Econometric Tests for Comparing Portfolio Strategies

Summary

The document poses an econometric question about comparing two long-only portfolio construction strategies tested on the same data. It contrasts testing a single strategy’s performance against the market using Jensen’s alpha with the separate problem of determining whether one strategy outperforms another.

The proposed possibility is to regress one strategy’s excess returns on the other’s and test whether the intercept differs from zero. However, the document contains only the question and no answer, data, or evidence that this regression is appropriate. It therefore identifies a comparison problem but does not establish a method; any analysis would need to account for the strategies’ shared sample, return dependence, and the specific definition of outperformance.

Key ideas

• Comparing two strategies is distinct from testing either strategy’s alpha against the market.
• The document asks whether one strategy’s returns should be regressed on the other’s.
• It proposes testing whether the regression intercept differs from zero but supplies no resolution.
• A valid comparison must account for dependence between strategies evaluated on the same data.

Tags

# Comparing two portfolio construction strategies


# Comparing two portfolio construction strategies












Suppose that I have two long-only portfolio construction strategies and that I backtest both of them on the same data. If I wanted to find out whether one of these methods would have outperformed the market, then I would regress the resulting excess returns of that method against the excess returns of the market and look at Jensen's alpha.

How would I compare (using econometric methods) whether one strategy outperforms the other? Would it make sense to regress one against the other and test whether the intercept is different from $0$?

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.