CLSBE - Dissertações de Mestrado / Master Dissertations
URI permanente para esta coleção:
Navegar
Entradas recentes
- When crime meets business : understanding the impact of crime on firms in BrazilPublication . Iovino, Aniello; Silva, JoanaThis thesis investigates the impact of crime on entrepreneurial performance in Brazil using firm-level data from the World Bank Enterprise Survey merged with municipal crime statistics. The final sample includes 1,290 firms. Crime exposure is measured using an index constructed through principal component analysis of robberies, thefts, and homicides.Baseline OLS estimates indicate a negative relationship between crime and firm performance. A one standard deviation increase in the Crime Index is associated with 7.3% lower sales, 6.1% lower value added, and 4.9% lower productivity. To address endogeneity, crime exposure is instrumented using leave-one-out state averages. Despite limitations, the results robustly suggest that crime harms firm performance.An ordered logit analysis examines how crime exposure affects firms’ distribution across performance categories: micro enterprises, small and medium enterprises (SMEs), mid-cap firms (250–499 employees), and large enterprises. Higher crime exposure increases the probability of firms being classified in lower performance categories while reducing the probability of being in higher categories, indicating that crime affects firms’ position within the performance distribution.The analysis reveals substantial regional and sectoral heterogeneity. Crime shows the strongest negative relationship in the Southeast, moderate impacts in the South, and negligible association in the Centre-West, North, and Northeast. Tobit regressions accounting for left censoring confirms these findings. Overall, the results indicate that crime represents a constraint on business performance in Brazil.
- AI integration in portuguese higher education : evidence from curricular changes before and after chatGPTPublication . Ferreira, Ana Filipa Marques; Matos, Miguel Godinho deThe integration of artificial intelligence into labor markets worldwide is creating demand for new competencies among workers. Universities face pressure to respond by preparing students with relevant AI skills, and ChatGPT’s release in November 2022 intensified this discussion. We investigate this response by analyzing whether and how Portuguese universities have integrated AI content into their curricula. We built a dataset combining data from three official sources: the higher education registry, accreditation agency records,and official government publications. The resulting data includes detailed curricular plans for over 3,000 distinct programs accredited between 2011 and 2025. We classified curricular units by AI content and computed metrics of AI integration for each program in each year.This allowed us to compare programs re-accredited before versus after ChatGPT’s release. Our results show that programs re-accredited post-ChatGPT are more likely to include AI content, though effects remain modest overall. The response is highly uneven across fields, with strong increases in engineering but essentially no change in arts, humanities,and health. Additionally, we observe considerable growth in new AI-specialized programs created since 2020. Overall, the results suggest that Portuguese universities are responding by integrating AI content, though the response remains in its early stages.
- The impact of artificial intelligence on sciencePublication . Rodrigues, André Costa; Matos, Miguel Godinho deIn this thesis, we analyzed how Artificial Intelligence may be impacting academicarticles, and two dimensions were considered: Writing and Topics Addressed. We applieda Latent Semantic Analysis (LSA) model to a dataset of academic articles published from2010 to 2025 to compute their similarities, and used two other models to add robustness.Using this model, we observed what there appears to be some suggestive evidence that theuse of AI tools may be, in a statistically significant way, causing an increase in writing andtopic similarity, and changing the frequency of words used. We finalize our analysis bysuggesting that in the field of Computer Science, those who are more prominent in using AItools have had a significant increase in the similarity of their works in the writing dimension.We conclude with what we believe to be the most crucial next step for a more complete andconclusive study of the impacts of AI on Science.
