School of Economics and Finance (ETDs)
Permanent URI for this communityhttps://hdl.handle.net/10539/37935
Browse
144 results
Search Results
Item Essays in Forensic Economics: Exploring issues of Lost Earnings from Personal Injury and Wrongful death in South Africa(University of the Witwatersrand, Johannesburg, 2025) Chirowodza, Joe; Akinkugbe, OluyeleEconomic loss estimates are fundamental in providing compensation to victims in personal injury or wrongful death cases. Such estimates, calculated by forensic economists assist judges and lawyers to make evidence-based decisions on civil litigation matters. In particular, lost earnings estimated for vulnerable groups in society can provide financial or quantifiable compensation vital for the sustainability of the victim or surviving family members. Yet, there is an inadequate evaluation of personal injury and wrongful death cases in the South African courts, due to challenges, such as insufficient resources, expertise, and court procedural delays. In addition, this has led to the inaccurate assessment of civil litigation cases in the country. The impact is devastating, especially to vulnerable groups, who may rely on the court systems to provide fair and reliable economic loss estimates. The thesis estimates work life expectancies for four labour states and future lost earnings for workers in formal and informal employment in South Africa, with the use of the National Income Dynamics Survey (NIDS) dataset from 2008 to 2017. Work life expectancies were estimated using the Stochastic Markov-chain transition model whilst future lost earnings were estimated the using Benham-La Croix present value formulae. Work life expectancies are an essential component for estimating future lost earnings. The results showed that male and female workers who are in formal employment have higher work life expectancies and greater future lost earnings, when compared to male and female workers in the informal sector. Furthermore, the results revealed that informal employment is an important sector for women because they had longer work life durations, when compared to men. However, future lost earnings for women remain lower in the informal sector as compared to future lost earnings for men. The thesis showed the presence of racial disparities in the labour market—White population had the highest work life expectancies and future lost earnings as compared to the Black, Coloured and Asian/Indian populations in the formal sector. There is a need for policy makers to legislate specific policies that especially integrate women into the formal labour market, such as policies that create a sustainable work-life balance, such as subsidising child care activities or programmes to incentivise women to participate in the formal labour force. In addition, there is a need for the government to provide regulations and social protection to informal workers as well as promote skills development programmes for minority groups and anti-discriminatory policies in the work place. iv The thesis estimates personal consumption rates in wrongful death in South Africa for single female parents’ households using a method known as Krueger’s expenditure allocation approach. Personal consumption rates are important in estimating compensation in wrongful death for households as the proportion of income which is consumed by the deceased is removed from household income. The remaining income is then awarded as compensation to the surviving family members. For this analysis, the thesis used wave 5 data from the NIDS dataset. The results showed that as household income increases, personal consumption rates (PCRs) for a single female parent with one child and a single female parent with two children decrease. Furthermore, the thesis found that personal consumption rates for single female parents with one child were higher, as compared to personal consumption rates for single female parents with two children. The results also show that there were substantial differences in the personal consumption rates for low, middle, and high-income groups. There is a need for the government to provide more financial support programmes to assist single female parents, in particular, to sustain their families when they are alive or even when they wrongfully die. To understand the compensation which could be awarded to minor children in case of wrongful death or injury, the thesis estimated educational probabilities and expected lifetime earnings for minor children in South Africa. The study used an ordered probit model with cluster robust standard errors together with the educational-probability -weighted lifetime earnings model to estimate educational probabilities and expected lifetime earnings. In addition, the study used the child dataset section of the NIDS for wave 1 and wave 5. The results of the study showed that on average, children in South Africa are more likely to achieve lower educational levels, such as high school education, as compared to achieving higher levels of education, such as an honours degree or higher. Furthermore, the results showed that male and female White children are more likely to have the uppermost expected life time earnings, as compared to the expected life time earnings of children of other races for all family structures. These results imply the urgent need for government intervention in subsidising education even at higher educational levels. In addition, there is also a need for government to embark on social and economic programmes that will reducing the systemic and intergenerational income inequality in the country. In conclusion, forensic economics estimates provide a platform towards compensation for personal injury and wrongful death cases. This thesis provides new insights into compensation estimates, given the current limited scope of this area in developing countries and in South v Africa. Such economic loss estimates can prove