Spatial dependency has been studied in several research areas, such as environmental criminology, economic geography, environmental sciences, and urban economics. However, it has essentially been overlooked in other subfields of economics and in the field of finance as a whole. A key element at stake is the definition of contiguity. In the context of financial markets, defining a metric distance is not a simple matter. In this article, we explore the notion of spatial dependency by formulating a spatial version of the capital asset pricing model (S-CAPM). Such a model specification makes it possible to account for alternative measures of distance between firms, such as market capitalization, the market-to-book, and other financial ratios. Our model is tested on a panel of 126 Latin American firms. In addition, we derive Value-at-Risk (VaR) measures from our S-CAPM formulation. We complement our discussion with Monte Carlo simulations aimed at quantifying the benefits of diversification in terms of VaR reduction.