Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
LSTM prediction capability on the South African JSE Top 40 of historical and live data
This study evaluates the efficacy of Long Short-Term Memory (LSTM) models in stock price forecasting using data from the South African FTSE/JSE Top 40 index, a domain yet to be extensively explored, particularly in real-time data analysis. Addressing the gap in existing research, this study assesses...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English English |
| Published: |
School of Economics
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613177264472064 |
|---|---|
| access_status_str | Open Access |
| author | Elhag, Mohsen |
| author2 | Ndlovu, Godfrey |
| author_browse | Elhag, Mohsen Ndlovu, Godfrey |
| author_facet | Ndlovu, Godfrey Elhag, Mohsen |
| author_sort | Elhag, Mohsen |
| collection | Thesis |
| description | This study evaluates the efficacy of Long Short-Term Memory (LSTM) models in stock price forecasting using data from the South African FTSE/JSE Top 40 index, a domain yet to be extensively explored, particularly in real-time data analysis. Addressing the gap in existing research, this study assesses LSTM model predictive capability in the South African stock market on historical data and its adaptability to the dynamic, real-time stock market environment over the period from January 2001 to January 2024. Various LSTM models were trained with different configurations, and the results show that a single-layer LSTM model performs better than its multilayer counterpart in processing historical data, in terms of the mean absolute error (MAE), the root mean square error (RMSE), Mean Absolute Percentage Error (MAPE) and the R-squared. However, when applied to real-time data, the accuracy of the single-layer model diminishes, underscoring the challenges posed by the dynamic and unpredictable nature of live stock market conditions. The findings contribute to the field of financial forecasting by demonstrating the strengths and limitations of the LSTM model in the context of the South African stock market. While showcasing significant potential in historical data analysis, performing on par with previous studies, the study underscores the need for further development of the model for real-time forecasting. Future research directions include extending the testing period, integrating diverse data sets, and exploring a combination of LSTM with other forecasting methodologies. |
| format | Thesis |
| id | oai:open.uct.ac.za:11427/41537 |
| institution | University of Cape Town (South Africa) |
| language | English eng |
| last_indexed | 2026-06-10T12:31:58.458Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UCTD — University of Cape Town Open Access Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | School of Economics |
| publisherStr | School of Economics |
| record_format | dspace |
| source_str | UCTD — University of Cape Town Open Access Repository |
| spelling | oai:open.uct.ac.za:11427/41537 LSTM prediction capability on the South African JSE Top 40 of historical and live data Elhag, Mohsen Ndlovu, Godfrey South African FTSE/JSE This study evaluates the efficacy of Long Short-Term Memory (LSTM) models in stock price forecasting using data from the South African FTSE/JSE Top 40 index, a domain yet to be extensively explored, particularly in real-time data analysis. Addressing the gap in existing research, this study assesses LSTM model predictive capability in the South African stock market on historical data and its adaptability to the dynamic, real-time stock market environment over the period from January 2001 to January 2024. Various LSTM models were trained with different configurations, and the results show that a single-layer LSTM model performs better than its multilayer counterpart in processing historical data, in terms of the mean absolute error (MAE), the root mean square error (RMSE), Mean Absolute Percentage Error (MAPE) and the R-squared. However, when applied to real-time data, the accuracy of the single-layer model diminishes, underscoring the challenges posed by the dynamic and unpredictable nature of live stock market conditions. The findings contribute to the field of financial forecasting by demonstrating the strengths and limitations of the LSTM model in the context of the South African stock market. While showcasing significant potential in historical data analysis, performing on par with previous studies, the study underscores the need for further development of the model for real-time forecasting. Future research directions include extending the testing period, integrating diverse data sets, and exploring a combination of LSTM with other forecasting methodologies. 2025-07-04T16:59:53Z 2025-07-04T16:59:53Z 2025 2025-07-04T15:38:03Z Thesis / Dissertation Masters MPhil http://hdl.handle.net/11427/41537 en eng application/pdf School of Economics Faculty of Commerce University of Cape Town |
| spellingShingle | South African FTSE/JSE Elhag, Mohsen LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| thesis_degree_str | Master's |
| title | LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| title_full | LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| title_fullStr | LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| title_full_unstemmed | LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| title_short | LSTM prediction capability on the South African JSE Top 40 of historical and live data |
| title_sort | lstm prediction capability on the south african jse top 40 of historical and live data |
| topic | South African FTSE/JSE |
| url | http://hdl.handle.net/11427/41537 |
| work_keys_str_mv | AT elhagmohsen lstmpredictioncapabilityonthesouthafricanjsetop40ofhistoricalandlivedata |