The version of the CAPM studied in this chapter specifies a simple regression model as
100 (St – rt) = α + 100 β (Mt – rt) + ε
where Mt are the returns on the market, St are the returns on the stock, and rt are the returns on risk-free investments. (See About the Data.) Hence, 100(Mt – rt) are the excess percentage changes for the market and 100(St – rt) are the excess percentage changes for the stock. What happens if we work with the excess returns themselves rather than the percentage changes? In particular, what happens to the t-statistic for the test of H0: a = 0?