Skip to content
All library documents

Constructing a Volatility Mimicking Factor from VIX Changes

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

Summary

The document addresses how to add a volatility factor to a Fama–French style asset-pricing model. It notes that changes in the VIX are not themselves the desired factor and proposes deriving a factor through regression. Regress VIX changes on the excess returns of the base assets being explained, such as the Fama–French portfolios, and use the fitted factor construction for the analysis.

The reply offers a compact methodology rather than a full implementation guide. It does not specify the sample period, return frequency, factor normalization, estimation window, or how to apply the resulting series in subsequent tests. No data source or empirical results are supplied, so researchers would need to make those choices explicit and evaluate whether the constructed factor fits their model and research design.

Key ideas

  • A volatility mimicking factor can be constructed by relating VIX changes to excess returns of the base assets.
  • The base assets in the regression should match the portfolios or returns the asset-pricing model aims to explain.
  • Raw VIX changes are distinguished from the mimicking factor derived through regression.
  • The suggested method leaves frequency, sample, scaling, and validation choices unspecified.

Tags

Factor investingVolatilityStatisticsEquities
Full text
# volatility factor


# volatility factor












I am trying to add a volatility factor to Fama-French factor model. Does anybody know of a source where I can get data for "volatility mimicking factor" or suggest a simple methodology for calculation. (Ang, 2006) introduces a FVIX factor and points that change in VIX cannot be used directly. I am open to trying other volatility factors if someone has a suggestion

## Answer by phdstudent (score 1)

https://quant.stackexchange.com/a/18632

FVIX is not hard to compute. Just regress changes in VIX on excess returs of your base assets (it can be the 25 FF portfolios if those are what you are trying to explain) i.e run the following:

\Delta VIX_t = X_t\beta+\epsilon

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.