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 A machine learning approach to quantifying and relating the determinants of unemployment in South Africa(2021) Mulaudzi, RudzaniUnemployment is a significant problem that South Africa faces. The rate was 30.1% in the first quarter of 2020: placing it amongst the top ten worst unemployment rates in the world. Public policy is a typical instrument used by governments to address unemployment sustainably. It is normally informed by forecasts derived through economic (traditional sta tistical) models. These models are, however, suitable when the data is stationary and linear. The South African unemployment rate, on the other hand, is asymmetric, seasonal, upward trending, and nonstationary. Vector autoregression (VAR), a traditional statistical model, was used to forecast the South African unemployment rate. It resulted in a mean absolute scaled error (MASE) of 41. Comparatively, twelve machine learning models were used to forecast the unemployment rate. The lowest MASE achieved was 0.39. Making the machine learning models 105 times more accurate than the VAR: the benchmark model. Additionally, through feature selection techniques, machine learning approaches enabled the identification of the most impactful features in forecasting the unemployment rate. These features were used to construct a Dynamic Bayesian Network (DBN) to determine how they influence each other and the unemployment rate. The DBN was then used to perform do-Calculus, a data-driven scenario analysis technique. One scenario tested the impact of increasing the GDP on the unemployment rate. This positively impacts the unemployment rate. However, a decline in GDP has a greater negative impact. Therefore, policymakers should avoid, at all costs, a decline in the GDP. This research, therefore, demonstrates the value of machine learning in forecasting the South African unemployment rate (a nonstationary macroeconomic variable) across the broad ma chine learning value chain: feature selection, forecasting, feature influence analysis, and do-Calculus scenario analysis. Previous research tends to only focus on one or two aspects of the value chainItem Understanding the relationship between business failure and macroeconomic business cycles: a focus on South African businesses(2017) De Jager, MarinusThis study examined the relationship between business failure and macroeconomic fluctuations within business cycles of South Africa’s economy for the time period 1980 to 2016. The study also sought to understand where, if any, immediate and lag correlations between fluctuations and business failure could be established. To understand this connection, this study used longitudinal data sets of different macroeconomic factors and studied their influence on business failure. The vector error correction model (VECM) was used to determine the long-term relationship between failure and each of the other variables. Additionally, Granger Causality was applied to establish whether the macroeconomic variables investigated in this study can be constructed to predict the probability of business failures. Three classes of macroeconomic predictor variables were considered. Firstly, well-known international variables in the form of GDP and CPI were used. Secondly, the study incorporated the three Composite Business Cycle indicators- leading, coincident and lagging. Lastly, behavioural indicators were used to incorporate the views of the actual businesses and their customers, which for this the study were the Business and Consumer Confidence Indices. After examining the effects the 7 macroeconomic variables had on business failure, the study found that there is a long-run relationship between the Composite Lagging Business Cycle indicator, the Business Confidence and Consumer confidence, which influenced Business Failure. Additionally, it was noted that Business Failure influence the Composite Lagging Business Cycle indicator in the long-run. The study additionally found that Business Failure may Granger Cause the Composite Leading Business Cycle indicator Outcomes of the study are potentially vital for entrepreneurs to understand the timing of entry into markets based on macroeconomic fluctuations through their cycles in certain industries. Business owners can make proactive financial and strategic decisions vital for survival of their business through the expansion and especially in the contraction cycles of the macroeconomic environments.Item Interaction between macroeconomic fundamentals and energy prices: evidence from South Africa(2017) Diale, Tumelo KGrowth in commodity exporting economies, such as South Africa, is highly dependent on the revenue generated from exports. It is thus evident that as commodity prices fluctuate, income and the balance of payments will be accordingly impacted. This is further exacerbated by strong dependence on the imports of certain commodities. Oil is one such commodity on whose imports South Africa is highly dependent. Although natural gas is also imported, it is in lower quantities and is as such expected to impact South Africa to a lower extent. Coal, on the other hand, is among the main commodity exports and was expected to have an impact on (and be impacted by) South African macroeconomic fundamentals. In this study, we use a VECM and MGARCH model to test the interaction between South African macroeconomic variables and these three commodities. Our VECM findings indicate that oil and exchange rates are inflationary. This implies that an increase in oil prices and/or exchange rates (indicating a depreciation of the Rand against the U.S. Dollar) results in an increase in inflation. Inflation, on the other hand, propagates higher coal prices and to a lesser extent, higher interest rates. We account the latter to South Africa’s inflation targeting regime and the former to demand and supply dynamics which occur at RBCT as production costs increase (short-term coal export contracts and spot market sales). Natural gas is found to have weak impacts on interest rates and exchange rates. Our MGARCH model shows that only the innovations in natural gas and oil prices spillover into interest rates and exchange rate. There is no direct spillover captured. However, there is strong direct spillover from oil to inflation. Lastly, interest rates are found to have a strong direct volatility spillover to both oil and natural gas. We attribute this to the exchange rate impact that interest rates have and is supported by the exchange rate impact on commodity price volatility. We conclude that an in-depth understanding of triggers is pertinent for monetary and fiscal policy decisions in South Africa. Although the South African economy is relatively diversified compared to other developing countries, commodity price fluctuations do have a significant impact on economic performance.Item Monetary policy and the stock market structure: some international empirical evidence(2016) Alovokpinhou, Sedjro AaronThis paper builds upon Blanchard's (1981) model of asset prices, and provides an empirical evidence for good news cases (GNC) and/or bad news cases (BNC) as de ned in Blanchard's paper. We update Blanchard's model by introducing Taylor's rule of monetary policy and explicitly incorporate income distribution in a small, open economy. The ndings indicate that, the labour share is a strong and signi cant variable that should be considered in asset pricing models. The real exchange rate plays a signi cant role in the determination of asset prices in most of the selected countries, but the signi cance is stronger in the emerging markets economies. As the main objective of the paper, the study has found four of the selected countries to be bad news cases and eight of them are good news cases.