Orthant Probabilities for Robust Correlation and Structural Performance Enhancement

Orthant probabilities applied in a two-dimensional framework are used to derive quadrant-conditional financial asset return correlations which fully capture both linear and non-linear components of co-variability. We investigate the potential for employing quadrant-conditional correlations in order to construct portfolios which generate a long-run average portfolio return which is more positive than long-run averages of individual assets' returns. Risk-based, but return-enhancing security selection applications involving assets co-variability criteria for investment management and high-frequency arbitrage trading are discussed