Excess Return✓ Mathematical
◆ The PatternCapturing risk-adjusted excess returns
Alpha (α) is the intercept of a factor regression — the return not explained by systematic risk exposures. Positive alpha means the manager added value beyond factor tilts.
α = Rp − [ Rf + β1F1 + β2F2 + … ]
Sources of alpha: information edges, execution speed, behavioral exploitation, or structural advantages (tax, regulation). Alpha decays — once a signal is widely known, it gets arbitraged away.
// Alpha generation — excess return decomposition
Alpha decay. The half-life of a quantitative signal is typically 2-5 years before crowding erodes it.
Pattern bridge: Feature importance for signal research connects to ML Math.