| dc.description.abstractEN | This thesis, Three Essays on Quantitative Methods for Finance, comprises three independent essays that develop and apply rigorous empirical methods to address central questions in mutual fund performance and credit‐rating measurement. Each chapter contributes novel methodological insights and empirical evidence to longstanding debates in quantitative finance.
The first essay re-examines the existence of diseconomies of scale in the active mutual fund industry. While theory predicts that fund size should constrain managers’ ability to generate abnormal performance, prior empirical studies, largely based on traditional risk-adjusted returns, have produced mixed results. This ambiguity is exacerbated by equilibrium considerations that drive expected alphas toward zero. Shifting the focus from percentage returns to value creation, the chapter adopts a novel bootstrap methodology to compare observed value added with a simulated benchmark absent scale effects. The results provide robust evidence that fund size imposes binding constraints on managers’ ability to create value, particularly among highly skilled managers, thereby offering new support for diseconomies of scale.
The second essay investigates how major U.S. regulatory reforms implemented between 2000 and 2003 reshaped the corporate information environment and, in turn, mutual fund performance. By improving disclosure quality and reducing information asymmetry, these reforms curtailed mispricing opportunities and limited managers’ ability to exploit private information. Consistent with these mechanisms, the analysis documents a pronounced contraction in the cross-sectional dispersion of mutual fund alphas and a sharp decline in performance persistence following the reforms. Falsification tests using funds investing in foreign markets unaffected by the regulations support a causal interpretation.
The third essay addresses a methodological issue in the credit ratings literature: the common practice of representing ordinal credit ratings with a linear numerical scale. Using stock market reactions to rating change announcements as a market-based benchmark, the chapter shows that rating categories are not equally spaced, particularly around the investment-grade threshold. Motivated by this evidence, it proposes a new market-based numerical representation that improves explanatory power and reduces bias in empirical applications.
Overall, the thesis highlights the importance of aligning empirical measurements with economic structure to obtain more accurate and informative results in financial research. | el |
