Interpreting Return Units in the Kenneth French Data Library

Summary

The note explains how to read return values in downloadable portfolios from Kenneth French’s Data Library. Its central point is that a reported value such as 1.37 represents a return of 1.37 percent, rather than a decimal return of 137 percent. This matters when importing the data into calculations, since treating percentage-point values as decimal fractions would greatly overstate returns.

The answer offers a practical sanity check: compare the magnitude of the downloaded values with what is plausible for the product and sampling frequency. The note does not document the library’s file formats in detail, discuss conversions for downstream analysis, or give a broader validation procedure. Its guidance is based on the respondent’s experience and is framed around the return field in the data under discussion, so users should still check the documentation for a particular file.

Key ideas

• Values in the discussed French Data Library return fields are percentages, not decimal fractions.
• A value of 1.37 denotes a 1.37 percent return.
• Check whether return magnitudes are plausible for the product being analyzed.
• Confirm the format for the specific file before using it in calculations.

Tags

# Are Fama-French returns in percentages?


# Are Fama-French returns in percentages?












I am working with data from Kenneth French Data library. While I am aware that they use simple returns, I remain unsure about their format in downloadable files.

This is an extract of the file "Portfolios formed on Book to Market" showing daily portfolio returns based on their Book to Market ratio.

I do not see it anywhere if the values are simple returns in absolute terms (1.37=137%), or in percentages (i.e. 1.37%).

Thanks!

## Answer by Matthew Gunn (score 5, accepted)

https://quant.stackexchange.com/a/69837

In my experience working with Ken French's data library, a "1.37" in the return field would correspond to a 1.37% return (with rounding to the hundredths place).

You can of course do some simple checks as to whether the order of magnitude makes sense in whatever particular product you're downloading.

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.