A version of the Black and Scholes dividend yield experiment is conducted, with a view to determining whether there exists a preference for dividends versus capital gains. Using data from the 1960s the experiment reveals a relatively large relationship between dividend yields and risk-adjusted...

In reply to Graham, Johnson and Schnabel (1977), we point out that they have raised a relatively difficult question in scientific method. Provided our earlier work is accepted as operating in the context of the two-period model, we reassert that it revealed no anomalous evidence.