|
|
|
| Untersuchte Arbeit: Seite: 5, Zeilen: 6-14 |
Quelle: Schmidt 2003 Seite(n): 5, 6, Zeilen: 5: 25-27; 6: 5-11 |
|---|---|
| Indeed, standard portfolio selection is usually based on this approach, the Markowitz mean-variance theory and the Sharpe-Lintner-Mossin capital asset pricing model (CAPM). Further, the dependence structure of the asset returns is solely described by the Pearson correlation of the returns. This suffices in case the asset return distributions are multivariate normal. However, if we leave the Gaussian or more generally the elliptical world, a mere consideration of the correlation matrix often explains the dependence structure in a quite unsatisfying way. Pitfalls like a non-existing correlation or zero correlation for dependent random variables may occur. Especially the dependence structure of extreme events is usually poorly or incorrectly described by this measure. | [Seite 5: 25-27]
Standard portfolio selection is usually based on the Markowitz mean-variance theory of risk and return, from 1952 on, and the Sharpe-Lintner-Mossin capital asset pricing model (CAPM) of 1964-66. [Seite 6: 5-11] In particular, the dependence structure of the asset returns is solely described by the covariance matrix Σ. This suffices in case the asset-return distributions are multivariate normal. However, if we leave the Gaussian world, a sole consideration of the covariance matrix often explains the dependence structure in a quite unsatisfying way. Pitfalls like a non-existing covariance or zero covariance for dependent random variables may occur. Especially the dependence structure of extreme events is usually poorly or incorrectly described by the covariance matrix. |
Die Quelle ist nicht angegeben. |
|