There is a considerable body of literature that examines the behaviour of institutional investors as a potential source of market price movement. Most existing studies focus on the market timing abilities of active fund managers and find mixed evidence for their fund timing skills. However, few studies have investigated fund manager timing abilities within segments of the market, such as factor timing and sector timing. This study investigates the style timing behaviour of US domestic equity funds existing at any time during the period 1_2-2002. Specifically, I examine the timing activities of actively managed mutual funds within different market segments based on such established systematic risk factors as size, book-to-market, momentum, and across different fund styles such as aggressive growth, growth and income, and small company funds, etc. Mutual fund timing strategy can be viewed as the fund manager's response to hislher private information regarding future factor premiums. Instead of directly observing how fund managers make their timing decisions, an alternative approach is to look at the direct outcomes of their decisions, which are related to the factor timing loadings derived from a factor timing model. I significantly expand on the work of Bollen and Busse (2001) and Volkman (l~ by combining systematic risk factors unique to equity markets with timing factors unique to actively managed portfolios. Within this empirical timing-activity evaluation framework, I additionally investigate fund timing behaviour in the context of Morningstar star rating performance record, investment objectives, fund age, turnover, and load expense, etc. This Ph.D. is an original contribution to the literature of fund timing activities, which seeks to contribute to our understanding in terms of investigating mutual fund managers' timing strategies with respect to specific systematic risk factors and their evolution over time. This research has important implications both for extant asset pricing theories and for practitioners especially in the evaluation of portfolio performance and investigation of fund managers' timing activities.

This thesis consists of four main research papers in the area of portfolio design, performance and evaluation. The thesis is motivated to empirically test the main models in the literature on the Egyptian Emerging Capital Market (EECM), and this contributes to our understanding on the applicability of these theories to markets outside the environment of developed economies. Specifically, this thesis is organised in six chapters. Chapter (1) as the introductory chapter contains the aims and scope of the study, as well as the research methodology and the major findings. Chapter (2) provides a survey of the key literature on portfolio performance evaluation models, in order to identify the leading theoretical and empirical issues in this area. Chapter (3) provides an empirical study for examining the risk and return characteristics of actively managed equity funds in the EECM. Chapter (4) provides another empirical study focused on deriving and designing optimal portfolios from within alternative investment to assist the investor in rationalizing his investment decisions. In response to the weak performance of both mutual funds and securities, together with the low level of diversity in the mutual funds, Chapter (5) uses the APT model to examine the relationship between the economic fundamentals - especially inflation, the exchange rate, interest rate, government borrowing and gross domestic product (GDP) - and the performance of both stock market and mutual funds, as well as the causality between them in the context of the APT model. Chapter (6) presents a summary and conclusion as well as eleven ideas for future research.

This thesis sheds light on evaluating portfolio performance in terms of investment styles the portfolios follow. Investment styles imply that a fund invests in a distinct group of stocks that share some common characteristics. Investment managers with similar investment styles, who select securities based on same principle criteria, are likely to perform more like each other than like the overall market, or like managers with different styles. This thesis aims to provide an investigation into the effects of styles on portfolio performance. To this end, the characteristics of various investment styles are analyzed, and then the validity of benchmarks used to measure portfolio performance is examined. By using company accounting and monthly return data, I set up portfolios as proxies of value, growth, small-cap, large-cap, momentum, contrarian, and technology investment styles that are widely recognized by professionals and academics. The properties of style portfolios are investigated in order to understand which firms actually enter into the portfolios and their sector breakdown and the extent to which styles overlap. As a general rule, I find differences between styles in terms of accounting ratios and sector emphasis. However, some style portfolios demonstrate similarity with others. Through comparing the overlapping rate between style portfolios with that of a set of simulated random portfolios, I find significant overlaps between selected portfolios for various styles, suggesting that the evaluated styles are not independent. The risk factors used to group style portfolios contribute to establishing benchmark models for evaluating portfolio performance. Before using risk-factor models to evaluate portfolio performance, I investigate the validity of the models through examining abnormal returns of simulated randomly selected portfolios that are expected to have zero average values, since there is no investment skill involved in random portfolios simulation. The results of statistical tests raise doubts as to the validity of various benchmark models, given the fact that most of the benchmark models generate non-zero average abnormal returns for randomly selected portfolios. Finally, a return-based style analysis is used to classify investment styles of the selected UK unit trusts. Generally the unit trusts underperfonn the benchmarks that take account of various risk factors, and very few of the funds showing significantly positive abnormal returns. Regarding investment styles, the small-cap style shows consistently better performance than the other styles. The size and value factor appear to dominate in classifying styles. However, after adding more risk factors into models, the 6-month momentum and 24-month contrarian factors exhibit ability to deviate the unit trusts' exposure to size and book-to-market factors.

