This thesis uses a sample of German stocks traded on German stock exchanges to study the cross-section of average stock returns. Six different factors are constructed from scratch and proposed as potential explanatory variables able to explain the variation in average returns. These factors are related to different firm characteristics that potentially pose risk. In particular, size, value, momentum, profitability, investment, and covariation with the market portfolio are used in different combinations in the search for a factor model that is able to adequately explain return variations. The Fama and French  three-factor model, Carhart  four-factor model, Fama and French  five-factor model, and a six-factor model that combines all factors are analyzed and tested using Fama and MacBeth  regressions and the Gibbons, Ross, and Shanken  test. This thesis finds that the five- and six-factor models are superior to the three- and four-factor models. The models are able to explain variations in average returns, but they seem to have one minor weakness related to firms with low book-to-market ratios. Five out of six risk premia are not strong enough, and a more parsimonious model may be a one-factor model which only includes the market risk premium.
|Uddannelser||Cand.merc.fin Finance and Investments, (Kandidatuddannelse) Afsluttende afhandling|