Thesis
Do companies use ESG criteria as a bargaining tool with investors, just like dividends? This is the question I explored while developing this thesis.
Dividends are a key element in the relationship between companies and shareholders. They send a clear signal about a company’s financial health and stability. A company that consistently pays dividends inspires confidence, it demonstrates that it generates enough profits to reward its investors while sustaining its operations. However, with the rise of Environmental, Social, and Governance (ESG) criteria, a new dynamic has emerged. Today, more and more investors are turning to companies that uphold responsible practices, sometimes at the expense of immediate financial returns.
This leads to my reflection: Do companies with low ESG scores compensate by paying higher dividends to attract shareholders? Conversely, do companies with strong ESG performance use this as a justification for a more moderate dividend policy? In other words, does one replace the other, or can both coexist to appeal to investors seeking financial returns as well as those prioritizing sustainability?
To answer this question, I analyzed a large sample of companies from 2015 to 2023, examining their dividend distribution policies and ESG scores. I also explored whether these trends varied based on region, company age, or industry sector.
The results reveal very different strategies. Some companies manage to strike a balance between high dividends and ESG commitment, maximizing their appeal to a broad range of investors. Others, on the contrary, adopt a substitution approach, where ESG efforts serve as a justification for reducing dividend payouts.
This thesis seeks to answer a key question for the future of finance: Do companies still need to attract investors with dividends, or has ESG become a new tool for financial appeal? Far from being a simple binary choice, the reality is much more nuanced, offering new insights into how companies balance profitability and responsibility.
If this topic intrigues you and you’d like to learn more, I invite you to read my thesis to explore the detailed findings of this analysis.
Tutored Project
As part of my bachelor’s degree, we were required to conduct a research project in finance, and I focused on portfolio analysis and investment management through the lens of three legendary economists: George Soros, Peter Lynch, and Warren Buffett.
The goal was to examine their respective approaches and compare their investment strategies through a deep analysis of their portfolios. Each of these investors embodies a distinct vision of the markets: Soros: the bold and opportunistic speculator, Lynch: the advocate of diversification and stock picking and Buffett: the long-term value investing master.
Through this project, I aimed to understand how these strategies work in practice, their strengths and limitations, and how they can be applied in today’s investment landscape. By leveraging performance indicators, risk ratios, and financial analysis models, I dissected their decisions and highlighted the underlying logic behind building a high-performing portfolio.
This experience significantly refined my understanding of financial markets and deepened my grasp of the trade-offs between risk and return that drive investment decisions. If you're interested in learning more about these investment strategies and their impact on asset management, I invite you to explore my research further.
As part of the Applied Econometrics and Machine Learning in Economics course at HEC Montréal, we conducted a replication study on the impact of debt relief programs in developing countries. We chose to reproduce and expand upon the study What Does Debt Relief Do for Development? Evidence from India’s Bailout for Rural Households by Martin Kanz, which examines the effects of one of the largest debt forgiveness programs in history, implemented in India in 2008.
Our work went beyond a simple replication of the original results. We provided additional insights by testing new econometric models and integrating machine learning techniques to refine the analysis. In addition to the regression discontinuity design used in the original study, we applied LASSO regression to select the most relevant variables and employed dimensionality reduction techniques to test the robustness of the results. We also conducted additional sensitivity tests to explore the heterogeneous effects across different household categories.
Our findings confirm the main conclusions of the original study: debt forgiveness had no significant impact on investment or agricultural productivity among beneficiary households. However, our broader approach revealed differentiated effects based on household characteristics, suggesting that the impact of such policies strongly depends on the economic context and local constraints.
This project was a stimulating experience, combining advanced econometrics, machine learning, and critical policy analysis.
As part of the Risk Management course, I explored the impact of dividend policies on financial risk management. The goal was to understand how dividends influence risk perception, stock volatility and investor decisions.
To examine this question, I relied on three complementary approaches from academic literature. The first analyzes the impact of dividends on return volatility, assessing whether they stabilize or amplify market fluctuations. The second examines the relationship between firm size, leverage, and risk perception, exploring how structural factors influence investor reactions to dividend policies. The third takes a behavioral perspective, applying prospect theory to study how investors interpret dividend changes and their risk implications.
