Browsing by Subject "Bayesian Network"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Open Access
    A semantic Bayesian network for automated share evaluation on the JSE
    (2021) Drake, Rachel; Moodley, Deshendran
    Advances in information technology have presented the potential to automate investment decision making processes. This will alleviate the need for manual analysis and reduce the subjective nature of investment decision making. However, there are different investment approaches and perspectives for investing which makes acquiring and representing expert knowledge for share evaluation challenging. Current decision models often do not reflect the real investment decision making process used by the broader investment community or may not be well-grounded in established investment theory. This research investigates the efficacy of using ontologies and Bayesian networks for automating share evaluation on the JSE. The knowledge acquired from an analysis of the investment domain and the decision-making process for a value investing approach was represented in an ontology. A Bayesian network was constructed based on the concepts outlined in the ontology for automatic share evaluation. The Bayesian network allows decision makers to predict future share performance and provides an investment recommendation for a specific share. The decision model was designed, refined and evaluated through an analysis of the literature on value investing theory and consultation with expert investment professionals. The performance of the decision model was validated through back testing and measured using return and risk-adjusted return measures. The model was found to provide superior returns and risk-adjusted returns for the evaluation period from 2012 to 2018 when compared to selected benchmark indices of the JSE. The result is a concrete share evaluation model grounded in investing theory and validated by investment experts that may be employed, with small modifications, in the field of value investing to identify shares with a higher probability of positive risk-adjusted returns.
  • No Thumbnail Available
    Item
    Open Access
    Banking regulation: a bayesian network approach to risk management
    (2025) Gross, Eden; Kruger, Ryan; Toerien, Francois
    The ever-evolving regulation surrounding banks and market risk, coupled with increased computing power, make for favourable conditions in employing machine learning techniques to estimate and forecast market risk metrics such as value at risk (VaR) and expected shortfall (ES). This study consists of three sections. First, this study comprehensively examines the performance of various market risk models when producing VaR and ES, and their stressed counterparts, using Standard and Poor's (S&P) 5 00 index returns from 1991 to 2020. The initial results show that autoregressive models are the most accurate of the traditional market risk models. Second, the first section's results are then used as the basis against which a novel and comprehensive Bayesian network (BN) methodology for producing VaR and ES forecasts, and those of their stressed counterparts, is assessed in the context of banking regulations, using four learning algorithms. The forecasts generated by the BNs are not found to offer any improved accuracy when incorporated into the market risk metric calculations, primarily due to the limited weight of the forecast in the return distribution relative to the historical returns in the return probability density function. Finally, a novel integrated forecast dynamic Bayesian network (IFDBN) methodology is developed, whereby, for each metric, the best -in-class autoregressive model and the best-in-class BN learning algorithm are coupled to produce market risk forecasts. The results of the IFDBNs are mixed, with the stressed ES metric IFDBN being the only IFDBN to produce more accurate forecasts relative to its traditional autoregressive counterpart. While certain market risk metrics may benefit from using IFDBNs in the forecasting process, this result is not universal, and the risk practitioner must evaluate the usefulness of IFDBNs on a case-by-case basis.