Asymptotic arbitrage and numéraire portfolios in large financial markets (Q928500)

Large financial market is a sequence of traditional market models with a finite number of risky assets. The paper is devoted to the concept of asymptotic arbitrage and strong asymptotic arbitrage (both of the first kind) on such markets. It is demonstrated that the arbitrage properties of a large market are completely determined by the asymptotic behaviour of the sequence of numéraire portfolios related to small markets. It is known that the numéraire portfolio possesses a number of optimality properties, in particular, it maximizes the expected logarithmic utility and it is uniquely defined that can serve a computational tools for checking asymptotic arbitrage conditions. The correspondent criteria of absence of asymptotic and strong asymptotic arbitrage are expressed in terms of contiguity, entire separation, and Hellinger integrals, provided that these notions are extended to sub-probability measures. In general, the assumptions are not restrictive since in the traditional setting the existence of the numéraire portfolio is implied by the existence of an equivalent local martingale measure. Diffusion market models are also treated and in this case the existence of local martingale measures in the small markets is not required. A discrete-time infinite horizon market model with one log-normal stock is also examined.