Skip to content
All library documents

Considering Skewness in Daily Expected-Range Estimates

Article Quant Q&A · Author: c.m.
Read the original

Summary

The document raises a practical question about adjusting daily expected ranges for return asymmetry in SPX. The author describes an existing estimate that blends daily implied volatility with daily realized volatility using different weights, and asks whether trailing skewness over shorter and longer windows could refine those estimates. It also asks whether skewness should be measured from log returns or from daily realized volatility calculated with a modified Garman–Klass-style method.

No answer, tested adjustment, or comparative evidence is included, so the text does not establish that skewness improves range forecasts or specify how to convert it into adjusted sigma values. It frames a research problem rather than presenting a validated strategy. Any proposed weighting scheme would need evaluation against realized outcomes and care around the distinction between skewness of returns and skewness of volatility estimates; the document supplies no results or guidance on that validation.

Key ideas

  • The author estimates daily ranges by combining implied and realized volatility with separate weights.
  • The proposed refinement is to incorporate recent skewness measured over two trailing windows.
  • The document leaves open whether return skewness or realized-volatility skewness is the more useful input.
  • It provides no adjustment formula, validation results, or evidence that skewness improves forecasts.

Tags

OptionsVolatilityStatisticsRisk management
Full text
# Taking skewness into account when determining daily expected ranges


# Taking skewness into account when determining daily expected ranges












I use a method to determine daily expected ranges by combining both daily IV and daily realized vol. with different weights to get the expected range, and it worked pretty accurately. However I want to finetune by taking the skewness into account. My idea is to meassure the 20D and 5D trailing skewness and weight the value of the last 5 days.

Is there any legit way to adjust the daily sigmas by skewness? Which one is better: meassuring the skewness of logreturns or the skewness of daily realized volatility? (I use a modified GKYZ formula)

kind regards (I'm talking about SPX)

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.