Mieczysław Kowerski
pages: 1-23, JEL classification: C10, C50, G12, Keywords: efficient market hypothesis, Fama-French three-factor model, Generalized Least Squares Method, Abstract: The purpose of the paper is to show that the three-factor Fama-French model can be a good instrument for analysis of investment risk on emerging capital markets if, because of the relatively small number of quoted companies, for calculation of the SMB and HML values we applied division of all companies into four portfolios (contrary to Fama – French who propose division of all companies into six portfolios). The usefulness of the above concept was verified on the Warsaw Stock Exchange. The models estimated with the Generalized Least Squares Method on monthly data within the period 1994 – 2008 have the signs of coefficients which are consistent with those of the Fama-French three-factor model and there is no autocorrelation of disturbances and no ARCH effect. Models are relatively high adjusted. Estimated coefficients are also robust. The models fully confirm the thesis posed by Fama and French that in addition to market risk there are two other risk factors which influence the return on investment. These are: risk associated with investing in small companies and risk connected with investing in companies undervalued by the market.