- Venture capital without borders? : a gravity model analysis of inbound cross-border venture capital flows in europe, 2007-2024Publication . Solli, Simen; Shuwaikh, FátimaThis thesis examines the drivers of cross-border venture capital flows into European countries from 2007 to 2024 and explores how these factors vary across different macroeconomic regimesand regions. Using a panel of Refinitiv deal-level data aggregated into bilateral country-pair flows for 32 European host countries, the study estimates a gravity model using Poisson pseudo-maximum likelihood. The analysis considers both the number of cross-border deals and the total investment amount, including fixed effects and interaction terms to capture changes across the periods 2007-2012, 2013-2019, and 2020-2024. Key findings show that origin-country economic mass (GDP) is a strong positive predictor of inbound venture capital activity, while hostcountry market size has no positive effect (and is negative or insignificant in some specifications). Geographical and informational proximity significantly facilitates cross-border venture capital flows. Shorter distances and shared language or borders are associated with more deals and, more weakly, with higher investment amounts. In contrast, host-country macro-financial conditions do not show a consistent significant influence on venture capital inflows. The results reveal limited changes over time across the three periods, but there is some evidence that the effect of distance on invested amounts was smaller from 2013 to 2019. Regional differences are noticeable. Small Nordics and peripheral European markets appear more sensitive to distance and host-country fundamentals than the major venture capital hubs. These findings highlight the importance of economic size and information frictions in shaping cross-border venture capital flows in Europe, suggesting that integration challenges remain for smaller ecosystems despite recent market growth.
- ESG performance and corporate venture capital activity : evidence from european listed parent companiesPublication . Olivier, Alexander Maximilian; Shuwaikh, FátimaThis dissertation investigates the link between corporate venture capital (CVC) activity and non-financial corporate performance for publicly listed European parent companies. Using firm-year panel data that consolidates CVC transactions at the parent-company level, the analysis explores two types of relationships. First, whether firms with higher ESG performance are more likely to participate in CVC and to invest more heavily, and second, whether CVC activity correlates with future governance outcomes, including tests for environmental and social pillar scores and reported Scope 1+2 greenhouse gas emissions. Panel regressions with year fixed effects are estimated, with random effects as the baseline based on Hausman-test results and firm-clustered standard errors. Robustness checks include fixed-effects models, lagged-ESG analyses, timing tests for lead outcomes, interaction models, and alternative outcome measures. The findings show that higher ESG performance is positively linked to CVC involvement, especially regarding the likelihood of being CVC active (extensive margin), while the association with deal intensity is less precise. Conversely, higher CVC activity is consistently associated with better governance pillar scores, with timing models indicating the effect is strongest in the short term and diminishes over longer periods. In contrast, the relationships with environmental and social pillars are weaker, and no significant link is observed between CVC activity and “hard” emissions outcomes.
- Crowding-Out effect : the impact of government debt on corporate leveragePublication . Silvério, Inês Margarida Silva; Bonfim, DianaThe objective of this master’s thesis is to investigate the crowding out effect, that is, the impact that increasing government debt has on corporate leverage. This relation is studied in 40 countries between 1990 and 2024. This research has Government Debt-to-GDP as the main variable of interest Government Debt-to-GDP. Additionally, the regression analysis also includes macroeconomic variables, as GDP per capita, unemployment rate and consumer price index, firm-specific control variables, including firm size and ROA leverage. Time and country fixed effects are also considered. The analysis is split into two phases: country level and firm level. In the first regressions with country data, the results show no significant relationship between public and private debt. However, when the timeframe is shortened to 1990-2014, significance emerges. As such, an investigation is conducted to identify whether extraordinary measures that happened after 2014, such as quantitative easing policies (QE) and the COVID-19 pandemic, could be impacting the results. Interaction regressions show that for countries implementing QE, the negative link between government and corporate debt is no longer statistically significant. A similar result is found for the pandemic period. This thesis concludes that Government Debt-to-GDP has a crowding-out effect on corporate leverage but is highly muted by monetary and fiscal policy, more specifically, the ones that took place following the Global Financial Crisis and during COVID-19.
- Equity valuation of Siemens Energy AG : a multi-method assessment in the context of the global energy transitionPublication . Horndasch, Niklas; Martins, José Carlos TudelaThis thesis employs a multi-method valuation framework to determine the equity value of Siemens Energy AG as of 01.10.2025. Siemens Energy is a global provider of power generation, grid technologies and industrial decarbonisation solutions, positioning the company at the forefront of the ongoing energy transition. While long-term demand fundamentals are strong, recent operational challenges, particularly at Siemens Gamesa, have resulted in volatile financial performance and heightened investor uncertainty. The analysis integrates a discounted cash flow (DCF) model, an adjusted present value (APV) framework, forward-looking trading multiples and Monte Carlo simulation. Segment-level forecasts draw on projections from the IEA and European Commission and incorporate structural trends such as renewable capacity expansion, grid modernisation and hydrogen-related investment. A global peer comparison ensures that valuation outcomes remain consistent with sector-wide market expectations. The DCF model yields an equity value of EUR 78.46 per share, while the APV framework results in EUR 79.27 per share, primarily due to differing treatments of leverage and tax shields. Forward-looking EV/EBITDA and P/E multiples indicate a valuation range of EUR 74 - 76 per share, broadly aligned with intrinsic estimates. Monte Carlo simulation highlights the sensitivity of results to WACC, margin recovery and terminal growth, producing an average equity value of EUR 78 per share.Overall, the findings suggest that Siemens Energy’s market valuation does not fully reflect its long-term strategic positioning within the energy transition, while short-term execution risks remain a major influence on investor sentiment.
