Electronic Theses and Dissertations (Masters/MBA)

Permanent URI for this collectionhttps://hdl.handle.net/10539/37942

Browse

Subject: search.filters.subject.Johannesburg Stock Exchange

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Mean-Variance Optimisation of A South African Index Based Portfolio Using Machine Learning
    (University of the Witwatersrand, Johannesburg, 2021) Makgoale, Katlego; Jakubose, Sibanda
    This study embarked on a comparison of the effectiveness of the Markowitz Mean- Variance Portfolio Optimisation against utilising a Machine Learning Technique to construct an optimal portfolio. The study aimed to: Construct an optimal portfolio using the Mean-Variance Analysis Framework, Construct an optimal portfolio using a Machine Learning Technique (Support Vector Regression), Contrast the results of the Minimum-Variance Portfolio and the Machine Learning Portfolio. The stocks of the FTSE JSE FIN15 index were chosen to construct the portfolio. The historical returns of the stocks in the index were used to trained (December 2014 to June 2019) and test the models(June 2019 to December 2020). The Mean-Variance Analysis and Minimum-Variance Portfolio were constructed using Python code that the author compiled. Similarly, the Support Vector Regression model was built in Python. The weights for the Machine Learning portfolio were calculated using the pseudo-inverse matrix and the predicted value of the Regression Model. It was found that the Minimum-Variance and Machine Learning portfolio produced different portfolios, but both containing fewer holdings than the original index. The performance of the Minimum-Variance Portfolio exceeded that of the index and the Machine Learning Portfolio with regards to relative(excess) returns and total returns in the out of sample period. It was found that the Machine Learning portfolio performs well at replicating the index returns but fails to exceed them and typically has a higher risk associated with it. It was concluded that the Minimum-Variance portfolio would be the most attractive to a risk-averse investor and the Machine Learning portfolio underperforms the Minimum variance and the index. Therefore confirming the effectiveness of Mean-variance Optimisation in a South African context against a Machine Learning Technique
  • Thumbnail Image
    Item
    The performance of South African Socially Responsible Investments: a comparative analysis of listed equity
    (2021) Olen, Davin
    In recent years, Sustainable and Responsible Investments (SRI) have undergone significant advancements in terms of both assets under management and investor attention. Concomitantly, the metrics which inform SRI methods, Environmental, Social and Governance (ESG) factors, have increasingly been incorporated within global investment approaches. This shift in approach suggests a permanent alteration to investing practices for some authors and investment houses. For South Africa, however, there is not yet consensus regarding the longterm comparative financial performance of securities focussing on SRI, considering the purported benefits of SRI’s incorporation within dominant investment approaches. In an attempt to address this lacuna, the following research dissertation unpacks the South African understanding of SRI and evaluates the comparative performance of portfolios constructed from rated ESG securities on the Johannesburg Stock Exchange. This research piece commences with an overview of recent global SRI developments followed by an evaluation of SRI as applied within South Africa alongside the country’s legislative framework. Provided with the relevant background, this research dissertation constructs a set of nine portfolios of equities listed on the Johannesburg Stock Exchange, based on both the security’s Bloomberg ESG Disclosure score and market capitalisation. Utilising the Fama and French Three- and Five-Factor asset pricing models, this research dissertation then gauges the financial performance of the constructed portfolios from May 2009 until April 2021 in terms of portfolio alpha values. Finally, this research dissertation reports that portfolios constructed from highly rated ESG companies with small and medium market capitalisation provide statistically significant positive alpha values at the 5% limit. For highly rated ESG companies with a large market capitalisation, statistically significant positive alpha values are identified at the 10% limit while a portfolio of medium rated ESG securities with the same market capitalisation report positive alpha values with significance at the 5% limit. A number of factor tests are further undertaken in order to determine the pricing accuracy of the two models in consideration. It is concluded iv that both the asset pricing models considered fail to explain the excess returns of the constructed portfolios at the 5% level of statistical significance.
  • Thumbnail Image
    Item
    Testing the pecking order theory of capital structure in South Africa
    (2021) Andries, Lydia Sejo
