I am trying to compare the performance of the compositions of a single portfolio determined by unconstrained mean variance optimization, minimum variance optimization (expected returns equal to 0 in the constraints) and 1/n allocation via backtesting. However, the issue is that mean variance optimization yields portfolio weights in terms of expected portfolio return, while the latter two do not. How could I compare their performance? Keep in mind this I am aiming for a single naive comparison, not a statistically significant one of a larger sample, as I am writing a high school level paper on this topic.

  • 1
    $\begingroup$ Nice project you are doing. How did you estimate the Expected Returns for MVO? Historical information ? That is OK, provided the historical period used is prior to the period you will use for backtest. Then I don't see any problem. $\endgroup$
    – nbbo2
    Feb 25, 2023 at 12:49


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.