- Equity valuation of A.P. Møller–Mærsk A/SPublication . Peisser, Luca; Martins, José Carlos TudelaThis thesis conducts an equity valuation of A.P. Møller–Mærsk A/S (Maersk), one of the largest global container liner operators with significant operations in logistics and terminals. The primary approach is a sum-of-the-parts discounted cash flow valuation using Free Cash Flow to the Firm (FCFF), discounted at the Weighted Average Cost of Capital (WACC). A complementary relative valuation based on trading multiples is also applied. To implement the intrinsic valuation, the thesis develops an independent and data-driven view of future developments in the container liner market and freight rates. The analysis incorporates an anticipated multi-year period of overcapacity and depressed freight rates, followed by a normalised mid-cycle terminal-period freight rate, and uses scenario and sensitivity analysis to assess the results’ dependence on key assumptions. The valuation is benchmarked against an equity research report by ABG Sundal Collier, which published a target price of DKK 16.426 on 29 September 2025. To ensure comparability, the valuation date is set to 30 September 2025, and no information beyond this date is incorporated. The resulting target share price is DKK 17.279, implying 39% upside versus the prevailing share price of DKK 12.465 at the valuation date. Based on both intrinsic and relative valuation, the results indicate that Maersk is undervalued as of the valuation date, even under pessimistic assumptions.
- Certification or strategic hype? : venture capital backing and IPO pricing in the modern U.S. market (2010–2024)Publication . Linden, Youp Eric van der; Shuwaikh, FátimaThis thesis examines whether venture capital (VC) backing in the modern U.S. IPO market primarily acts as value certification or instead reflects momentum-creating incentives through strategic underpricing. Using a hand-collected sample of 1,001 U.S. common-stock IPOs listed on NYSE and Nasdaq between 2010 and 2024 from SDC Platinum, the study analyzes first-day underpricing, gross underwriting spreads, and post-IPO market-adjusted performance. Long-run performance is measured using market-adjusted buy-and-hold abnormal returns (ABHARs) over horizons of 28 to 365 trading days. The empirical approach combines multivariate OLS regressions with year and industry fixed effects and a propensity-score-based inverse probability weighting (IPW) framework to address endogenous selection into VC backing. The results show that VC backed IPOs experience significantly higher underpricing, with first-day returns approximately 10–12 percentage points larger, and pay higher underwriting fees, with gross spreads about 20–30 basis points higher. In contrast, VC-backed IPOs do not exhibit superior post-IPO market-adjusted performance, as any initial outperformance dissipates over time. Overall, the findings are more consistent with grandstanding and momentum-creating mechanisms than with pure certification.
- The impact of minimum wage increases on capital structure : evidence from PortugalPublication . Freitas, Tiago Luís Carvalho; Bonfim, DianaOver the last decade, Portugal has experienced one of the steepest increases in the national minimum wage of the European Union countries, following a long period of wage freeze under the Economic Adjustment Program. This dissertation studies how Portuguese firms adjustedtheir leverage during the sequence of minimum wage hikes. Using employer–employee data from Quadros de Pessoal and firm accounts from SCIE between 2010 and 2022, the analysis compares firms that were highly exposed to the minimum wage in 2014 with less exposed firms. By using a difference in differences approach with propensity score matching, the results show that highly exposed firms increased leverage relative to less exposed firms, and that this pattern is concentrated among low profitability firms. However, treated and control group firms did not follow parallel pre trends, showing that these firms already had diverging financing trends before 2014.