Non-U.S. Multi-Factor Data Sets Should be Used with Caution
36 Pages Posted: 4 Feb 2014 Last revised: 3 Nov 2015
Date Written: November 3, 2015
Due to the success of the Fama/French three-factor model, many factor sets for non-U.S. stock markets have been estimated and applied. Exporting a specific factor model from the U.S. to another country seems to be an easy and well-defined task. We use the example of Germany to illustrate that this is not the case. The factor sets offered by seven providers who all intend to exactly replicate the four-factor model with German data take the country-specific institutional settings into account in different ways. As a consequence of these differ-ences and of quality problems in the underlying databases, the factor time series differ consid-erably and produce very different results in two standard applications. We can well imagine that similar problems exist for other countries, especially because four of the seven providers of factors for Germany offer identically calculated factors for a large number of countries. In addition to noting problems, we give advice to providers and users of non-U.S. factor sets.
Keywords: data quality, factor model, risk factors, MSCI indices, Germany
JEL Classification: G12
Suggested Citation: Suggested Citation