Motivating, constructing and testing the Fama-French three factor model on the Johannesburg Stock Exchange
The purpose of this dissertation is to motivate, construct and test the suitability of the Fama and French (1993) three-factor model in pricing equities listed on the Johannesburg Stock Exchange. Before this can be achieved, however, the existence of the size and the value effects needs to be established, and their resistance to risk adjustment with traditional asset pricing models needs to be ascertained. Once, these two empirical facts are documented, the three-factor model is built and tested. Results of Fama and French (1992) can be replicated on the Johannesburg Stock Exchange in that a firm‟s size and its value-growth indicator have reliable power to forecast stock returns. However, the value effect and, in particular, the size effect, attenuate after market microstructure is controlled for. Both effects are found to be independent of one another and the book-to-market ratio is found to be the best value-growth indicator. The static CAPM and an APT variant cannot explain the size and the value effects. This result is robust to time-series and cross-sectional tests. The three factor model of Fama and French (1993), and its variant, are constructed. The models can capture a substantial amount of time-series variation in most assets. When applied to the size and book-to-market sorted portfolios, they are not rejected in the vast majority of asset pricing tests. In tests on ungrouped data, the three factor model can explain the value effect, but not the size effect. However, in cross-sectional tests that use the size and book-to-market sorted portfolios as well as industry portfolios, the pricing errors of the three factor model are not substantially different from the ones obtained from the static CAPM.
MCom (Research) , Faculty of Commerce, Law and Management, University of the Witwatersrand, 2007
JSE , Johannesburg Stock Exchange