    This study was conducted with the aim of testing the pecking order hypothesis as well as examining the variables influencing the capital structure decisions using companies that are listed on the main board of the Johannesburg Stock Exchange over 12 years from 2009 to 2020. It has been noted that several listed companies in South Africa are facing the challenge of survival in the current economic climate. The companies are constantly found to be in risky situations in relation to their finances, which makes it difficult for them to meet their responsibilities to creditors and other stakeholders. Making the correct decisions regarding capital structure enhances the financial well-being of any organization. When companies fail to make the right capital decisions, the result can be financial distress, bankruptcy, and liquidation. The research will also inform both potential and existing investors in South Africa on the general approach to funding of the JSE listed companies and therefore give investors some basis for their investment decisions. A total of 197 companies excluding financial companies were selected to be used in the study using stratified random sampling. Pooled ordinary least squares, fixed effects and random effects models with heteroscedasticity robust standard errors were used. The results from the model employed to test the pecking order hypothesis do not show strong evidence of company management going in line with the pecking order theory when making decisions regarding the financing of the company. The conventional leverage regressions run to determine the factors that influence the capital structure were found to be consistent with theoretical predictions except for firm size which had the correct sign as predicted by theory but statistically insignificant. Asset tangibility and firm size were both positive while Market to Book and profitability were both negative in line with theoretical predictions. All the conventional predictors of leverage were statistically significant at the 1% and 5% significance levels except for firm size which was not statistically significant at any of the conventional levels of significance. Thus, for the sample of companies used in this study and the sample period selected, there is no strong support for the pecking order theory while leverage is determined by growth options as proxied by MTB, asset tangibility as well as profitability. A limitation of the model used in this study is that it only considered firm-specific factors in determining the predictors of leverage. Future research could include macroeconomic variables as additional predictors of the financing decisions of listed firms in South Africa as empirical evidence shows that these macroeconomic are important in the leverage-financing deficit nexus
  • Thumbnail Image
    Item
    The impact of monetary policy related variables on the performance of the South African Real Estate Investment Trust (REIT)
    (2020) Mbhokota, Vongani
    Since the introduction of the Real Estate Investment Trust (REIT) in the Johannesburg Stock Exchange (JSE), there has been huge interest from both local and international investors to invest in the SA REIT. The reason for this rise in interest is that the SA REIT offers investors ownership of tangible assets that are backed by rental income and capital growth in the underlying asset value. In recent years this has been a good diversification strategy for investors, especially in emerging markets like South Africa. Whilst the SA REIT has given investors good returns since its inception in 2013, there is still not enough literature on the factors that affect the performance of the SA REIT in relation to monetary policy movements. Capital market movements in most emerging countries are dependent on macroeconomic factors like monetary policy movements, political stability, the input and output of production, etc. Most studies on how these macroeconomic factors impact on the performance of the REIT have been done in developed markets, and most of them have focused on monetary policy variables, e.g. interest rates, inflation, exchange rates, and GDP. This has necessitated this study on “The impact of monetary policy related variables on the performance of the South African Real Investment Trust (REIT)”. The purpose of the study was to determine the impact of monetary policy related variables in the short and long run on the performance of the SA REIT to assist investors and other role players as a tool to make investment decisions. The data gathered consisted of information on the overall performance of the SA REIT in the last nine years, and the source of this information was the SA REIT Association, which is a statutory body of the SA REIT that represents international REIT. SA REIT data was taken from the SA REIT Association’s database and was compared against each monetary policy variable taken from Statistics South Africa, which is deemed to be a reliable source. Unit root testing, the Autoregressive Distributed Lag (ADRL) model, and Bound tests were utilised to test if the monetary variables impact the performance of REIT returns in the short and long run to answer the hypothesis. The findings were presented in a graphical representation of the analysed data using time series plots in the short and long run. Methods used are as follows: unit root testing, the ADRL model, and Bound test . The outcome of the research results showed that none of the variables have a significant relationship with the performance of the SA REIT in the short run, but interest rates and exchange rates have a significant relationship with the performance of the SA REIT in the long run. This may prompt other studies to investigate how interest rate and exchange rate can be used positively to maximise returns in the SA REIT for investors, since it has been established that the two variables impact on the performance of the REIT.