Principal Curves for Statistical Divergences and an Application to Finance

摘要: This paper proposes a method for the beta pricing model under consideration of non-Gaussian returns by means generalization mean-variance and use principal curves to define divergence optimization model. We rely on q-exponential so consider properties divergences which are used describe statistical fully characterize behavior assets. derive minimum portfolio, generalizes Markowitz’s (mean-divergence) approach relying information geometrical aspects distributions Capital Asset Pricing Model (CAPM) is then derived characterization data, all approach. discuss possibility integration our into an adaptive procedure that can be search optimum points finance applications.