Using Sector-Neutral Long–Short Portfolios to Compare Companies
Summary
The document asks how to compare company performance while reducing the influence of macroeconomic changes such as interest rates, crises, and inflation. It also raises the idea of expressing company financial metrics relative to other companies instead of in currency terms, but does not develop a method for adjusting financial statements or inflation measures.
The response describes a common investment approach: compare firms within the same market or sector and build a long–short portfolio based on a view about which company will perform better. Since firms in a sector often share risk exposures, this can reduce exposure to sector-specific factors and focus the position more on relative company performance than on the direction of the broader market or economy. The answer frames this as a practice used by fundamental portfolio managers. It gives no specific normalization formula, portfolio construction rules, empirical evidence, or guarantee of macro neutrality; the degree of neutrality depends on how well the selected firms' risk exposures actually match.
Key ideas
- Comparing firms within the same sector can help control for shared sector risk exposures.
- A long–short position expresses a view on relative company performance.
- Relative positions may reduce dependence on broad market or economic direction.
- The proposed approach does not provide a formula for normalizing financial statements or removing inflation.
- Sector matching does not guarantee full neutrality to macroeconomic or other risk factors.
Tags
Full text
# How to normalise companies relative to each other? Remove the effect of macro-economy? # How to normalise companies relative to each other? Remove the effect of macro-economy? I want to focus on companies-performance only and as much as possible remove the effect of macro-economy. Interest rate changes, crisis, inflation etc. There are 2 goals: - Simplification of analysis for value investing. When I look at company financial report numbers and historical time series - I want to see simple curves, without jumps and falls caused by macroeconomy. - Better account for inflation. I don't know how to account for inflation correctly (I don't trust CPI and other measures of inflation). So it would be better to completely ignore USD and measure numbers in something else, thus eliminate the inflation. The general idea - measure company numbers in numbers of other companies, not in USD. That's probably not a new idea and I guess some people already did something like that. What are the popular approaches to do that? Maybe calculating market averages and measuring company financial reports in market average metrics? Looking for links examples and ideas. ## Answer by Chris (score 1) https://quant.stackexchange.com/a/55643 The most common way this is applied is via trading within a given market or sector. For instance, biotech or telecom stocks are largely exposed to similar risk factors so creating a long/short portfolio using whatever unique info you think you have you're roughly neutral to sector-specific risk factors and are just betting company A will do better than company B (relatively easier) versus what the sector or economy as a whole will do (directionally, much more difficult). This is in fact a version of what a lot of professional fundamental PMs do.
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.