to be vital for compensation especially to the vulnerable groups in society. The methodologies and estimates in the study may also assist the court systems to improve efficiency in civil litigation matters that involve personal injury and wrongful death cases.Item Energy and Commodity Price Volatility in Sub-Saharan Africa: The Role of Uncertainty(University of the Witwatersrand, Johannesburg, 2024) Makumbe, Happiness; Fasanya, Ismail OlalekeThis thesis, focusing on sub-Saharan Africa, investigates the relationship between energy prices, agricultural commodity markets, and policy/events uncertainties. The research is divided into three distinct yet interconnected papers, each employing unique methodologies and data to address specific aspects of this issue. The first paper examines the dynamic relationship between oil prices and food security in sub-Saharan Africa, accounting for the moderating role of global uncertainty, as measured by the World Uncertainty Index. Utilizing a Pooled Mean Group Autoregressive Distributed Lag (PMG-ARDL) model, the study reveals that oil price shocks and heightened global uncertainty negatively impact food security in the region. The effects are found to be homogenous across countries in the region. The second paper analyses the role of geopolitical risk in the extreme spillovers between energy prices and agricultural commodity prices. Employing a time-varying parameter vector autoregression (TVP-VAR) model and disaggregated geopolitical risk measures, the study finds evidence of significant time-variation and asymmetry in extreme spillovers. Geopolitical risk events significantly influence the intensity and direction of these spillovers, with different risk categories exhibiting varying impacts. In the third paper, the study examines the effect of climate risk on the dynamic spillovers between oil prices and food prices in the sub-Saharan region. Utilizing a GARCH-MIDAS approach for 10 countries, the study reveals that climate risk significantly affects the volatility and interconnectedness of these markets. The findings underscore the importance of accounting for climate risk in assessing the vulnerability of food systems to oil price shocks. Specifically, the results for paper one show a strong positive significance at 1% level of significance for the five variables including oil prices, WUI, GDP per capita, exchange rate, and the dummies used to control for the structural breaks. The three forms of geopolitical risk are net shock receivers, except for the narrow definition, which is a net shock giver. The narrow definition, likely referring to direct conflicts or wars, shows the highest net connectedness in all countries except Nigeria, where it is second to oil. The lagged impact of climate risk on the volatility of the oil-food connectedness index is captured by 𝜃 in paper three. The results show positive and significant values of 𝜃 in six countries (Angola, Ghana, Kenya, Senegal, South Africa, and Tanzania), indicating that, on average, an increase in the climate change variable, like higher temperatures, is associated with an increase in the volatility of the connectedness between oil and food prices. iii Overall, across all three papers, the findings have important implications for policymakers and stakeholders seeking to enhance food security and economic stability in the face of increasing uncertainties. The findings of this thesis underscore the critical need for policymakers in sub- Saharan Africa to adopt a holistic and multi-dimensional approach to address food security challenges in the face of increasing uncertainties. Policymakers should explicitly acknowledge and account for policy uncertainties in their decision-making processes. This involves developing flexible and adaptive policy frameworks that respond effectively to unforeseen shocks and changing market conditions. Given the heterogeneous impact of oil price volatility and uncertainty on different developing economies, policymakers should prioritize targeted interventions that protect the most vulnerable populations. Reducing dependence on imported oil and diversifying energy sources can enhance resilience to oil price shocks. Similarly, promoting agricultural diversification and supporting local food production can reduce vulnerability to global food market fluctuations for papers one and two. Moreover, there is a need to enhance regional cooperation and integration to help mitigate the adverse effects of geopolitical risks and climate change on food security for papers two and three respectively. This may involve sharing information, coordinating policies, and investing in regional infrastructure and early warning systems. Encouraging sustainable agricultural practices and investments in renewable energy can contribute to long-term food security and environmental sustainability. This includes promoting soil conservation, water management, and agroforestry practices. By adopting these policy recommendations, policymakers can create a more resilient and sustainable food system in sub-Saharan Africa, capable of withstanding the challenges posed by oil price volatility, geopolitical risk, and climate change. It highlights the need for tailored policy interventions that address the specific vulnerabilities of different countries and account for the time-varying and asymmetric nature of market interactions.Item The Political Economy of Political Regimes and Trade Flows: A Case of Zimbabwe 2005-2020(University of the Witwatersrand, Johannesburg, 2024) Mufandaedza, Dhibhora; Mondi, Lumkile PatriarchThis thesis delves into the intricacies of the political economy of trade within the context of Zimbabwe. Similar to numerous other developing African nations, Zimbabwe experienced a persistent trade deficit from 2005 to 2020, thus being categorised as a net importer due to consistently higher import volumes compared to exports. Presently, Zimbabwe grapples