M.Comm. (Financial Management) / Since the late 1990s South African companies have started to dual list their shares in different countries, mainly to source capital from larger and more developed economies. In addition to this the level of participation by foreigners in the buying and selling of South African shares has increased. This leads to the question: should a local or a global CAPM (capital asset pricing model) be used to value shares that are traded in integrated global capital markets? This study focuses on dual-listed South African shares as these shares are most likely to be traded by investors globally. This study replicated aspects of earlier studies conducted in the Unites States of America and the United Kingdom, which are developed economies. By applying the same principles within a South African context, valuable insights might be derived relating to companies from developing economies. The main purpose of this study is to investigate the impact of using a global CAPM instead of a local CAPM to determine the cost of equity of South African companies. To this end, a sample of 26 dual-listed South African companies was selected using non-probability judgement sampling. Descriptive research was undertaken using quantitative analysis of secondary data. The cost of equity using the local and global CAPM was calculated for each of the selected dual-listed South African companies. The historical monthly returns of the dual-listed shares as well as each of the local and global risk factors during the period from 1 January 2005 to 31 December 2009 were used to calculate the local and global beta coefficients. The estimates of the local and global cost of equity were compared to ascertain whether there were significant differences for individual shares, as well as across different market sectors. While the results from similar previous studies on shares of developed countries by Koedijk and van Dijk (2004:474); Koedijk et al (2002:911); and Mishra and O‟Brien (2001:28) indicated insignificant differences between the local and global CAPM, this study indicated differences of 400 basis points and above for the sample of dual-listed South African companies. The findings in this study therefore suggest that the findings from studies conducted in developed economies cannot be generalised for companies in developing economies. In the South African market, shares across different sectors behave differently towards global risk factors; therefore this study highlighted the need for financial analysts to carefully consider using the global CAPM instead of the local CAPM when valuing shares that are traded in globally integrated capital markets. Using the incorrect cost of equity may result in incorrectly valuing a company as well as incorrect decision making.

This work presents a class of models in asset pricing, whose underlying has dynamics of Affine jump diffusion type. We first present L´evy processes with their properties. We then introduce Affine jump diffusion processes which are basically a particular class of L´evy processes. Our motivation for these is driven by the fact that many financial models are built on them. Affine jump diffusion processes present good analytical properties that allow one to get close form formulas for a wide range of option pricing. The approach we use here is based on the paper by Duffie D, Pan J, and Singleton K. An example will show how incorporating parameters such as the volatility of the underlying asset in the model, can influence the resulting price of the financial instrument under consideration. We will also show how this class of models incorporate well known models, specially those used to model interest rates dynamics, like for instance the Vasicek model. / Dissertation (MSc (Mathematics of Finance))--University of Pretoria, 2008. / Mathematics and Applied Mathematics / unrestricted

Since Engle and Granger formulated the concept of cointegration in 1987, the literature has extensively examined the unbiasedness of the commodity futures prices using the cointegration-based technique. Despite intense attention, many of the previous studies suffer from the contradicting empirical results. That is, the cointegration test and the stationarity test on the differential contradict each other. In marked contrast, my dissertation develops the no-arbitrage cost-of-carry model in the NYMEX light sweet crude oil futures market and tests stationarity of the spot-futures differential. It is demonstrated that the primary cause of the "cointegration paradox" is the model misspecifications resulting in omitted variable bias.

Asset prices and macroeconomics: towards a unified macro-finance framework Aleš Maršál March 30, 2020 Abstract The dissertation consists of three papers focused on fiscal policy and explaining what determines the dynamics of cross-sectional distribution of bond prices. The connecting factor of the thesis is however not just its main theme but also the used methodology. The valuation of bonds and effects of studied policies are endogenous outcome of the full-fledged macro-finance dynamic stochastic general equilibrium model. The first chapter provides broader context and non-technical summary of the three papers in following chapters. The first paper studies the role of trend inflation in bond pricing. Motivated by recent empirical findings that emphasize low-frequency movements in inflation as a key determinant of term structure, we introduce trend inflation into the workhorse macro-finance model. We show that this compromises the earlier model success and delivers implausible busi- ness cycle and bond price dynamics. We document that this result applies more generally to non-linearly solved models with Calvo pricing and trend inflation and is driven by the behavior of price dispersion, which is i) counterfactually high and ii) highly inaccurately approximated. We highlight the channels be- hind the undesired performance...

