Erasmus university rotterdam

Methodology

In this part of the article the methodology to test the theoretical models is defined. Several parameters have to be computed in order to be able to test the versions of the CAPM. The methods of constructing the parameters of the models are described. To facilitate comparison across countries ,the analysis is conducted from the perspective of a US-based internationally diversified investor.

Global risk free rate

The appropriate risk free rate is that at which an investor can tie his money at the current point in time in that market (Pereiro, 2002). The US is considered by many as the most efficient market in the world. As a result the US Treasury bill is frequently used as proxy for the risk free rate. Since there is more than one choice of risk free rate, this study will use the most frequently used option. This will make comparison with earlier research more straightforward. A 10-year US Treasury bond is used as a proxy for the risk-free rate, the rates are converted to weekly frequency. In doing so, investors are expected to price stock based on anticipated US dollar returns. This approach to the global CAPM does not ignore currency risk because currency risk is included in the regression (Bodnar et al, 2003). Bodnar et al. (2003) suggest that currency risk should be included in the global CAPM as a separate factor. However, this is not done in this study. Koedijk et al. (2002) show that les than 4% of a stocks total variance are explained by currencies.

There is no agreement among academics on the true historical value of the US risk-free rate. Pereiro (2001) finds a median return for 30-year T-bonds for the period 1993-1999 of 6.72%. The median return for the 10-year US Treasury bond found in this article is 6% for the period 1987-2009, when looking at the period 2001-2009 an average of 8.3% is found.