CLSBE - Dissertações de Mestrado / Master Dissertations
Permanent URI for this collection
Browse
Recent Submissions
- Synthetic data, real impact : a framework for augmenting tabular datasets with synthetic data in machine learningPublication . Bitzer, Jann Noah; Fernandes, Pedro AfonsoAccess to high-quality data is an ever-occurring challenge in machine learning due to scarcity, cost, privacy constraints, and biases. While synthetic data has gained traction in large-scale AI applications to overcome these challenges, its practical implementation for small to mid-size businesses remains underexplored. This study bridges this gap by developing a structured and universal framework to integrate synthetic data augmentation into various machine learning processes. The approach systematically assesses augmentation ratios, selective filtering strategies, and their impact on predictive performance. This research provides a scalable and actionable framework for businesses to use synthetic data, offering practical guidance on augmentation strategies and performance evaluation. By addressing technical and ethical considerations, this study advances the adoption of synthetic data as a transformative tool for data-driven decision-making in business environments.
- Adoption of AI tools in software development in Germany : curse or blessing for the SME sector?Publication . Blischke, Robin; Rajsingh, PeterThis study investigated the adoption of Artificial Intelligence (AI) tools in Germany's software development sector, focusing on small and medium-sized enterprises (SMEs). AI technologies are reshaping the industry, presenting opportunities and challenges. Using a mixed-methods approach, including 14 expert interviews and a survey of 262 participants, the research identified key factors affected by AI adoption, such as efficiency gains, code quality, competitive pressure, security risks, organizational readiness, technical debt, and ethical concerns. These factors were identified through a literature review, tested via expert interviews, and validated through a survey, adhering to the principle of triangulation. Currently, AI adoption in German SMEs remains in its infancy, primarily focused on enhancing productivity in routine tasks, with strategic integration still limited. Expert insights highlighted SMEs' agility in deploying off-the-shelf AI tools but noted constraints from limited resources and technical expertise. In contrast, large enterprises (LEs) leverage robust infrastructure and R&D investments for more comprehensive AI integration. While AI tools were viewed as an efficiency innovation, findings indicated their disruptive potential to democratize coding, bridge skill gaps, and drive long-term transformations. However, systemic barriers, including security vulnerabilities, ethical dilemmas, and insufficient organizational readiness, continue to hinder widespread adoption. By integrating dynamic capabilities and innovation theories, this research extrapolated AI’s trajectory from incremental efficiency gains to disruptive innovation, fundamentally altering workflows and competitive dynamics. The study offers actionable recommendations to foster readiness, address ethical and security concerns, and promote targeted upskilling for a transformative future.
- Echo chambers in a social finance platform and option-implied moments of stock performancePublication . D'Isep, Lorenzo; Karimli, TuralThis thesis examines the role of echo chamber in financial markets. Using measures built by Cookson et. al. (2022) and variables based on option prices (DeMiguel et. al. (2013)), I analyzed how selective exposure shapes investors' perceptions of implied risks and expected excess stock returns. The results highlight that increased disagreement expressed by users stimulates trading activity, in line with Cookson et. al. (2022). Next, I document that a greater dispersion in the messages a member receives leads to a higher probability of extreme payoffs; higher implied skewness, volatility and expected returns. The analysis shows that the investors' tendency to interact with information that confirms their pre-existing beliefs, creates a polarised environment that biases users' trading decisions and market stability in the short run.
- Market reactions to the corporate sustainability due diligence directive : insights from the European UnionPublication . Byback, Diana Katarina; Venter, ZoëThe objective of this thesis is to study the market impact of the Corporate Sustainability Due Diligence Directive (CSDDD) on EU companies, which was proposed in February 2022 and implemented in July 2024. Using an event study methodology, the research assesses affected firms9 abnormal returns during periods surrounding the directive's announcement and implementation, focusing on variations within two and five-day windows. It also investigates how firm-specific control variables like emissions and ESG factors affect these returns and examines industry-specific differences. Results indicate a significant negative market reaction to both the proposal and implementation of the CSDDD, particularly in the longer window analysis. The ESG score,along with the environmental and governance pillar scores, significantly impacted returns, with pronounced variances across industries based on the levels of emission and ESG variables. These findings offer valuable insights for investors and policymakers regarding the directive's implications for affected companies.
- NVIDIA corporation equity valuationPublication . Vigário, João Maria Medeiro Godinho Dias; Martins, José TudelaThis thesis explores the application of equity valuation to NVIDIA Corporation, a leading player in the semiconductor and artificial intelligence industries. The primary objective is to determine NVIDIA’s intrinsic value through two valuation methodologies: the Adjusted Present Value (APV) model and market multiples. The APV model yields a target share price of $105 as of January 26, 2025, supported by a relative valuation analysis to ensure consistency with the market pricing of comparable companies. When measured against the market value on November 20, 2024, the findings indicate a potential -28% return, leading to a SELL recommendation. Finally, these results are contrasted with equity research published by Deutsche Bank, highlighting differences in assumptions and valuation methodologies.
