City Research Online

Abstract

The aim of this thesis is to investigate the nature of determinants of equity returns as suggested by the CAPM model, in particular, alphas, betas and equity premium and to outline implications for investment managers that statistical and structural analysis of the aforementioned variables may suggest. The thesis contributes to the existing literature in the following areas. First, it addresses the question of predictive power of historical risk-adjusted portfolio performance measures on determining future equity returns in short and long term hoirsons. Second, it investigates the stability of beta coefficients and its impact on portfolio risk and seasonality in equity returns. Third, it assesses the question of dividend yield as determinant of portfolio alphas. Finally, it addresses the question of a common factor that may be influencing movements of equity premiums across European markets. All the aforementioned empirical work is the first of this kind, at least to our knowledge, in the UK. In Chapter One we provide an indirect test of alpha stability. We test if past alphas, information ratio and alpha-to-beta ratio of positive and negative alpha portfolios can be used to determine future portfolio returns. We find that chosen portfolio performance measures do not have any predictive power in the short term investment horisons. However, in the longer term horisons of 24 to 36 months, we document the mean reversion in our portfolio returns and conclude that one can use historical measures of performance to predict returns in the longer run. In Chapter Two we proceed to investigate if stocks with higher beta (systematic risk) also exhibit higher instability in betas as well, thus causing even greater risk for investors. We also examine the seasonality effect in the UK size-based portfolios and try to relate it to seasonality in betas. Our findings suggest that higher beta stocks do have more time-variant betas. Additionally, we find that equity returns are much higher in December-April than in May-November period but we find no robust evidence that such seasonality in returns is due to seasonality in betas but rather due to investors' psychology. In Chapter Three, we assess the relationship between excess returns and dividend yields in the UK market. The econometric analysis reveals U-shaped yield-return relationship in the 1980s and quadratic, bell-shaped, relationship in the 1990s. It seems that such a change in the relationship is driven by the change in the returns pattern of small size stocks in the 1990s. We find no evidence of the tax effect as the explanation of yield-return relationship that we observe. In Chapter Four we try to identify what may be the common determinant of equity risk premium across European markets. We test for the serial correlation in the stock market returns and the results suggest that serial correlation is not in the level of returns but in the volatility of returns. Hence, if shocks to returns and in turn equity premium are persistent, there can be a scenario of a world-wide shock, which may influence the equity premium across countries in the similar manner driving them in the same direction. The overall findings of the thesis are indicating instability of CAPM determinants of UK equity returns. If investors are aware of these instabilities, they can adjust theirinvestment strategies accordingly and generate excess returns on their investment

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HG Finance
Departments:	Bayes Business School > Finance Doctoral Theses Bayes Business School > Bayes Business School Doctoral Theses

Export

Downloads