with one of the most severe crises among countries in the Global South. The economic and political crises situation actually calls into question the fact that Zimbabwe had a Government of National Unity (GNU) between 2009 and 2013 aimed at serving as a transitional authority overseeing the economic and political governance of the nation. This study examined the interaction between political regimes and trade flows in Zimbabwe from 2005 to 2020 and how political regimes affect economic policies, institutions and economic management. This study contributes to extant literature by deepening an understanding of the relationship between politics and trade in the Zimbabwean context. The study was organised into three periods, and was intended to answer the three research questions: What was the state of political relations in Zimbabwe from 2005 to 2020? What was Zimbabwe's trade-flow level between 2005 and 2020 and why? Ultimately, what are the characteristics of the import and export concentrations in Zimbabwe between 2005 and 2020? The first period covered the years 2005 to 2008, i.e. the period before the formation of the Government of National Unity (GNU). The second period analysed the years 2009 to 2013 - A period under GNU. The third period covered the period after the dissolution of GNU, the years 2014 to 2020. The study was premised on an exploratory design; this mixed-methods research relied on secondary data to investigate the above three research questions. To answer the first research question, the study relied on documentary and historical data analysis to determine the status of political relationships in Zimbabwe. To answer the second research question, the study deployed descriptive analysis to describe the evolution of trade flows, taking into account the underlying theoretical frameworks of the political economy namely neopatrimonialism and political settlement framework. Finally, the third research question was based on time trend analysis. The study concludes that political instability had a huge impact on the economy as Zimbabwe faced political and economic challenges for several decades, which negatively affected its trade with other nations. These challenges can largely be attributed to the political and economic decisions made by the ruling ZANU-PF elites. This study also contends that the heterodox economic planning of the ruling government from 2005 to 2020 led to a mix of redistributive and indigenisation policies on the one hand and the acceptance of neoliberalism on the other, which iii has negatively impacted on trade. Zimbabwe’s political environment has also been characterised by severe restrictions on political freedoms as evidenced by the dominant political relationship themes between 2005 and 2020. Some of the identified themes in this study are political tensions, economic challenges and power struggles, factions within the ruling ZANU-PF, protests, and post- election violence, to name a few. All these factors were an obstacle to attracting investors and gaining access to more trading partners. For example, export growth was rather weak during the period analysed. The export basket exhibited reduced diversification and heightened reliance on a limited array of products, indicating a decline in both diversification and technological sophistication of exports. This study attributes the causes of this poor performance of the trade sector to various political economy challenges such as poor economic governance, neopatrimonialism, weak political settlement in the form of GNU that created an unfavourable climate for private investment and trade, and industrial policies (especially indigenisation policies) that undermine investor confidence and discourage private investment. While trade volumes recovered from the deep recession of 2007-2008, this cannot compensate for other worrying longer-term trends unveiled in this study. With regard to recommendations, the good news from this study is that the rectification of these policy deficiencies rests within the hands of the Zimbabwean government. To achieve significant transformation in trade, the Zimbabwean government should consider collaborating with international partners and stakeholders to redefine itself as a major player in the global economy. However, this will pose a significant challenge to the Zimbabwean government as research reveals that democratic governance promotes economic growth and prosperity, while autocratic regimes stifle progress, with China being an outlier in this case. Further to that, literature suggests that authoritarian governments are reluctant to reduce distorting red tapes and other unofficial trade barriers. Another recommendation is that Zimbabwe needs to adopt effective policies to leverage and harness its abundant natural and human resources for increased production and export earnings. The current focus on raw materials exports needs to shift to value-added goods and services. The government should actively attract local and foreign investors to set up production facilities in Zimbabwe that cater aiming for both the domestic and international markets. It is important that the government does not make abrupt policy changes to gain political advantage, as this creates uncertainty and deters investment.Item Enhancing Global Equity Returns with Trend- Following and Tail Risk Hedging Overlays(University of the Witwatersrand, Johannesburg, 2025) Schwalbach, João Bruno Meneses; Auret, ChristoItem A Structuralist Analysis of Income Distribution and its Influence on the Macroeconomy(University of the Witwatersrand, Johannesburg, 2025) Capazario, Michele; Mokoka, TshepoThis study undertakes a comprehensive structuralist analysis of income distri- bution and its macroeconomic implications, framed through three distinct but interrelated essays. These three essays aim to generate novel, theoretical modelling contributions to macroeconomic literature, prior to