- Title
- Market timing and portfolio returns: an empirical analysis of the potential profitability of buy-sell strategies, based on South African equities 2009-2018
- Creator
- Mulweli, Ramulongo
- ThesisAdvisor
- Nel, Hugo
- ThesisAdvisor
- Khumalo, S A
- Subject
- Johannesburg Stock Exchange
- Subject
- Stocks -- Charts, diagrams, etc.
- Subject
- Investment analysis -- South Africa
- Subject
- Stocks -- South Africa
- Subject
- Stocks -- South Africa -- Cast studies
- Date
- 2020
- Type
- text
- Type
- Thesis
- Type
- Masters
- Type
- MCom
- Identifier
- http://hdl.handle.net/10962/144487
- Identifier
- vital:38350
- Description
- South Africa’s financial markets have become larger and more complex over recent decades. The number of market participants who are using technical analysis techniques to predict the market’s movement has been growing rapidly. This research aims to investigate if historical share prices can be used when forecasting the market’s direction and to examine the profitability of the Japanese candlestick patterns. The study is based on ten companies selected from the JSE top 40 2019 composition. These are Aspen Pharmacy Holding, Capitec Bank Holding LTD, Discovery LTD, Kumba Iron Ore LTD, Mondi PLC, Mr. Price Group LTD, MTN Group LTD, Naspers LTD, SASOL LTD, and Shoprite Holdings LTD. These were selected from the JSE top 40 based on market capitalization and sector. This research analyzes eight candlestick reversal patterns; four are bullish patterns namely: doji star, hammer, bullish engulfing and the piercing lines and the other four are bearish patterns namely: shooting star, hanging man, bearish engulfing and the dark cloud cover. The ARCH and GARCH models are used to test for correlation between past share prices and future share prices and the binomial test and the mean return calculations were used to test the profitability of candlestick patterns. The sample is from Thomson DataStream 2019 and IRESS SA 2019 and covers ten years with 2496 observations starting from 02 January 2009 to 31 December 2018. The findings from the ARCH and GARCH tests revealed that there is a serial correlation between the returns from the previous day and the returns for the current day. The results from the mean returns and the binomial tests show strong evidence that the shooting star, hanging man, bearish engulfing and the bulling engulfing are statistically significant in predicting the share price movements. On the other hand, there was no evidence that the dark cloud cover, piercing lines, and the bullish doji can predict share price movements. Additionally, further studies on this topic could be improved by adding different candlestick patterns and the total number of companies analyzed. The results could also be improved by analyzing the candlestick reversal patterns when they are used with other trading rules such as support resistance levels and oscillators.
- Format
- 114 pages, pdf
- Publisher
- Rhodes University, Faculty of Commerce, Economics and Economic History
- Language
- English
- Rights
- Mulweli, Ramulongo
- Hits: 3946
- Visitors: 4146
- Downloads: 358
Collections
RU Department of Economics and Economic History
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details | SOURCE1 | RAMULONGO-MCOM-TR20-231.pdf | 1 MB | Adobe Acrobat PDF | View Details |