Universidad Pontificia Comillas. Madrid (España)
21 de septiembre de 2026
Resumen:
The relationship between firm characteristics, risk, and realised returns remains fragile and sensitive to modelling choices. In particular, the use of ex-post returns as proxies for ex-ante required returns, together with the reliance on segmented portfolio constructions and the limited treatment of interactions, hinders a unified interpretation of asset pricing mechanisms. This is a central question in asset pricing and extends to Corporate Social Responsibility (CSR) as a trait defining the firm. In this context, the literature offers mixed evidence on the relation between CSR and return and risk. To address these limitations, in the context of factor investing, this thesis develops a new factor model in which factor returns are reformulated as graded factors—cross-sectional slopes— using the full distribution of firms rather than segment averages. This eliminates arbitrary breakpoints, incorporates interactions structurally, and provides a mathematically consistent link between characteristics and returns and risk metrics. Its definition allows for a more advanced and direct analysis of valuation dynamics. By decomposing returns into earnings and price–earnings multiple components and with the implicit cost of capital (ICC) as a forward-looking proxy for required returns, the framework yields a unified model linking cross-sectional and time-series drivers of returns. Putting the framework to the test, we show that CSR characteristics are robustly associated with lower risk across specifications and time periods, but do not yield stable return premia. Second, we prove that ICC aligns consistently with firm characteristics in theoretically expected directions, supporting its role as a proxy for required returns. Third, our test underscores the importance of interactions in computing jointly operating factors. Fourth, we showcase the need to examine the time profile of factor returns. For example, size and low reinvestment display stable, time-dependent premia, whereas value and greenness are linked to episodic shifts in investor preferences. We then illustrate the causal channels by decomposing returns into earnings and valuation multiples. Both contribute over long horizons: the overperformance of smaller firms is largely driven by multiple expansion, while that of low-reinvesting firms is more closely tied to realised earnings. The time-series analysis shows that changes in required returns are the primary driver of valuation dynamics, while announcement-related shocks are the main channel through which new information affects prices. Overall, the thesis contributes to asset pricing by integrating valuation mechanics with factor theory in a consistent causal chain. The results have implications for asset pricing, investment, and corporate finance, providing a more structured basis for risk attribution, valuation, and the interpretation of apparent abnormal returns.
Resumen divulgativo:
Esta tesis aborda una cuestión central en la valoración de activos: si la responsabilidad social corporativa (RSC) incide en el riesgo y en los rendimientos, y cómo las características empresariales influyen en la distribución transversal y temporal de los rendimientos.
Descriptores: Gestión Financiera
Palabras clave: Asset Pricing, ESG, Corporate Social Responsibility, Risk, Firm Valuation, Factor Returns, Factor Investing, Cross-sectional Returns, Expected Returns
Cita:
J. Merladet, "The Impact of Corporate Social Responsibility on Firm Risk and Valuation", Tesis Doctoral, Universidad Pontificia Comillas, Madrid, España, 2026.
