Style migration: evidence from the European equity markets

Abstract

PURPOSE OF THE STUDY

This paper examines how migration contributes to value and size premiums in stock returns. Migration is defined as stocks moving from one valuation or size category to another. Specifically, this thesis is the first to replicate the methodology of Fama and French (2007a) using European data. Fama and French (2007a) find that migrating stocks form most of the size and value premiums in the US. Therefore, European data gives the opportunity to provide an out-of-sample test for the Fama and French (2007a) study. The secondary objective of this study is to predict migration using the contextual fundamental analysis method of Piotroski (2000). As Piotroski (2000) is able to show that investors can enhance value investing strategies, it is interesting to test whether the same phenomenon drives migration.

DATA

The sample used in this study consists of companies from over 23 countries in Europe between 1980 and 2011. The total amount of firm-year observations is 52,154. Accounting data is acquired from the Worldscope database using Thomson One Banker and market data from Datastream. To be able to fully replicate the Fama and French (2007) method supplemental data for buyouts (35,255 observations during 1980-2011) from SDC Platinum is acquired.

RESULTS

The results of this thesis indicate a significant value effect in Europe during 1980-2011. Further, migration has a significant contribution to the value effect. Additionally, size effect is positive in 1980-1988 but seems to disappear after 1988, which is in line with previous research (van Dijk 2011). In accordance to Chen and Zhao (2009), the size effect appears to be positive if small growth stocks are omitted from the portfolios. Migration also has a major role in the formation of the size effect. The fundamental analysis method of Piotroski (2000) is able to enhance portfolio returns, although the benefits are mainly concentrated on small firms. However, the results concerning the method's ability to predict migrating stocks are mixed. Firms in a very good financial condition have a higher probability to migrate favourably than firms in a poor financial condition.