This thesis investigates the inconsistency in modelling financial markets between the sophisticated analytic models based in efficient market theory and persistence in the practical finance world of relatively simplistic approaches to investment, supported by the views of wealthy investors which seem to provide evidence of superior ex-ante performance inconsistent with even simple efficiency concepts. What is often termed 'value' investing is typical of such an approach. Although the term originally implied determining an intrinsic value for the common stock based upon a comprehensive fundamental analysis of a firm, it has more recently become associated with selecting stocks based upon relative values of common publicly available price-related accounting ratios. So-called 'value' stocks are frequently contrasted with 'growth' stocks, being those where higher growth rates are regarded as compensation for low current yields. Despite a general presumption of market efficiency, naive value strategies are supported by numerous academic studies reporting superior returns even after adjusting for risk using the mean-variance capital asset pricing model, arbitrage pricing theory, or in relation to volatility. However, the findings of many of these studies have been challenged on various grounds. The composition of and returns to portfolios of stocks selected from a representative sample of United Kingdom listed companies between 1979 and 1999 by four common price denominated accounting ratios, are examined. The study finds significant differences in returns between portfolios selected on four valuation ratios using a methodology that is robust to many of the criticisms levelled against previous studies, thus supporting the claims that value stocks tend to show superior returns over long holding periods without attracting any obvious element of additional risk. The study also finds that observed differences in returns predicted by one ratio as against another arise because each ratio tends to select different types of firm. It is argued that the observed persistence of certain firms and sectors in high value, high return portfolios, especially those based on cashflow, is evidence that investors do not regard stocks as homogenous, and that theories of investor over or under-reaction do not fully explain the differences in portfolio returns. This study challenges the view that higher returns to value strategies are compensation for higher systematic risks. The findings suggest that multifactor models, in which size and book to market ratio are proxies for unobserved risk, may be miss-specified and that weak or negative cashflow is a more direct indicator of vulnerability to exogenous stress. The evidence reviews securities that are inconsistently classified and the incidence offailure and takeover and concludes that it is simplistic to regard value and growth stocks as representing the extremes of an accounting ratio scale.

Much of portfolio theory has ignored the investment decisions of fund managers despite their important role as financial intermediaries or agents. This study employs agency theory to study the behaviour of UK portfolio managers towards risks. Conventional portfolio theory suggests that there should be a clientele effect. That is, funds with real and long term liabilities should specialise in riskier assets whilst those with nominal and short term liabilities specialise in less risky assets. We find little evidence of such a clientele effect based on a study of the portfolio compositions of four investing institutions: pension funds, life insurance companies, unit trusts and investment trusts. The absence of a clientele effect may be due to the limitation of conventional portfolio theory, which is mostly single-period. We examine the theory's implications for fund management in a more general multiperiod setting. To some extent, our findings are able to reconcile the discrepancy between theory and practice. Agency theory argues that fund managers may become asymmetrically averse to downside risks when their performance are competitively evaluated over short intervals. As a result, herding behaviour may occur. That is, managers may mimic the portfolio decisions of other managers to avoid grossly underperforming the average fund. We find evidence in agreement with agency theory, based on a cross-sectional study of pension funds and life insurance companies. Agents may also mimic the trading decisions of other managers over time. Studies of asset volatilities have indirect bearing on this temporal portfolio behaviour. If contemporaneous trading takes place with little reference to information signals, asset prices become excessively volatile in relation to these signals. Using the volatility test of Kleidon(1986) and co integration tests for speculative bubbles (Engle and Granger 1987), we investigate whether the UK stock market has been excessively volatile relative to the present value of expected dividend flows. The evidence are against the hypothesis of excessive volatility, and suggest that herding behaviour did not have a pervasive influence on UK share prices.

