Why Exclude Financial Firms from Quantitative Studies?

I often get insightful questions from readers.

A recent question was as follows:

…applied works most often exclude Financials and Utilities in their universe under study. You exclude similar constituents in your “valuation horse race” study. Why? Might the inclusion of these sectors materially impact the conclusions?

This is a fair and important question.

First, we actually only exclude financials in our work (this is a mistake in the writing, although, the exclusion of utilities doesn’t affect results).

Second, the primary reason we exclude financials:

• Because Eugene Fama and Ken French said so
  • We exclude financial firms because the high leverage that is normal for these firms probably does not have the same meaning as for non-financial firms, where high leverage more likely indicates distress
  • See The_Cross-Section_of_Expected_Stock_Returns pg 429 in the “data” section
  • This is the “common practice” in research. This is not a good reason to do something, so we always do our tests with and without financials, however, to stay in line with common practice and to ensure that skewed financials fundamentals data doesn’t drive our results, we exclude them in published and working papers.

Hope that answers any questions folks might have…