testing these contributions empirically for 7 countries. The first essay investigates the role of income distribution in maximising real GDP and evaluating cycles in the labour share around some GDP-optimising value, proposing a novel, parsimonious single-equation model to estimate the GDP-optimal labour share without reliance on Cobb-Douglas production functions. Employing Nonlinear Least Squares (NLS) and Dynamic Ordinary Least Squares (DOLS) techniques, the essay demonstrates that the relation- ship between the labour share and output is non-linear and time-dependent, with substantial variations across the seven countries analysed. The second essay explores the dynamic interplay between output, household debt, and the labour share (using the abovementioned novel specification to identify labour share cycles suggested in the first essay) positing that deviations between the actual and GDP-maximising labour share influence both business and financial cycles. Generating a two-equation structuralist framework, the study finds (through employment of a Structural Vector Autoregression - SVAR - approach) that income distribution shocks have asymmetric effects on debt accumulation and output, with wage-led economies in the sample under analysis (Australia, France, Germany, Italy, South Africa, the United Kingdom and the United States of America) exhibiting larger increases in household debt volatility and accummulation, than profit-led economies. The final essay extends the structuralist model found in Flaschel and Krolzig (2003) to an open macroeconomy, integrating wage and price Phillips Curves, monetary policy, and exchange rate dynamics. Employing a large SVAR approach, prior to conducting a stability analysis and utilising a novel, income distribution-sensitive monetary policy rule, the study establishes that shifts in income distribution can exacerbate macroeconomic stability. The findings suggest that structuralist-informed monetary policy frameworks can serve as effective tools for stabilising economies exposed to income distribution shocks. The study concludes by highlighting the policy implications of the findings, advocating for income distribution-sensitive approaches to macroeconomic management.Item The effect of banking sector competition on firm creation in Sub-Saharan Africa(University of the Witwatersrand, Johannesburg, 2025) Ramahuta, Katlego; Gwatidzo, TendaiThis study determines the effect of banking sector competition on firm creation in Sub- Saharan Africa (SSA). The study uses country-level data from multiple World Bank databases covering the period 2006-2021. While the analysis initially included all 48 SSA countries, only 40 countries were part of the final sample due to missing data in 8 countries. For a comprehensive view of the banking sector competition in the region, we use two non- structural measures of competition (Boone Indicator and the Lerner Index) and one structural measure (The concentration ratio calculated based on the top five banks, CR5). The Fixed Effects (FE) model and the two-step system Generalized Method of Moment (system-GMM) estimators are used to estimate the results of the analysis. Employing two estimators enhances robustness of the findings and ensures their consistency across different estimation techniques. Our findings support the market power hypothesis for SSA countries, indicating a positive relationship between banking sector competition and firm creation. This supports the view that increased competition in the banking sector will lead to an increase in the rate of firm creation. Based on the results, economic growth and higher levels of education lead to opportunities for firm creation. Our findings also show that new firm registration rates are lower in upper-income countries compared to lower-income countries in the region. The findings from this paper can inform policy decisions around banking sector competition regulation, and support initiatives that leverage firm creation as a tool to address unemployment and poverty in the region.Item The Bias ratio: An effective fraud identification tool(University of the Witwatersrand, Johannesburg, 2025) Haddad, Remon; van Vuuren, GaryFinancial fraud poses significant risks with far-reaching consequences, particularly in the context of growing assets under management and expanding equity markets. This thesis underscores the urgent need for robust measures to safeguard investors from fraudulent activities by exploring the consequences of notorious fraud cases such as Bernie Madoff’s Ponzi scheme. Through analyses of hedge fund, index fund and stock price return data in the US and SA, over various periods starting in 1997 to 2024, it becomes evident that tools such as the Bias ratio, kurtosis, and skewness can serve as effective mechanisms for detecting fraudulent behaviour. The Bias ratio emerges as a dual-purpose tool. Beyond its fraud detection capabilities, it functions as a performance measurement metric akin to the Sharpe ratio, offering additional value during security analysis. By highlighting suspicious historical outperformance and signalling securities with unusual performance patterns, the Bias ratio enriches the evaluation process, enabling investors to make informed decisions and avoid fraudulent investments. This thesis demonstrates the efficacy of the Bias ratio by examining its application in the notorious Madoff case, where it successfully flagged fraudulent activity that was overlooked by traditional measures like the Sharpe ratio. The findings emphasize the critical role of the Bias ratio in validating the legitimacy of returns and enhancing investor protection.Item The interplay of energy access and labor market outcomes in South Africa(University of the Witwatersrand, Johannesburg, 2025) Moyo, Nomathamsanqa; Adetutu, MorakinyoThis study investigates the relationship between energy access and employment opportunities in South Africa