Return Sampling Frequency and Volatility Estimation

Summary

The document raises whether annualized volatility estimated from monthly returns is typically lower than estimates based on daily returns. The questioner considers using daily observations to increase the sample size, while worrying that the measured volatility may rise as a result. No study or empirical calculation is presented to resolve this comparison.

The response points to a chapter on covariance matrix estimation in a portfolio management book. It says the chapter discusses daily, weekly, monthly, and yearly returns and how these frequencies affect covariance estimates and the risk calculated from them. This is a reading recommendation rather than a direct explanation of the mechanics or a reported result. The document does not establish that one sampling frequency systematically yields lower annualized volatility, nor does it offer a specific estimation method or address the tradeoffs of serial correlation and market microstructure effects.

Key ideas

• The question concerns whether annualized volatility estimates differ across return sampling frequencies.
• Daily returns provide more observations, but the document does not establish that they increase annualized volatility.
• Covariance matrix estimation and derived risk can depend on whether returns are daily, weekly, monthly, or yearly.
• The response recommends a book chapter for further study but provides no empirical evidence or definitive conclusion.

Tags

# Daily vs Monthly vs. other return for volatility calculation?


# Daily vs Monthly vs. other return for volatility calculation?












I thought I read/heard somewhere that annualized volatility, using monthly returns vs daily returns is usually lower. With that said, I can't seem to find any papers on this.

Does anyone have any studies that looks at annualized volatility using different return calculation time frames?

I'm tempted to use daily returns because it increases the number of observations, which I would like, but if it increases volatility, I'd prefer to keep volatility lower.

Thanks

## Answer by Paul Brennan (score 2)

https://quant.stackexchange.com/a/72149

I have not seen a paper on this result but there is a good section in a book "Modern Investment Management: An Equilibrium Approach" by Bob Litterman

The chapter is Covariance Matrix Estimation by Giorgio De Santis, Bob Litterman, Adrien Vesval, and Kurt Winkelmann. This talks about the issues around using daily weekly, monthly and yearly returns to estimate covariance matrices and the risk that can be calculated with them.

It was very informative, but unfortunately the book is not cheep.

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.