The thesis contributes to the literature in the following ways: First it contributes to the body of literature by extending our knowledge on the predictive ability of alternative Unconditional methodologies. Second it adds to the body of litareture by providing practical tests so as to assess the performance of Conditional models. Third the thesis extends our knowledge on the sensitivity of utilising different portfolio formation criteria, while testing both Unconditional and Conditional asset pricing inferences. Fourth it contributes to the body of literature by extending our knowledge on Unconditional and Conditional beta models and their comparative performance. Fifth the thesis adds to the existing literature by estimating the Industry cost of capital, using the following different models, Unconditional, Conditional, the Arbitrage Pricing Model and the Capital Asset Pricing model. Thus provides empirical evidence using a practical application, estimation of the Industry cost of capital, of which model provides a better description of UK returns. Chapter 4 introduces the portfolio returns used in the thesis and examines the size, price earnings ratio, dividend yield effect and their interactions. The time-series of the primary portfolios start in 1956 and ends in 1996. We find that for the 1976-1996 period, that the dividend yield and PE effect subsume the size effect. However the PE effect subsumes the dividend yield effect and it is the PE effect that is the most dominant. The best documented of all stock market effects, the small-firm premium went into reverse during 1989-1996. The size effect lives on, but for the latest decade, it is the largest firms that outperform the smallest ones by 10.26% per annum. Chapter 5, which examines Unconditional models, aims to examine the predictive ability of alternative Unconditional methodologies. Another objective that is explored is the sensitivity of results to different grouping techniques, of size; PE ratio and dividend yield portfolio groupings. The third issue examined entails the identification of priced factors in the UK market, over a twenty year of period, (1976-1996), and for a data-set (approximately 6000 companies), which provides a complete history of firms traded on the London Stock exchange, inclusive of Unlisted securities market. We find that that the choice of one methodology over another has important implications and that there is a sensitivity of results to different portfolio groupings. Chapter 6, which examines Conditional models, i. e., conditioned on a set of instrumental variables, models the dynamic behaviour of portfolio returns using a Conditional Asset Pricing Model and examine the behaviour of macroeconomic risk premiums over time. We provide practical tests of Conditional Asset Pricing Models and forecast (i) the sign of the price of risk using the probit model, (ii) the magnitude of the price of risk and (iii) portfolio returns for the size, PE ratio and dividend yield portfolios. We find that the instrumental variables show ability to predict variation of the price of risk of the return on FTSE, S&P 500, unexpected UK stock exchange turnover, change in money supply, imports, inflation and portfolio returns. Chapter 7 compares first Unconditional (constant) and Conditional (time-varying & conditioned on a set of instrumental variables) beta models and second the CAPM and the APM, estimates the industry cost of capital. We find differences, between constant unconditional betas and conditional betas cost of capital. The average Mean Square Error (MSE) for the conditional betas are smaller compared to constant betas. Moreover we find that the CAPM has larger MSE not only compared to the APT model with conditional betas, but with APT with unconditional betas. The Conditional beta model provides the best description of UK returns. We also run Monte Carlo simulations and test the statistical significance of the errors of the Conditional beta model. We find the errors to be statistically insignificant.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:340343 |
Date | January 2000 |
Creators | Koulafetis, Panayiota |
Publisher | City University London |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://openaccess.city.ac.uk/8108/ |
Page generated in 0.0016 seconds