The Capital Asset Pricing Model is amongst the most used models to calculate risk and expected
rates of return of a portfolio. This project thesis aims to optimise a portfolio applying variance
and CVaR as measurements of risk. The objective is to compare both of these methods. The
optimisation is applied to a model portfolio consisting of eight assets out of dierent asset classes.
We nd that both the mean-variance and mean-CVaR portfolio optimisations with a long-only
constrained portfolio lead to slightly dierent results. However, the obtained results may vary
if dierent assets are taken into consideration or other constraints are added.