Implied cost of capital based investment strategies: Evidence from Western-Europe

Abstract

Due to the limitations of traditional approaches to estimating expected returns, there have been attempts to generate alternative ways to estimate expected stock returns, such as an implied cost of capital (ICC). The term ICC was initially coined by Gebhardt et al. (2001), and it is defined as the internal rate of return that equates a firm’s share price to the firm’s expected cash flows.

The objective of this thesis is firstly to assess the connection between the average of ICC estimates of stocks included in a portfolio and the future realised returns of a portfolio. Secondly, this thesis aims to evaluate whether the ICC estimates can be used to generate trading strategies that yield positive abnormal returns before and after transaction costs. Besides these main objectives, this thesis analyses the factor loadings of the ICC-based portfolios.

The methodology used to achieve the objectives of this thesis consists of three main sets of methods: Firstly, to calculate stocks’ ICC estimates, I follow the approaches introduced in earlier literature and especially the methodologies used by Gebhardt et al. (2001) and Claus and Thomas (2001). Secondly, to construct the ICC-based portfolios and to calculate the absolute and risk-adjusted returns of these portfolios, I follow the methodology used by Esterer and Schröder (2014). Thirdly, to calculate the returns of ICC-based portfolios after transaction costs, I estimate the transaction costs following the methodology of Barber et al. (2001).

The results regarding my main hypotheses imply that investing in monthly rebalanced portfolios comprising stocks with very high ICC estimates generates statistically significant four-factor alphas both before and after transaction costs. On the other hand, observing the returns of portfolios that include stocks with ICC estimates closer to the sample average implies that there is no clear relation between the ICC estimates and the absolute or risk-adjusted returns of the portfolios. The reason for this seems to be that the so-called ICC effect documented by Esterer and Schröder (2014) is strongest for stocks with ICC estimates clearly deviating from the sample average.

My thesis contributes to the existing literature in three main ways: It assesses whether investment strategies designed to take advantage of stocks’ ICC estimates can yield positive abnormal returns, both before and after the transaction costs. In addition, this thesis contributes to the literature examining the association between ICC estimates and future realised returns by investigating the association between portfolio-level returns and the average of ICC estimates of stocks included in portfolios. Furthermore, this thesis contributes to the existing literature by analysing the factor loadings of the ICC-based portfolios.