Quantitative equity portfolio management strategy : a combination of fundamental value and risk-managed Momentum

• Autor(a) principal: Dey, Thomas
• Data de Publicação: 2016
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.14/21439
• topic: Domínio/Área Científica::Ciências Sociais::Economia e Gestão

This dissertation examines return predictability from B/M and Momentum for US stocks for the period 1970-2015. Particularly, it investigates whether a simple fundamental screening (F-Score) within the high B/M quintile helps separating winners (financially undistressed firms) from losers (financially distressed firms). Finally, it identifies whether a simple 50-50 combination of HML (High-Minus-Low) and risk-adjusted WML (Winners-Minus-Losers) portfolios generates significant abnormal returns (alpha) for the full sample and sub-sample periods. In accordance with the literature, Fama-MacBeth cross-sectional regressions reveal that Momentum and B/M offer significant and persistent return predictive ability. Conflicting with previous evidence (Piotroski 2000), no return predictability in the cross-section of firms is detected for the interaction term between the F-Score and B/M. Return improvements from conditioning the high B/M quintile on high F-Scores are reduced to the 1976-1996 sample period of Piotroski (2000). Contrary, the target volatility momentum adjustment (Barroso & Santa-Clara 2015) does yield significant risk-return improvements, duplicating the Sharpe-Ratio from the Raw WML portfolio, reducing the maximum drawdown and improving the third and fourth moments of the return distribution. The 50-50 HML and WML* (target vol-atility WML) portfolio strategy significantly outperforms the CRSP market-value weighted portfolio and the S&P500 from 1970-2015, although the outperformance was strongest from 1970-2000. Ultimately, both the pure HML - WML* and the HML_F-Score - WML* com-binations (50-50) generated highly statistically significant abnormal monthly returns of 0.8% when setting the Carhart Four-Factor Model as the relevant asset-pricing model benchmark.