Beyond analyzing theoretical and empirical models, I incorporated a critical dimension by questioning their limitations and real-world implications. For instance, while some studies suggest that dividends serve as a positive signal to investors, others argue that their effect is distorted by market conditions or internal corporate strategies. I aimed to deconstruct common assumptions and assess under what conditions dividends actually reduce risk or, conversely, introduce additional uncertainty.
My contribution was to bridge these different analyses, highlighting underlying dynamics that are often treated separately in the literature. I sought to determine when dividends reassure markets and when they might instead amplify uncertainty. This perspective allowed me to nuance traditional conclusions and explore the role of dividends as a strategic tool for risk management.
This project provided me with a broader understanding of the interactions between corporate finance, risk management, and investor behavior. If these questions interest you, I invite you to explore this analysis in more detail!
As part of the Economic Behavior and Project Evaluation course, we conducted a study on the factors influencing consumer decisions between installing solar panels for self-consumption and purchasing energy from the market.
The objective was to understand which economic, environmental, and behavioral factors drive households to opt for solar self-consumption rather than relying on the electricity grid. To achieve this, we combined several analytical approaches:
By combining these different approaches and cross-referencing our results, we identified the real drivers of the household energy transition. Contrary to common assumptions, our study shows that financial incentives are not necessarily the decisive factor and that technological accessibility and perceived profitability play a key role.
Thus, our goal was to highlight the limitations of current incentives and identify potential levers to further encourage solar self-consumption. If you are interested in learning more about the economic dynamics behind the energy transition, I invite you to explore this study in detail!
As part of the Financial Management course, we conducted an in-depth analysis of Metro Inc.'s capital structure and dividend policy. The objective was to evaluate whether the company follows an optimal financial strategy regarding debt financing and profit distribution.
We examined Metro Inc.'s financing choices, comparing its leverage and debt levels to those of its competitors in the grocery retail sector. Our analysis revealed that Metro maintains a lower debt level than the industry average, which reduces its financial risk but also limits its leverage effect.
Regarding its dividend policy, we analyzed the evolution of shareholder payouts and share buybacks to determine whether Metro prioritizes profit redistribution or reinvestment. A key aspect of our project was comparing these findings with financial theories, particularly clientele effects, agency cost reduction, and trade-offs between debt and dividends.
Our value-added approach was to take a critical perspective, questioning whether Metro’s current strategy is truly optimal given its mature market position and low debt levels. We highlighted the benefits and drawbacks of increasing leverage and the strategic implications of an aggressive distribution policy.
Case Study: Stock Valuation and Market Outlook of Canada Goose
As part of the Financial Investments course, we conducted an in-depth analysis of Canada Goose Holdings Inc., aiming to determine whether its stock was fairly valued by the market.
The objective was to combine multiple valuation approaches to arrive at a coherent estimate of the stock's fair value. We first examined the company’s strategic positioning, considering its international expansion, sales seasonality, and reliance on luxury fashion trends. We then applied various valuation methods, including the Discounted Cash Flow (DCF) model and market multiples, to compare our valuation with observed market prices.
One of the most insightful aspects of this study was analyzing the impact of economic and sector-specific risks on the stock’s valuation. We assessed the sensitivity of the target price to fluctuations in capital costs and growth prospects. Our findings suggested that the stock was slightly undervalued, although several uncertainties, particularly related to Asian market demand and brand image management, could influence its future performance.
I wanted to explore a different format by creating a podcast with artificial intelligence, focusing on program evaluation and causal inference methods. The goal was to see how an interactive audio format could make these concepts more accessible and concrete.
One episode explores randomized experiments and their central role in applied econometrics. It begins with a simple scenario: why do we need random assignment in both economics and medicine to measure causal effects? The episode explains how a well-designed experiment eliminates selection bias and provides a credible estimate of the Average Treatment Effect (ATE).
The AI then presents an example of a social assistance experiment, where one group receives cash transfers while another does not, in order to evaluate the impact on employment and consumption. The episode also discusses the limitations of randomized experiments, such as attrition issues and lack of generalizability, while exploring alternative methods like natural experiments and quasi-experimental approaches.
This podcast was an opportunity to experiment with a new way of popularizing applied econometrics, blending analytical rigor with fluid storytelling. While the AI format has its limitations, it offers an interesting way to explore how academic learning can adapt to more dynamic and accessible formats.