using the Autoregressive Distributed Lag (ARDL) model, analyzing macroeconomic data spanning 1990 to 2022. Unlike previous research focused on rural areas, this study adopts a macro-level approach, encompassing both urban and rural regions. The findings reveal a positive short- and long-run association between energy access and employment, indicating that increased electricity access correlates with higher employment levels. Inflation negatively affects employment, while foreign direct investment (FDI) and gross domestic product (GDP) have positive associations. Trade is significant in the short run but becomes insignificant in the long run. These results highlight key policy implications: expanding access to electricity, particularly in underserved areas, and investing in renewable energy infrastructure can enhance employment opportunities. Attracting FDI through improved infrastructure and incentives, implementing sound monetary policies to control inflation, and supporting economic growth through small and medium enterprises and innovation are critical for sustaining employment. Additionally, addressing South Africa’s high dependence on imports and promoting export-oriented industries can amplify the long-term benefits of trade on employment, emphasizing the need for targeted, structural policy interventions.Item Predicting Wind Energy Production in South Africa Using Machine Learning(University of the Witwatersrand, Johannesburg, 2025) Reddy, Sivasha; Kutela, DambalaSouth Africa faces an urgent and escalating energy crisis driven by ageing coal infrastructure, frequent load shedding, and rising electricity demand. Wind energy presents a viable renewable energy alternative with significant potential to alleviate these challenges; however, its inherent variability complicates grid stability and energy planning. Accurate wind energy forecasting is essential for optimising power dispatch, minimising curtailment, and enhancing energy security. Despite advancements, traditional forecasting methods, such as physical models and statistical techniques, struggle to capture the complex and nonlinear nature of wind patterns, particularly in data-scarce environments like South Africa. This study investigates the application of machine learning models to improve wind power forecasting in South Africa, where data constraints and fluctuating meteorological conditions pose unique challenges. The research examines the effectiveness of machine learning in predicting wind energy production and assesses the role of explainable artificial intelligence techniques, such as SHapley Additive exPlanations, in enhancing model transparency and interpretability. Using historical meteorological data and turbine performance records from a South African independent power producer, the study evaluates multiple machine learning approaches to determine their predictive performance. A comparative analysis of different machine learning models highlights the most reliable techniques for wind energy prediction. The findings demonstrate that XGBoost outperforms Random Forest, Decision Tree, and K-Nearest Neighbour. Furthermore, the machine learning methods show a significant improvement over traditional statistical techniques, offering improved predictive accuracy while providing insights into key meteorological and operational factors influencing wind power generation. The integration of explainable artificial intelligence further ensures interpretability, fostering trust and practical usability among stakeholders. This research contributes to the renewable energy forecasting literature by adapting machine learning solutions to a data-scarce environment and emphasising the role of interpretability in real-world adoption. The results provide valuable insights for policymakers, energy planners, and grid operators, supporting South Africa’s transition to a more sustainable and resilient energy future. However, the study's findings may be limited in generalisability and accuracy due to the analysis focusing on a single wind turbine chosen based on data availability rather than representativeness. Consequently, the results may not extend to other wind farms or turbines operating under different geographic or climatic conditions.Item Forecasting Exchange Rate Dynamics: A Comparative Study of Traditional Econometric Models and Machine Learning Models(University of the Witwatersrand, Johannesburg, 2025) Ndlovu, Teyven; Farrell, GregoryAccurate exchange rate forecasting is crucial for economic policy, risk management, and financial decision-making. This study compares traditional hybrid econometric models, specifically the Autoregressive Integrated Moving Average with Exogenous Variables and Generalized Autoregressive Conditional Heteroskedasticity (ARIMAX-GARCH), with machine learning approaches, including Random Forest, Multi-Layer Perceptron, and Long Short-Term Memory networks, to predict the South African Rand (ZAR) against major global currencies: the United States Dollar (USD), Euro (EUR), British Pound (GBP), Japanese Yen (JPY), and Chinese Yuan (CNY). Using daily exchange rate data from 2000 to 2024, model performance is evaluated based on Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Symmetric Mean Absolute Percentage Error (sMAPE), and Mean Directional Accuracy (MDA). The results indicate that no single model consistently outperforms across all currency pairs. The ARIMAX-GARCH model excels in trend prediction, the Random Forest model balances predictive accuracy and adaptability, the Multi- Layer Perceptron model minimizes absolute errors but struggles with directional accuracy, and the Long Short-Term Memory model captures long-term dependencies but underperforms in volatile markets. These findings highlight the need for hybrid forecasting models that integrate machine learning and econometric techniques while incorporating macroeconomic indicators to enhance predictive reliability.