the no-arbitrage condition. In section 4 they study the notion of superreplication price and they are able to prove the standard duality results for this notion which also hold infinite dimension. In section5 they introduce the von Neumann-Morgenstern utility functions and they prove the existence of optimal trading strategies i.e. that the maximization problem of the economic agent actually has a solution. The last section investigates the behavior of varying risk-aversion and the convergence of the reservation price to the superreplication one. My only very minor criticism of this paper is related to the introduction. In that section the authors mention the work of \textit{M. A. Khan} and \textit{Y. Sun} [J. Econ. Theory 110, No. 2, 337--373 (2003; Zbl 1074.91017)]. In my opinion the important point in that paper is the use of nonstandard analysis rather than just the use of a measure-theoretic setting. I wish to add that the paper is very well-written and the proofs are very clear. It is a very good read and a valuable contribution to the current research.