This paper considers the possible spatial interactions among the probabilities of dying at certain age, on particular regions, in the valuation process of a financial derivative whose underlying variable is an annuity. Taylor's (2001) methodology, which proposes a linear model to study the impact of geographic variations on a diversity of insurance products, is complemented with a spatial model applicable to the valuation of annuity derivatives. The proposed methodology is applied to an Argentinian case, constructing mortality tables that incorporate implied spatial correlations, and evaluate their effects on the price of the derivatives.