Title: Reconstructing an economic space from a market metric

Abstract: Using a metric related to the returns correlation, a method is proposed to reconstruct an economic space from the market data. A reduced subspace, associated to the systematic structure of the market, is identified and its dimension related to the number of terms in factor models. Example were worked out involving sets of companies from the DJIA and S&P500 indexes. Having a metric defined in the space of companies, network topology coefficients may be used to extract further information from the data. A notion of "continuous clustering" is defined and empirically related to the occurrence of market shocks.