Evaluating and improving methods to estimate the implied cost of capital and return decompositions

The dissertation analyzes existing accounting-based methods to measure expected stock returns and proposes methodological advances to improve expected return estimation. The primary object of study are methods that estimate the implied cost of equity capital (ICC). The analysis includes a simulation study to provide direct evidence concerning the quality of existing ICC methods in capturing the true cost of capital. The analysis shows that the structure of the underlying valuation models is a crucial determinant for the quality of ICC methods. Utilizing a large empirical dataset, the dissertation proposes a new approach to estimate the implied cost of equity capital that addresses specific problems of existing ICC methods. Finally, the dissertation presents a decomposition of realized stock returns based on implied cost of capital estimates and applies the return decomposition to shed light on the return anomaly that the quality of corporate governance helps to predict stock returns