3404526: MasterPROJECT 3

Description

-
Covariance forecasting: A comparison of implied volatility and conditional heteroscedasticity models - Group 1

Background:

• OLZ operates “smart indexing” approach
• Academia and practice have shown that market capitalisation weighted portfolios are a bad approximation of the market portfolio
• Alternative approaches have been introduced: minimum-variance, maximum-sharpe, equally-weighted and fundamental indexation
• OLZ applies a minimum-variance optimisation as a proxy for the market portfolio
• Advantage of only estimating variance/covariance of asset returns and not forecasting returns themselves

Lecture Goals

• Students are expected to compare approaches for the estimation of variance/covariance of asset returns
• Primary interest is on the quality of estimates derived from implied volatility
• Above method should be compared to conditional heteroscedasticity models (e.g. ARCH, GARCH, etc.)
• Approaches should be tested in an empirical portfolio setting
• Data set applied is in-line with the interest and practice of OLZ & Partners

Comments

Expected group size: 3-4 students.