Skip to content
All library documents

Choosing Equity Risk Factors for Portfolio VaR Sensitivities

Article Quant Q&A · Author: Wane Mamadou
Read the original

Summary

The document explains two ways to represent equity risk when validating value-at-risk for a stock portfolio. One approach models shocks for each individual stock and applies them directly. Another models shocks to shared factors, such as a market or financial index, then estimates each stock’s beta to those factors to translate the shocks into portfolio changes.

The discussion is conceptual and offers no worked calculation, validation results, or guidance on choosing a factor model. Its main caveat is that factor-based estimates rely on assumptions about how stocks respond to common drivers; those assumptions can be made more flexible, but doing so requires a more detailed model.

Key ideas

  • Equity VaR sensitivities can be based on shocks to individual stocks or to common risk factors.
  • A common factor may be a market index or a financial index.
  • Factor-based estimates require each stock’s sensitivity, such as its beta, to the chosen factor.
  • More flexible assumptions about stock behavior require a more detailed model.

Tags

EquitiesRisk managementPortfolio construction
Full text
# Value-at-risk and Equity delta


# Value-at-risk and Equity delta












How to validate value-at-risk calculation on an equity portfolio using equity sensitivities? I don't have trouble doing that for rates instruments or options but I don't know which underlying risk factor can we use on an equity to compute sensitivities.

## Answer by Bram (score 1)

https://quant.stackexchange.com/a/43384

Well, you can either come up with a model that provides shocks on individual equities and then directly use them. Or you can come up with a model that provides shocks on a few risk factors and the you have to determine the sensitivity (beta) of the individual equities towards your common risk factor (which can be a market index or a financial index). From here it's of course possible to go crazy by making less strong assumptions regarding functional behavior.

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.