- Pricing sustainability risks : ESG scores and stock returns in emerging marketsPublication . Eisen, Lisa Maria; Venter, ZoëSustainable investing has become an increasingly important topic with investors seeking ethical investment options and financial gains through risk mitigation. Extensive research exists on how investors incorporate sustainability in U.S. and European markets. In emerging markets, investor responses to sustainability risks are less explored. My research aims to answer whether investors in these markets demand higher risk premia for firms more exposed to ESG risks. Using emerging market data from 2014 to 2023, I examine how ESG, ESGC, and Social Scores affect stock returns through cross-sectional and time-series analyses. The findings indicate that investors demand a premium on stock returns for higher ESG risk exposure during economic crises. Further, profitability moderates the impact of ESG risks in emerging markets. Finally, country-and-industry-specific characteristics significantly influence sustainability risk premia. Overall, this study contributes to the understanding that investors price sustainability-related risks in emerging markets.
- Hedging against warmer days : the relationship between local temperature and stock returnsPublication . Freitas, Martim Cascais de; Silva, António Baldaque daThis research aims to find how mood swings regarding climate change affect stocks returns and what strategy could be used to hedge against these changes in perception. Based on prior literature I use local abnormal temperatures as a predictor of changes in the attention given to global warming. I assess the impact abnormal temperatures have on two different long-short portfolios for each of the seven country indices studied; one is based on industries, the other is based on Environmental (E) Scores provided by the London Stock Exchange Group (LSEG,previously Refinitiv). I find that stocks from companies operating in polluting industries as defined by the Intergovernmental Panel on Climate Change (IPCC) tended to underperform those who operate in clean industries in periods of high abnormal temperatures from 2016 to 2022. On a parallel experiment I also find that stocks with low E Scores tend to overperform stocks with high E Scores over the same period. These phenomena are observed essentially in the European Indices. Facing results pointing in opposite directions I study the relationship between E Scores and the industries companies operate in. I find that for the last 22 years,operating in high emissions industries as defined by the IPCC is associated with higher LSEG/Refinitiv E Scores.
- Determinants of long-term stock performance : how Domino’s IPO outperformed its 2004 cohort and surpassed alphabetPublication . Azeredo, Maria da Assunção Machado Nunes Mexia de; Cerqueiro, GeraldoThe dissertation investigates the exceptional long-term performance of Domino’s following its 2004 Initial Public Offering (IPO), comparing it to 85 other companies that went public in the United States during the same year. Inspired by various press articles claiming that Domino’s outperformed not only its industry peers but also tech giant Google (Alphabet Inc.), this study aims to validate these assertions while identifying the key determinants of long-term IPO performance in general and the specific factors behind Domino’s success. The findings confirm that Domino’s delivered the highest Long-term returns among its 2004 cohort, challenging traditional assumptions about the dominance of technology companies in generating sustainable financial growth. Regression analysis reveals that Domino’s re- markable performance can be attributed to its notably high Sales-to-Assets ratio and significant leverage at the time of the IPO, which reflected operational efficiency and financial discipline. Furthermore, the research acknowledges that Domino’s success extends beyond these initial conditions, emphasising how corporate and strategic decisions, including early adoption of digital innovations and a scalable franchising model, enabled the company to thrive across global markets and sustain its exceptional performance.
- The impact of interest rate volatility on the leverage choices of S&P500 US-listed firmsPublication . Miranda, Matilde da Gama; Bonfim, DianaThis thesis examines how interest rate volatility affects the borrowing decisions of S&P500 listed companies. The study includes data from 205 S&P 500 companies from 2000 to 2023 in order to explore how interest rate volatility impacts capital structure decisions. By using panel data regression models, it analyzes both long-term and short- term interest rate changes, along with firm-specific factors, to understand leverage decisions during a 24-year period. The results show that interest rate volatility has a big impact on leverage, especially long-term, which discourages borrowing due to financial risks. Firm-specific factors like cash flow strength, size, and market-to-book ratios play a key role in shaping borrowing strategies. Industry-specific differences matter less overall, though Technology and Telecommunications firms are more likely to reduce leverage when faced with long-term volatility. This is probably because these sectors rely on intangible assets and have less predictable revenues. For most other industries, macroeconomic conditions and firm-level traits are more important. These findings give useful insights for financial managers to create flexible borrowing strategies and for policymakers to ensure stable borrowing conditions during periods of economic uncertainty.
- Deepwater Horizon oil spill : stock performance effects on the energy industryPublication . Pereira, Pedro Miguel da Fonseca; Stahl, JörgThis dissertation investigates the financial implications of the largest marine oil spill in history (Deepwater Horizon oil spill) on the Oil and Gas and Alternative Energy industries. Employing event study and regression methodologies, the analysis spans US and European markets to assess intra-industry and cross-industry effects, with a focus on the role of Gulf of Mexico exposure in driving abnormal stock returns. Findings confirm significant negative market reactions for BP and its partners, with severe negative cumulative abnormal returns reflecting strong intra-industry contagion. The research further highlights the disproportionate impact of negative news on market reactions compared to positive developments, suggesting investors tend to overreact to adverse events. Additionally, exposure to the Gulf of Mexico during the spill was identified as a key driver of abnormal returns. The study provides key insights for stakeholders in assessing financial risks associated with environmental disasters and offers a foundation for further exploration into long-term industry ramifications, such as regulatory impacts and shifts in energy policy.