Recovery rates are mostly treated as exogenous and constant in structural models. However, this assumption creates a number of valuation problems: default probability is disassociated from the recovery rate; recovery rate is uniform for all classes of bond; there is often a problem of discontinuity in payoff at expiration; and there is a possibility of a negative duration. This thesis adheres to the original Merton (1974) framework and proposes endogenously determined recovery rates, thus avoiding the above problems. This is achieved by modelling the process of firm value as the underlying asset of a ParAsian option, whose features can capture possible bankruptcy resolutions and violations of the absolute priority rule. Armed with this technique, a range of models are developed to value straight bonds, subordinated bonds, callable bonds as well as convertible bonds. Numerical results are obtained by means of the Crank-Nicolson finite-difference method, with parameter configurations adopted from Huang and Huang (2003). Investigation into credit spreads, bond values, durations and optimal exercise strategies, sheds light on the proposition of endogenous recovery rates as well as the ParAsian feature of corporate bonds. This research should have implications for fixed income investment and its hedging strategies.

This thesis answers two major research questions: what are the investment preferences of institutional block-holders in the UK listed firms and whether their presence is associated with firm operating performance and earnings quality. Cross-sectional and panel analyses are conducted for FTSE All-Share Index constituent firms in three discrete years (1996, 1999 and 2003) and eight continuous years (1996-2003), respectively. Institutional block-holdings are positively associated with board composition, but negatively associated with directors' ownership. These findings imply that directors' ownership (board composition) is perceived by UK institutional investors as a substitute (complementary) mechanism to institutional monitoring. Institutional block-holders in the UK generally prefer smaller firms; this is interestingly different from findings in contemporary studies in the US. Dynamic models are developed for panel analyses to institutional block-holdings. The results show that institutional block-holdings are persistent up to three years. Endogenous relation does exist between institutional block-holdings and directors' ownership. Interestingly, board composition is a positive determinant to the institutional block-holdings in UK listed firms but not the other way around. Institutional. block-holders are categorized into different groups according to their domiciles, trading strategies, portfolio turnovers and business natures. Different types of institutional block-holders have heterogeneous investment preference in terms of internal corporate governance mechanisms, the firm performance and other firm specific characteristics.

