Whether to Exclude Stocks with Negative Expected Returns

Summary

The document poses a portfolio construction question: when choosing a subset of stocks with Markowitz mean-variance theory, should stocks with negative average weekly returns be discarded? It provides no proposed selection rule, numerical example, covariance estimates, or answer, so it frames a decision rather than resolving it.

The question points to a central feature of mean-variance optimization: an asset’s expected return is considered together with its covariance with the other holdings and the portfolio’s risk objective. A negative expected return may weigh against inclusion, but by itself does not establish whether the asset worsens an optimal portfolio; its diversification contribution and the investor’s constraints also matter. The discussion gives no information about forecast uncertainty, transaction costs, short-sale constraints, or how expected returns were estimated. Any conclusion therefore requires more inputs and a specified optimization objective than the post supplies.

Key ideas

• The post asks whether negative average weekly returns justify excluding stocks from a Markowitz portfolio.
• Mean-variance selection depends on expected returns together with covariances and the portfolio objective.
• A negative expected return alone does not establish whether an asset belongs in an optimized portfolio.
• The question supplies no covariance estimates, investment constraints, or answer to the selection problem.

Tags

# Construct portfolio with assets having expected negative returns


# Construct portfolio with assets having expected negative returns












I have been asked to select a n stocks among N stocks, to construct a portfolio. Some of them have have negative weekly returns on average. If I want to select these n stocks by constructing an 'optimal' portfolio using Markowitz theory, for example, does it make sense to remove the stocks with negative returns?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.