Fama-French five factor asset pricing model validation test for Turkey

Five factor asset pricing model which is devised by Fama and French is tested for equities in Borsa Istanbul(BIST) between 2007-2016 years for financial reports and for returns between 2008 April-2018 March by regression analysis. Constant terms resulted out from regression analysis are tested by using GRS test statistics to check that whether they are equal to zero. 157 equities trading in national market in BIST are researched. Mimic portfolios diversified from equities are used as independent variable, and diversified portfolios according to firm size, book value, operational profit, and investment criteria are used as dependent variables. According to regression analysis, for the most of portfolio, null hyphotesis imply that coefficiant of factors are zero could not be rejected for the most of regression models for five factor and sub models including three and four factor. Five factor model is not better than other three and four factor models according to adjusted R square value. Null hyphotesis that constant terms are equal to zero for five factor model could not be rejected by GRS test satatistics for 2x2 and 2x2x2x2 factor models for some portfolios within %90 confidence level. 2x3 factor model is rejected for all portfolios except some models and portfolio selected according to size, operational profit and investment criteria by GRS test. In spite of GRS test, with respect to regression coefficient, Fama-French five factor model is rejected in Turkey for 2007-2016 years for financial tables and stock return data between 2008 April-2018 March for 157 stocks. Two factor model including SMB and market risk premium factor is better than CAPM model.