Eigenvalue
Also called: eigenvector, eigenvalues and eigenvectors, spectral decomposition
An eigenvector is a direction a matrix stretches without rotating, and the eigenvalue is how much it stretches. For a covariance matrix, the eigenvectors are the independent directions of risk and the eigenvalues are how much variance sits in each. That's why a negative eigenvalue is a red flag: it means the correlation matrix is inconsistent and implies a portfolio with negative variance.
Want more than a definition? Learn it in Math Masters →