Skip to content
All library documents

Choosing Industry Factors for a Portfolio Risk Model

Article Quant Q&A · Author: Rcwilkin1993
Read the original

Summary

The discussion considers whether a risk model can include only selected industry factors, using tobacco and healthcare as examples, and whether an additional category for all other industries is required. The answer gives a basic redundancy principle: a factor need not be included when its information can be derived from the other factors. It also cautions that a model limited to a few sectors may be inadequate for representing broader portfolio risk.

As an alternative for investors seeking to avoid exposure to particular sectors, the answer suggests screening those industries out of the portfolio. The exchange is brief and offers no quantitative model, assumptions about factor construction, or empirical validation. Whether omitted industries can be represented by the included factors depends on the model’s structure and data, so the stated principle alone does not establish that two factors are sufficient.

Key ideas

  • A factor may be redundant when its information is derivable from the other factors.
  • A risk model with only a few industry factors may not capture broader portfolio risks well.
  • Screening out unwanted industries is suggested as an alternative to representing them with a limited factor set.
  • The answer provides no detailed model construction or empirical test.

Tags

Risk managementPortfolio constructionEquities
Full text
# Having only 2 Industry risk factors?


# Having only 2 Industry risk factors?












Is it possible to build a risk model that only has 2 industry risk factors? For example, if I wanted just Tobacco and Healthcare industries as risk factors can I do that? If I did that do I have to create a third bucket called other?

## Answer by AlRacoon (score -1, accepted)

https://quant.stackexchange.com/a/46356

If the information from a 3rd factor can be derived from the information in the other 2 factors, then the 3rd factor is redundant and is not necessary.

But as others have commented, probably not a great risk model. If one is uncomfortable with the risk in those sectors, it might be best to use a filter to screen those industries out of your portfolio.

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.