Investors have strong incentives to assess the expected return of common equity as an important variable in portfolio management, capital budgeting, investment appraisal and resource allocation decision. A relatively novel methodology of estimating the expected return links market prices with analyst-issued expected cash flows to deduce the implied cost of equity capital. The purpose of this thesis is to evaluate five earnings-based implied cost of capital estimates and to assess the sensitivity of earnings-based implied cost of capital against a number of parameters introduced by alternative equity valuation models and analyst behaviour. The dissertation utilizes a relatively large dataset ofU.K. firms for the period 1994-2003. The individual and joint assessment of implied cost of capital estimates bears upon the investigation of several empirical questions. In particular, an examination of the average magnitude of implied cost of capital estimates measures the degree of mean-point accuracy of these estimates, in relation to expected return. Comparative statistics and a set of (non)parametric tests of independence are used to assess the proximity/departure of the distributions of alternative implied cost of capital estimates. Cases of extreme divergence among the estimates are marked by relative values of ex-ante characteristics and a break-point level of implied cost of capital. The ability of implied cost of capital to characterize cross-sectional variation in expected equity return is evaluated through tests of linear association between implied cost of capital estimates and a number of firm/industry specific risk factors, including-the special cases of Research and Development and Advertising expenditure. Incrementally, the usefulness of implied cost of capital estimates as parsimonious predictors of realized return is assessed, with return predictability tests carried out against various forecasting horizons, at firm and industry levels, across particular clusters of stocks, and under the assumption that a firm's risk premium is a positive function of the market excess return. Finally, the impact of analyst behaviour on the quality of earnings forecasts and, pervasively, on the quality of earnings based implied cost of capital estimates is examined. Several attributes of analyst behaviour are initially identified as sources of analyst bias. These attributes are then quantified with the use of easily measurable proxy variables facilitating the development of an econometric model that predicts analyst forecasting errors, and the consolidation of fitted values of expected errors in the estimation process of' adjusted' implied cost of capital estimates. The variation of implied cost of capital estimates across analyst investment recommendation is hypothesized and tested, and the cases of analyst forecast sluggishness and/or thin trading leading to information mismatch between analyst expectations and market prices are considered. The results of this thesis are of interest to researchers dedicated to the implied cost of capital, the cross-section of expected returns, earnings based equity valuation, model equivalence, analyst earnings forecasts, analyst behaviour, and, to a minor extent, the U.K. aggregated equity risk premium. The tests performed are sufficient in clearly designating the preferred implied cost of capital estimate. A number of implications for academics and practitioners are discussed.

This thesis comprises four essays on optimal investment in mathematical finance. The first two are concerned with optimal stock buying/selling w.r.t. its global maximum (minimum). We first aim to determine an optimal selling time so as to minimize the expectation of the square error between the selling price and the global maximum. Then, we formulate four stock buying/selling problems by minimizing/maximizing the expectation of the ratio of the buying/selling price to the global maximum (minimum) price. These are optimal stopping problems that can be formulated as variational inequality problems. We solve them by a partial differential equation approach. The latter two essays are related to optimal investment with behavioral preferences. In Essay Ill, we consider optimal asset liquidation for an investor with an S-shaped utility. We characterize the value function in the sense of viscosity solution due to the nonsmoothness of the payoff function and show that the optimal liquidation strategies are consistent with the disposition effect. In Essay IV, we consider a mean- semi-variance portfolio selection problem with probability distortion. Using a dual argument, we change the decision variable to a quantile function. We then apply the Lagrange method to solve the problem. It turns out that the distorted mean-semi- variance problem only admits an optimal solution with some proper distortion functions.

This dissertation examines the role of new asset classes and optimal dynamic strategies for tailored portfolios with a focus on passive investment portfolios. With a passive investment strategy a fund manager does not seek to outperform the rest of the market; rather, the fund manager tries to construct, at a low cost, a well-diversified portfolio that matches the risk appetite of the investor. Traditionally, passive strategies have relied on two major financial asset classes: bonds and equities. This study shows there are merits in expanding the range of acceptable assets. One should consider securities defined on real estate, commodities, hedge funds, private equity, commodities or even volatility. One should also consider the interrelationships of these securities over the long run and, further, one should consider options or dynamic investment strategies that produce return profiles that are not linear in the constituent asset returns. So, passive investment strategies do not necessarily need to be limited to index tracking strategies based on equities and bonds. As a result, this study is focused on the design of portfolios tailored to passive investment strategies. This study addresses these broader issues in four main chapters. Firstly, we address the choice of asset classes to include in optimal passive portfolios in a single period context: Should optimal passive portfolios include allocations to commodities, volatility, hedge funds and private equity (PE)? What is the most attractive path to an investment in hedge funds: single manager hedge funds (HF), fund of hedge funds (FoHF) or investible hedge fund (IHF) indices? Are the relatively new lHF indices merely disguised FoHF? These questions are answered using a methodology similar to the Black-Litterman (BL) model. We find that, if one assumes that recent performance of alternative investments persists, it is optimal to tilt the global market portfolio towards HF, commodities and PE and away from traditional assets. There is no place for lHF indices or FoHF. The similarity in performance of these two investment vehicles is so similar to suggest that IHF indices are merely disguised FoHF.