With data covering 20 years, we test three different liquidity measures' explanatory power in explaining asset returns on the Swedish stock market, and if an illiquidity premium exists. After establishing whether an illiquidity premium exists or not, we test whether the asset pricing models CAPM and the Fama-French three-factor model can benefit from including a liquidity factor. We use t-tests and regressions to test the liquidity measures and whether our chosen asset pricing models benefit from including a liquidity factor. Our findings do not support the definition of an illiquidity premium for our liquidity measures Return to dollar volume, Turnover rate, and Zero trading days. We apply a common definition of the illiquidity premium, which is that the least liquid portfolio should outperform the most liquid portfolio. We do, however, find that liquidity can to some degree explain asset returns, and when constructing a liquidity factor and including it into our asset pricing models, their explanatory power increases. Therefore, liquidy still might be a variable one should consider when explaining asset returns. Both CAPM and Fama-French three-factor model becomes better when including liquidity, and the result is consistent in our robustness test where we divide our sample into subperiods.

Using an APT model where global temperature shocks are a systematically priced factor, the risk premium is significant and positive. Evidence is provided that positive exposure to temperature shocks is related to increasing CO2 emissions by industry. The global impact on the cost of equity could be as high as 2.8% per year, implying a global GDP loss of $2.2 Trillion per year due to global temperature shocks.

Magister Commercii - MCom / Market Segmentation and style investing have become an essential part of security management over the past 40 years. There are many factors that separate the market, these include economy, investor behaviours, and specific anomalies. Apart, from the segmentation, investors lean towards a few tested investment styles and sectors, which hinder growth, while, dividing the market further. Thus, a major question arises on what really drives asset performance in the South African equity market. An evaluation of the relationship between sector performance and style anomalies over time is essential.

Asset allocation and portfolio decisions are at the heart of money management and draw great attention from both academics and practitioners. In addition, the segmentation of fund investors (i.e., the clientele effect) in the money management industry is well known but poorly understood. The objective of this dissertation is to study the implications of regime switching behaviors in asset returns on asset allocation and to analyze the clientele effect as well as the impact of portfolio management contracts on fund investment. Chapter 2 presents an innovative regime switching multi-factor model accounting for the different regime switching behaviors in the systematic and idiosyncratic components of asset returns. A Gibbs sampling approach for estimation is proposed to deal with the computational challenges that arise from a large number of assets and multiple Markov chains. In the empirical analysis, the model is applied to study sector exchange-traded funds (ETFs). The idiosyncratic volatilities of different sector ETFs exhibit a strong degree of covariation and state-dependent patterns, which are different from the dynamics of their systematic component. In a dynamic asset allocation problem, the certainty equivalent return is computed and compared across various models for an investor with constant relative risk aversion. The out-of-sample asset allocation experiments show that the new regime switching model statistically significantly outperformed the linear multi-factor model and conventional regime switching models driven by a common Markov chain. The results suggest that it is not only important to account for regimes in portfolio decisions, but correct specification about the structure and number of regimes is of equal importance. Chapter 3 proposes a rational explanation for the existence of clientele effects under commonly used portfolio management contracts. It shows that although a fund manager always benefits from his market timing skill, which comes from his private information about future market returns, the value of the manager's private information to an investor can be negative when the investor is sufficiently more risk-averse than the manager. This suggests different clienteles for skilled and unskilled funds. Investors in skilled funds are uniformly more risk-tolerant than investors in unskilled funds. Moreover, a comparative statics analysis is conducted to investigate the effects of the manager's skill level, contract parameters, and market conditions on an investor's fund choice. The results suggest that the investors who are sufficiently more risk-averse than the manager should include fulcrum fees in the contract to benefit from the skilled manager's information advantage.

No description available.

This dissertation consists of two chapters. The first chapter examines the role of growth options on stock return continuation. Growth options are both difficult to value and risky. Daniel, Hirshleifer and Subrahmanyam (1998) argue that higher momentum profits earned by high market-to-book firms stem from investors' higher overconfidence due to the difficulty of valuing growth options. Johnson (2002) and Sagi and Seasholes (2007) offer an alternative rational explanation wherein growth options cause a wider spread in risk and expected returns between winners and losers. This paper suggests that firm-specific uncertainty helps disentangle these two different explanations. Specifically, the rational explanation is at work among firms with low firm specific uncertainty. However, the evidence is in favor of the behavioral explanation for firms with high firm specific uncertainty. This is consistent with the notion that investors are more prone to behavioral biases in the presence of firm-specific uncertainty and the resulting mispricings are less likely to be arbitraged away. The second chapter examines how investors capitalize differences of opinion when disagreements are common knowledge. We conduct an event study of the market's reaction to analysts' dispersed earnings forecast revisions. We find that investors take differences of opinion into account and do not exhibit an optimism bias. Our findings indicate that the overpricing of stocks with high forecast dispersion is not due to investors' tendency to overweight optimistic expectations, but rather due to investor credulity regarding analysts' incentives. Our findings support the notion that assets may become mispriced when rational investors face structural uncertainties as proposed by Brav and Heaton (2002). / Ph. D.