3. Electronic Theses and Dissertations (ETDs) - All submissions
Permanent URI for this communityhttps://wiredspace.wits.ac.za/handle/10539/45
Browse
4 results
Search Results
Item Application of machine learning algorithms to predict the closing price of the Johannesburg Stock Exchange all-share index(2020) Makgwedi, Precious MakganotoStock markets are regarded as one of the most important indicators of the economy’s strength and development. Predicting stock prices is of critical importance for investors who wish to minimise the risks of investments. Stock price prediction is a difficult task since stock prices are influenced by factors such as the financial status of the company, socioeconomic conditions of the country, political atmospheres, and natural hazards. The Efficient Market Hypothesis (EMH) states that stock markets behave like a random walk and due to this reason, it is complex to forecast the stock market. Researchers use time series forecasting, technical, and fundamental analyses to predict the stock values while proving or disproving the EMH. In the past, researchers used traditional methods such as Autoregressive Integrated Moving Average (ARIMA) to predict stock prices. Currently, deep learning architectures are widely used to solve time-dependent problems and can provide a huge push to the problem of stock price prediction. The main objective of this study is to develop a framework that forecasts the daily closing price of All- Share index data based on deep learning techniques. To achieve this objective, Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) are employed. A Vector Autoregressive (VAR) model is used to benchmark the deep learning techniques. The analysis is based on the Financial Times Stock Exchange (FTSE)/ Johannesburg Stock Exchange (JSE) All-Share (J203) data collected from Iress Expert. The results show that all the methods are able to predict the closing price of the index. GRU predicted the future closing price with an average Mean Absolute Percentage Error (MAPE) of 9.349% maximum while LSTM was able to predict with the maximum average error of 9.459%. A VAR model performed with the maximum average error of 2.152%.Item Analysing the efficiency of the Johannesburg Stock Exchange using the magic formula(2018) Vincent, Christopher JohnThis study examined the efficiency of South African markets, namely the Johannesburg Stock Exchange (JSE) through the use of a value investing strategy called the “magic formula”, which was created by Joel Greenblatt and published in his 2006 book “the little book that beats the market”. This study back tested the magic formula on the JSE from 2000 to 2016. It ranked stocks according to the magic formula methodology, using earnings yield and return on capital to derive portfolios. The portfolios were then compared against the JSE All Share Index (the market). The magic formula showed evidence of outperformance of the market over the period, even when accounting for risk. The magic formula was compared against other portfolios derived from value investing ratios, namely ROA, ROE and EY. The ROA portfolio produced the best risk-adjusted results, but all value investing portfolios outperformed the market providing evidence against efficient markets.Item Sectorial herding: evidence from the JSE(2017) Mekwa, Itumeleng EskiaThis study investigates the existence of herd behaviour within the Johannesburg Stock Exchange (JSE) and three sectorial indices using monthly closing prices for all shares listed on the JSE for the period 31 January 2003 to 31 May 2016. No evidence of herding was found on either the JSE or in any of its sectors during the sample period. Furthermore, no evidence of herding was found during bull and bear markets within the sample period.Item Individualism as a driver of overconfidence, and its effect on industry level returns and volatility across multiple countries(2016) Horne, ChadThis study attempts to determine the possible effects of individualism on industry volatility. The implications of this for behavioural finance are extensive, showing firstly that different industries react differently to behavioural biases and secondly that overconfidence is a possible driver of the positive effect of individualism on industry volatility. The country selection process was relatively objective, taking two countries with high individualism indexes and two with low indexes and including one with a medium index value. The result was a sample of the United States of America, the United Kingdom, South Africa, China and Taiwan. The industry selection process was more subjective. Industries were selected which should have a higher propensity to behavioural biases with lower book to market ratios (software and computer services industry and pharmaceutical and biotechnology industry) and other industries which should not be as strongly affected by behavioural biases (banks, mining, oil and gas producers, and mobile telecommunications industries). In order to correct for ARCH effects the series’ were modelled using a GARCH (1, 1) model. The resulting residuals, which showed no autocorrelation, were then used to conduct panel data regressions on each of the industries. The results confirmed that individualism had a positive effect on volatility in the industries which were expected (software and computer services and pharmaceuticals and biotechnology industries). However, it was also determined that the banks industry was significantly affected by individualism, an effect which it was hypothesised, was due to the individualism of employees as opposed to investors.