BBSE Research Projects
Permanent URI for this community
Browse
Browsing BBSE Research Projects by Subject "Artificial neural networks (ANN)"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- ItemOn the profitability of technical trading rules based on artificial neural networks: evidence from the Kenyan stock market(Strathmore University, 2017) Ngolobe, Loreine DottiThe aim of this study is to investigate the profitability of technical trading rules based on artificial neural networks in the Kenyan stock market. The technical trading rule is also compared to the buy-and-hold strategy to determine which strategy is more profitable. The study is carried out on the NSE 20 Index using the Excel software. A three-layered feedforward network model is used following (Rodriguez, Martel, & Rivero, 2000). The data set is divided into three sub-periods corresponding to the bear, stable and bull markets. Nine stock indices are then sampled from each sub-period and used as the inputs to the model so as to obtain the ANN forecasts. To determine the forecast accuracy, both the percentage of correct predictions and the Pesaran and Timmermann (1992) non-parametric test proportion of correctly predicted signs are used. To test for economic significance, the total return of the technical trading rule is compared to the buy-and-hold return. The results suggest that, in the absence of trading costs, the technical trading rule is always superior to the buy-and-hold strategy for both bear and stable markets. The buy-and-hold strategy, however, generates higher returns than the technical trading rule for bull markets. These findings are especially beneficial to traders since they can achieve significant trading advantages through the adoption of ANN based technical trading rules. Traders can therefore alternate between the ANN based technical trading rule and the buy-and-hold strategy, depending on whether it is a bear, bull, or stable episode, so as to maximize profits and minimize losses