International research indicates that portfolios formed on various stock characteristics produce different returns. These stock characteristics under which portfolios are formed include; past stock returns, size, earnings yield, leverage, and book to market ratios. The differences in returns cannot be explained using the Capital Asset Pricing Model (CAPM), suggesting that the CAPM may not be an appropriate asset pricing model. In Australia a number of market anomalies have been reported, although there is limited research on the book to market effect and conflicting results using price momentum. Thus the First objective of this thesis is to study whether size, book to market ratios and price momentum are related to returns in Australia over the time period 1982 to 2006. Evidence that the CAPM cannot explain asset returns has led to the development of empirically derived asset pricing models, which are based on the Intertemporal CAPM (ICAPM) or the Arbitrage Pricing Theory (APT). These models include the market risk premium and either two or three additional factors, which capture the size (SMB), book to market (HML) and momentum (UMD) effects. In the United States of America (USA) these empirically derived models explain a significant proportion of the difference in returns between portfolios formed on various stock characteristics. In Australia there have been few studies that have investigated whether these empirically derived asset pricing models can explain the cross-section of portfolio returns. Therefore, the second objective of this thesis is to study whether the CAPM or a three- or four-factor model can explain the cross-section of portfolio returns. In this thesis portfolios are formed based on size, book to market ratios and price momentum. Results indicate three effects: First, portfolios made up of stocks whose market capitalisation is less than $3 million out-perform portfolios formed on larger stocks. This result would indicate that the size anomaly is not economically exploitable in Australia. Second, portfolios formed on past price momentum indicate that past winners continue to outperform past losers. However, past winners do not outperform the majority of portfolios formed on momentum. This suggests that to exploit the momentum effect would require continual short positions of the loser portfolio, which may be unsustainable in Australia. Finally, portfolios formed on stocks with high book to market ratios outperform portfolios formed on stocks with low book to market ratios. This result is consistent across all size groupings, suggesting that it may be possible to exploit the book to market effect. Results in Chapter 5 show that the negative relationship between market capitalisation and returns, and positive relationship between book to market ratios and price momentum and returns, are present even after exposure to market risk is taken into account. These results indicate a strong positive return during the months of January and July which are not fully captured by market risk, market capitalisation, book to market ratios or price momentum. These results present compelling evidence that these three characteristics are important determinants in explaining differences in portfolio returns. Such results lead to testing whether the CAPM or a three- or four-factor model can better explain the time-series and cross-sectional variation of portfolio returns. These results indicate that both the three- and four-factor models provide increased explanatory power over the CAPM. In particular, there is a strong relationship between size and the SMB factor. In contrast to previous Australian studies, there is strong evidence that the HML factor plays a significant role in explaining the time-series and cross-sectional variation in portfolio returns. Results indicate a strong monotonic relationship between book to market portfolios and the HML factor. Finally, these results indicate that a number of portfolios have significant exposure to the UMD factor. Overall, results indicate that a four-factor model explains the greatest proportion of portfolio returns. Even though the four-factor model has increased explanatory power over the CAPM, it has a number of limitations. That is, results indicate that the model cannot fully explain the size, value and momentum effects in Australia. This would suggest that future research is required in this area, in particular research should focus on applications of the conditional CAPM, consumption CAPM, and the intertemporal CAPM.
