Realizing Value - Empirical Evidence on Multivariate Fundamental-based Investment Strategies

Abstract: This thesis examines whether fundamental-based indicators can build the foundation of a zero-cost portfolio strategy that earns statistically significant excess returns. Our empirical analysis can be divided in two steps. First, we separately investigate a wide range of prominent accounting metrics, related to core value investing criteria, on their individual characteristics and predictive ability over future stock returns. We find that gross profits-to-assets, equity-to-assets, current ratio, gross profit growth and book-to-market ratios are best suited to cover these fields. In a second step, we thus build a zero-cost portfolio strategy based on a combination of these five metrics. Such a strategy offers statistically significant average monthly excess returns of 1.26% over the observed sample period (July 1966 to December 2016). The performance analysis displays that observed returns remain largely unexplained by the CAPM and Fama & French (1993, 2015) risk factor models.